Acoustics & Music Theory
why it sounds like that
Theme

What the listener supplies

A downbeat, a key, a category, a preference — none of them is in the sound, and every one of them can be measured. These essays are about the part of music that is contributed by the person listening, which is also the part that varies between listeners and between traditions.
A 7-semitone sequence at 120 ms a tone. Tones drawn as pitch against time, one bar per tone. The events are the same in both readings of this pattern; what changes is whether a listener assigns them to one line that leaps back and forth or to two lines that each stay put. Nothing in the drawing decides which, and nothing in the sound does either. Perception and the listener

The ear builds objects, and sometimes offers a choice

What arrives at an ear is one pressure signal. What a listener gets is a set of separate things — a violin, a voice, a car outside. The assignment is a construction, and the clearest evidence is that it can be flipped by changing nothing but the speed: one sequence of tones is a single line when slow and two lines when fast, with a wide region in between where the listener may choose.

 Where A has been. Documented pitch standards and surviving instruments, plotted as cents from A440. The extremes are 392 Hz and 465 Hz, which is 296 cents apart — 3.0 semitones, close enough to a minor third that a piece written at one and played at the other is in a different key. Nothing here is a preference; each is a decision somebody recorded. Pitch and tuning

A memory for the note itself, and it is dated

Absolute pitch is usually described as a rare perceptual gift. It is better described as a memory for a convention — and conventions have dates. Possessors trained on A=440 mis-name Baroque pitch by a semitone, their own labels drift sharp with age, and meanwhile most listeners without it start familiar songs within a semitone of the record.

 Where one interval stops being itself. Identification as a function of interval size: the probability that a listener names each category, modelled as a logistic with the boundary positions and sharpness a study reports. One category's share falls from three-quarters to one-quarter over 24 cents, against the 100 that separate adjacent categories — so the change of mind happens in 24% of the gap and the rest of it is not in doubt at all. That is what makes a mistuned third a third that is out, rather than a different interval. Perception and the listener

The chord is still major, and that is why temperament works

A major third can be seventeen cents wrong and still be a major third. That tolerance is not a failure of hearing — it is the reason the whole subject of tuning is a discussion rather than a catastrophe. Every temperament ever proposed moves intervals around inside their categories, and the one thing none of them may do is push one across a boundary.

 The probe-tone profile, major key. How well each of the twelve pitch classes was rated as fitting, after a context establishing the key — Krumhansl and Kessler, 1982. The shading is not part of the measurement: it is the tonic, the rest of the tonic triad, the rest of the scale and the remaining five notes, which are categories this subject had before anybody ran the experiment. The profile separates all four without overlap. Perception and the listener

Counting produced the hierarchy

Ask listeners how well each of the twelve notes fits after a passage in C major and the answers are not a smooth gradient. They fall into four groups with no overlap at all: the tonic, then the rest of the tonic triad, then the rest of the scale, then everything else — categories the subject had names for centuries before anybody ran the experiment.

 Two models of consonance, and where they disagree. Every chord scored twice: horizontally by summed Plomp–Levelt roughness, vertically by the largest integer needed to write it as members of one harmonic series. Both are supposed to be measuring consonance and both are computed here from the chord itself. They correlate, but not tightly enough to be the same claim, and the chords furthest from the diagonal are the ones any experiment has to be run on. Perception and the listener

Consonance is half learned, and this is the half

This site's founding claim is that consonance is small whole numbers. Two computable models say so and they disagree about which chords — which is already awkward. The cross-cultural evidence is worse: listeners with little exposure to Western music discriminate roughness exactly as anyone does, match octaves exactly as anyone does, and rate consonant and dissonant chords as equally pleasant.

 The metrical levels of 120 bpm, against the window. The range of inter-onset intervals that can be heard as a beat at all, from about 100 to 2000 milliseconds, with the preferred rate near 550. Each mark is one metrical level of a piece at 120 beats a minute. Which of them a listener taps is decided by which falls nearest the preferred rate, not by which one the notation calls the beat. Rhythm and metre

The beat has a preferred rate, and it is not the notation's

Metre is a hierarchy of levels and only one of them is tapped. Which one is decided by the listener's own clock rather than by the time signature: the window in which a series of events can be a beat at all runs from about a tenth of a second to two seconds, with a marked preference near half a second — so at the extremes of tempo the notated beat and the felt one part company, predictably.

 A bell tuned to 294 Hz, and the note it is heard at. The partials of a well-tuned church bell, as ratios to the prime, with the three that imply the strike note marked. The nominal, twelfth and double octave sit at 2, 3 and 4, which is a harmonic series on 1 — so they imply a fundamental at 294 Hz, an octave below the loudest partial the bell has. The tierce at 1.2 is a MINOR third above the prime, which is why a bell has a minor quality by construction rather than by choice. Instruments and their design

A bell has no fundamental

The note a listener names when a church bell is struck is not any partial the bell has. Its nominal, twelfth and double octave sit at 2, 3 and 4 times the prime, which is a harmonic series on a pitch an octave below the loudest thing in the sound — and that pitch is supplied by the listener. Founders have been tuning it by ear since the fifteenth century.

 What is struck, and where its partials land. Partial ratios for an ideal string, an ideal membrane, a kettledrum, drawn on a logarithmic axis so that a whole-number ratio is a fixed distance from the last. An ideal string: 1, 2, 3, 4, 5, 6. An ideal membrane: 1, 1.59, 2.14, 2.29, 2.65, 2.92. A kettledrum: 1, 1.50, 1.99, 2.44, 2.89. The membrane's are Bessel zeros — a derivation, and they land nowhere near the whole numbers. The kettledrum's have been pulled toward 2 : 3 : 4 : 5 by the enclosed air, on a fundamental an octave below the lowest partial present. Instruments and their design

What a drum is doing instead

An ideal membrane's modes are zeros of Bessel functions — 1, 1.59, 2.14, 2.30 — which support no common fundamental, so a drum rings without a note. A timpani is that problem solved — the kettle's air and the radiation load drag four modes onto 1, 1.5, 2, 2.5, and the pitch a timpanist tunes is the fundamental those four imply and none of them is.

 Boundaries found by a local operator, at three kernel widths. Foote's checkerboard novelty computed on the self-similarity matrix of thirty-two-bar AABA, at kernel widths of 2, 4, 8 bars. The dashed verticals are where the encoding's sections actually change; nothing about them enters the computation. A peak is a place where the bars before resemble each other, the bars after resemble each other, and the two groups do not resemble each other. Form and structure

The boundary is where the neighbourhood changes

A section boundary can be found by an operator that never sees a section. It walks the diagonal of a similarity matrix asking one local question — do the bars behind me resemble each other, do the bars ahead resemble each other, and do the two groups resemble each other — and where the answer is yes, yes, no, there is an edge. What it cannot find turns out to say more than what it can.

 A phrase is a number of seconds, and the bars follow the tempo. Phrase durations for 1, 2, 4, 8, 16-bar phrases at seven tempos, on a logarithmic seconds axis, with the 2 to 8 second window shaded. The window is a property of the listener and does not move; which bar count falls inside it is decided entirely by the tempo. Form and structure

A phrase is a number of seconds

Musical phrases are described in bars, and four is the number everybody names. But the constraint that fixes a phrase is a property of the listener and is measured in seconds, so the bar count is whatever the tempo makes it. Across seven ordinary tempos the bar count that lands inside the window moves by a factor of eight, while the window itself does not move at all.

 Two ways to fill eight bars, and only one of them accelerates. The period against the sentence, drawn as the lengths of their constituent units against position on a grid of eight bars. The ratio beside each row is the mean unit length in its second half divided by the mean in its first: the period at 1.00, the sentence at 0.67. A ratio below one is an acceleration — the unit shortening as the phrase approaches its arrival — and a ratio of one is a plan whose unit never changes length. Form and structure

One of these eight-bar phrases accelerates

The sentence and the period both occupy eight bars, both end with a cadence, and both are recognised by ear rather than counted. What separates them is arithmetic. One halves its unit halfway through and the other does not, and the difference comes out as a single ratio — 0.67 against 1.00 — computed from nothing but the lengths of the parts.

 Five signals, computed separately, and no total. The five components of closure for 6 chord pairs. The first three are computed from the chords alone; the last two are properties of where the goal lands and how long it is held. There is no total column: the components are not commensurable and the ordering of these cadences depends on which is weighted. Form and structure

What makes an ending an ending

Cadences are ranked. The authentic one is strong, the plagal weaker, the deceptive weaker still — and the solver here measured the quantity that ranking is usually explained by and found it says something else entirely. What survives is not a weaker version of the ranking but a different kind of object, with five components and no total.

 Five signals, computed separately, and no total. The five components of closure for 4 chord pairs. The first three are computed from the chords alone; the last two are properties of where the goal lands and how long it is held. There is no total column: the components are not commensurable and the ordering of these cadences depends on which is weighted. Form and structure

An ending that exists so a bigger one can

Half of the cadences in tonal music are built to fail. A phrase that stopped convincingly at bar four would be a piece four bars long, so the ending at bar four is engineered to arrive and not to settle — and the components it withholds are exactly the ones its partner at bar eight supplies. Closure is nested, and the nesting is what turns two phrases into one thing.

 The same induction, one level up. Bar-level onsets from thirty-two-bar AABA — a bar is marked where a section or a key begins — scored against hypermetres of 2, 3, 4, 6, 8 bars with the identical function the beat-level figures use. The best-fitting period is 8 bars, which at 108 beats a minute lasts 17.8 seconds. Form and structure

The bar above the bar

A four-bar group is a bar whose beats are bars. That is not an analogy — it is the same computation, and the same metre-induction model produces one when it is handed bars instead of beats, unchanged. What decides where the hierarchy of levels stops is not in the arithmetic at all, and it is a number the phrase essay already measured.

 How surprising each chord is, in bits. Each step's information content, −log₂ of the probability the root-motion weights used here give it. a perfect cadence totals 6.4 bits over 3 steps; a deceptive cadence totals 7.3 bits over 3 steps; I – IV – V – vi totals 7.3 bits over 3 steps. The single most surprising move drawn is IV to V at 2.7 bits, which is 42 per cent of everything its passage spends. The eight weights are ordinal and stipulated rather than counted, so these are the numbers that ordering implies and not a measurement of any repertoire. Harmony and voice leading

The chord that did not come

A deceptive cadence is described as a surprise, and the explanation offered is that the wrong chord arrived. Measured against the tonal hierarchy already in use, the wrong chord is the second best-fitting triad in the key — and two of its three voices do exactly what they would have done in the right one. The surprise is not statistical. It is one voice, and it is the bass.

 How often the chord changes, and what a room allows. Chord changes a second implied by each style's stated rate and tempo, on a logarithmic axis, with the rate above which a room leaves more than one earlier chord above 20 dB marked for six rooms. The style rates are conventions rather than corpus measurements and the figure says so; the room rates are arithmetic from the reverberation time. Harmony and voice leading

How often the chord changes

Two pieces can use the same chords in the same order and be nothing alike, because a progression says which chords and not how fast. Harmonic rhythm is the second variable, it runs over a factor of thirty between the styles that use it, and both of its limits are set by things that are not harmony — a listener's memory at the slow end and a building at the fast one.

 A cycle has no ending to compute, so it uses density instead. 5 layers over 32 cycles of a 12-step pattern, with each layer's entry and exit marked, and the onsets per step summed underneath. No chord changes and no cadence occurs; the closure vector is zero throughout, and every change a listener hears is a change in how many things are playing. Form and structure

A cycle cannot cadence

Every component of closure is defined by a first time and a last time. Music built on a repeating cycle has neither, so the whole apparatus returns zero on it — not a small value, zero, at every setting. What such music uses instead is how many things are playing, and that is a curve which can be computed from the onsets and nothing else.

 The 7-note sets in which every interval occurs a different number of times. Each set of 7 notes containing C whose six interval counts are all different, with the counts printed. Every one of them is a rotation of one of two shapes, and only one of the two has steps a scale could use — the other is a run of semitones with the gap at the end. Scales and modes

Every interval a different number of times

Count the intervals inside a major scale and the six answers are 2, 5, 4, 3, 6 and 1 — six different numbers, no two alike. That is not decoration. It means the number of notes a key shares with a transposition of itself identifies the distance uniquely, so a listener who can only count common tones can still tell exactly how far a modulation went.

 The period, as the piece goes by. The strongest lag of thirty-two-bar AABA computed on only the bars heard so far, against how many bars that is. The final answer is 4 bars; it is revised 6 times on the way, and is not reached for the last time until bar 29 of 32, which is 91 per cent of the way through and 64 seconds at 108 beats a minute. Nothing about the boundary operator is involved: this is the global statistic, and it is the half of the form that a first hearing cannot have. Form and structure

The form a first hearing cannot have

Every figure so far was computed with the whole piece in hand. Run the same methods over only the bars already heard and one of the two methods survives intact — the boundary operator turns out to be causal at a fixed delay of a few bars — while the other collapses. The period of a piece is not knowable until the piece is nearly over, and in two of the six schemes here not until its last bar.

 The ranking is settled either side of one narrow band. Remembered repetition — each bar's best match to an earlier bar, discounted by exp(−Δt/τ) with Δt in seconds — for 6 schemes at 108 beats a minute, against the decay constant τ on a logarithmic axis. The order of the schemes changes only between 8 and 13 seconds; outside that band it is fixed, so an estimate of τ wrong by any amount that stays outside it leaves the ranking alone. Perception and the listener

A return has to be remembered

A stripe four bars off the diagonal and a stripe twenty-four bars off it are the same ink and are not the same experience. Convert the lag axis to seconds, discount every comparison by how long ago it was, and the ranking of these six schemes by how repetitive they are changes — and the decay constant and the tempo turn out to enter the arithmetic as one number rather than two.

 What 2 partials imply, and how many answers there are. The partials are at 32.7, 49.0 Hz. Each row is a harmonic series they are consistent with to within 30 cents: the filled dots are the partials in their assigned slots and the open dots are slots the series predicts that nothing occupies. There are 5 such series with harmonic numbers up to 16, the best fitting them to 1.4 cents with a fundamental of 16.34 Hz and no unoccupied slot. The rest sit at one half, one third, one quarter, one fifth of it and their arithmetic is exactly as exact; what separates them is the count of empty slots, which is why a residue pitch built from few partials is reported an octave out by a minority of listeners and not by the rest. Instruments and their design

Two pipes for a note neither makes

The resultant stop sounds a sixteen-foot pipe and a ten-and-two-thirds-foot pipe and asks the listener for a thirty-two-foot note. It is the missing fundamental built on purpose from the fewest partials that can imply anything — and counting how many fundamentals two partials actually imply is the arithmetic behind three centuries of builders disagreeing about whether it works.

 The most a 6 cm cone can make of a low note. The maximum sound pressure level at one metre from a circular radiator of effective radius 3.2 cm, moving 1.5 mm at its limit, in a system resonating at 250 Hz. Below resonance the cone is already at that limit and the pressure a piston makes goes as the square of frequency, so the curve falls at twelve decibels an octave: 66 dB at 40 Hz, 78 dB at 80 Hz, 94 dB at 200 Hz. The 40 Hz figure is 28 decibels below the 200 Hz one, and that gap is arithmetic about a radius and a displacement rather than a property of any particular loudspeaker. The dots are the harmonics of a 41.2 Hz note with a one-over-n source spectrum; the loudest of them is the 6th. Instruments and their design

The bass a small loudspeaker does not make

A three-inch cone at its excursion limit produces sixty-six decibels at forty hertz, which the ear converts to seventeen phons — barely above nothing. The note is heard anyway, because its harmonics are radiated and its fundamental is supplied by the listener. Computing what arrives turns the residue from a curiosity into a design decision, and finds that the fundamental of a low note is not the loudest part of it on any system a listener is likely to own.

 The thirteen intervals at C4, ordered by roughness, against Fux's species, 1725. Every interval within an octave above 262 Hz, ordered by the Plomp–Levelt roughness of a string spectrum — from 0.0029 for the unison to 0.2635 for the minor second — beside whether Fux's species, 1725 counts it a consonance and how many of the thirteen may follow it under that rule set. The best single threshold on roughness misclassifies 2 of the thirteen, among them the major third and the minor third, and the gap it falls in is 4.2 per cent of the roughness either side of it. Harmony and voice leading

A dissonance is what has to be resolved

The perfect fourth is a consonance between two upper voices and a dissonance against the bass, and the same three notes are involved either way. Score both arrangements with the roughness model and the one the rules call a dissonance comes out thirty per cent smoother. Whatever the rule is tracking, it is not the sound.

 One pattern, four metres. The same 6-step onset pattern read under 3 candidate metres, each scored by a preference rule set: 3 for a strong position that carries an onset, -2 for one that does not, -1 for an onset that lands off every strong position. The scores are the 3 is the beat 8, the 2 is the beat 4, every step is a beat 8, so the 3 is the beat and every step is a beat tie and the model does not choose. Nothing about the sound differs between these readings; the bar line is supplied by the listener. Rhythm and metre

Which of the two is the beat

A polyrhythm is notated as a bar of two with three laid across it. Run the metre rules used here over the composite and they choose the three — at every ratio tried, without exception. Add the one other rule the rules have, and the answer changes hands at a bar of 1,013 milliseconds, which is 118 to the minute.

 Syncopation against a bar of 8. An 8-step pattern with 3 onsets, against the metrical weights of its bar. A position's weight is zero on the downbeat and one lower at each level down the subdivision tree, drawn here as the depth of the bar hanging beneath it. A note on a weak position followed by a rest on a stronger one costs the difference. the tresillo, in a bar of eight scores 2 — the note at step 4 against the rest at step 5, costing 2. Rhythm and metre

Syncopation is a number about the metre

Longuet-Higgins and Lee price a syncopation at the metrical weight a note skips over. That makes it computable, and it makes it a property of a pair rather than of a rhythm — the son clave scores 4 read from step one, 2 from step three and 8 from step four, and not one onset has moved. Worse, the induction rules pick very nearly the reading that scores lowest.

 One pattern, four metres. The same 16-step onset pattern read under 4 candidate metres, each scored by a preference rule set: 3 for a strong position that carries an onset, -2 for one that does not, -1 for an onset that lands off every strong position. The scores are downbeat on step 1 0, downbeat on step 2 -12, downbeat on step 3 -12, downbeat on step 4 0, so downbeat on step 1 and downbeat on step 4 tie and the model does not choose. Nothing about the sound differs between these readings; the bar line is supplied by the listener. Perception and the listener

The beat that is never sounded

A listener who has heard four bars of a groove and then hears two bars with the downbeats taken out does not move the downbeat. This site's rule set does, every time, on every pattern tried — and the direction it moves in says exactly what kind of model would be needed instead.

 The seven modes, brightest first. The same seven pitch classes started on each of its degrees in turn, ordered by how many of their notes are raised. Each row differs from the one below it by exactly one note, and that note moves down one semitone each time. Scales and modes

Nothing in the census knows which note is home

All four properties six earlier essays are about are invariant under rotation and transposition — one property tuple over all eighty-four rotations and transpositions of the diatonic set. So the census that separates 349 shapes cannot separate a major scale from its own Aeolian mode, and everything that makes one note a tonic is outside it.

 The price of a tonic. Every note of Dorian is given the same duration except its tonic, which is lengthened; the horizontal axis is the share of the total that goes to it. The key-finder answers with the parent key until 25.0 per cent of the time is spent on the modal tonic, and with D minor above it. At the left-hand edge every note has equal weight, which is the pitch-class set itself — and with every weight identical the correlation is not merely low but undefined, because a flat histogram has no variance to correlate with anything. Perception and the listener

What a tonic costs in seconds

The standard key-finding algorithm cannot be run on a pitch-class set at all — a flat histogram has no variance and the correlation is undefined. Give it durations and it answers with the parent key for all seven modes identically, and it takes between 15.8 and 30.0 per cent of the total time spent on one note before it names that note instead.

 How many bars a key change takes to be heard. A twelve-bar progression that moves to G major at bar 6, read by the same correlation against all twenty-four profiles, with a window of 3, 4 and 8 bars. With 3 bars of history the new key is never the answer at all. With 4 bars of history the answer is G major from bar 7, one bar late, and it holds it from there. With 8 bars of history the answer is G major from bar 9, 3 bars late, and it holds it from there. The pivot bar is ambiguous by construction — it belongs to both keys, which is what makes it a pivot — so the lag is not a defect of the algorithm but a statement about how much evidence a key is. Perception and the listener

How much evidence a modulation needs

Run a key-finder bar by bar over a progression that moves to the dominant at bar six. With four bars of history the answer becomes the new key at bar seven and holds. With three bars it never gets there at all, and reports E minor and B minor on the way. The window decides the lag as much as the music does.

 Long-term average spectra: an orchestra, playing forte against a trained operatic soloist. Each source's mean spectrum over a long passage, in decibels below its own strongest region, on a logarithmic frequency axis. An orchestra, playing forte peaks at 250 Hz and is 30 dB down by 3,150 Hz; a trained operatic soloist peaks at 250 Hz and is 11 dB down by 3,150 Hz. The shapes are the same until about 1 kHz and separate above it: at 3153 Hz the difference is 19.0 decibels, which is the largest anywhere in the range. Nothing here is about level. Both curves are drawn against their own peaks, so what is being compared is shape. Timbre and acoustics

One voice over ninety players

A soloist heard over a full orchestra is not louder than it and could not be. What the trained voice does instead is put a peak of energy at three kilohertz, which is where the orchestra's spectrum has already fallen away and where the ear's own threshold happens to be lowest. Nineteen decibels of advantage, in a place nobody is competing for.

 a major triad, C–E–G. The partials are at 261.6, 329.6, 392.0 Hz. Each row is a harmonic series they are consistent with to within 30 cents: the filled dots are the partials in their assigned slots and the open dots are slots the series predicts that nothing occupies. There are 2 such series with harmonic numbers up to 16, the best fitting them to 10.1 cents with a fundamental of 65.54 Hz and no unoccupied slot. The rest sit at one half of it and their arithmetic is exactly as exact; what separates them is the count of empty slots, which is why a residue pitch built from few partials is reported an octave out by a minority of listeners and not by the rest. Intervals and chords

The root an ear supplies

A major triad's notes fit 4:5:6 with nothing missing and a fundamental two octaves below the bass. A minor triad's fit two different series with two different answers a major sixth apart, and the model cannot choose between them. The ambiguity theorists argued about for two centuries is a computable quantity, and the spectrum invented to remove it is one no object produces.

 How much of the rule a walk with no rule reproduces. Post-skip reversal in 20,000-note random walks with no melodic knowledge of any kind. An unbounded walk reverses after 50.0 per cent of leaps, which is the chance rate and is the check that the measurement is right. Confining it to 12 semitones raises that to 61.3 per cent. Reaching the 70 per cent that corpus studies report needs a central tendency of 0.95 — an almost deterministic pull back toward the middle at the edges of the range. A wall is not enough; there has to be a spring. Form and structure

The leap that pays itself back

Every melody textbook teaches that a leap should be followed by a step in the opposite direction, and every corpus that has been counted agrees — around seven leaps in ten are answered that way. A random walk with two walls, no memory of the leap and no rule of any kind reverses after 61 per cent of them, and the residue is not a rule either. What is left when the walls are accounted for is a prediction the rule does not make, and it is the prediction that decides between them.

 The arch is not a preference. Every sequence of 6 notes over 8 scale degrees — 262,144 of them, enumerated rather than sampled — classified by contour, under three constraints. With none, the nine classes are spread. Requiring the sequence to return to its starting degree leaves only the arch, the valley and the flat, at 42.2 per cent each for the first two. Requiring it to begin and end on the LOWEST degree leaves the arch alone, at 99.6 per cent. Nothing here prefers a rise followed by a fall; the constraint is that the melody comes home, and a melody that comes home from below has nowhere to go but up first. Form and structure

The shape that survives everything else

Throw away a melody's key, its tuning, its instrument and the sizes of its intervals, and what is left is a string of pluses and minuses. That string is what a listener who cannot name a note still has, and it costs 37 per cent of the tune to keep. The arch that melodic shape is famous for is not in it as a preference: enumerate every six-note sequence that begins and ends on the lowest degree it uses and 99.6 per cent of them are arches, because a melody that comes home from below has nowhere to go first but up.

 The vowel in "hod", sung at 110 Hz. The partials of a 110 Hz note, each drawn at the amplitude the vocal tract's resonances give it. The peaks of the curve are the formants — 730 Hz and 1090 Hz — and they stay where they are when the pitch changes, because they are a property of the shape of the mouth and not of the note being sung. Timbre and acoustics

The sound a listener knows best

A voice is recognisable across every vowel it says, across two octaves of pitch, down a bad telephone line and in a whisper where there is no pitch at all. Nothing that survives all of that can be a frequency. What survives is a ratio: the resonances of a vocal tract are set by its length, so a shorter tract multiplies every formant by the same factor, and identity is a scale on the spectral envelope rather than a position within it. Between an adult man and a child the whole pattern moves by a fifth, and the vowel does not change at all.

 The same eight notes, read three ways. The scale, eight quavers, scored against every triad and seventh at every root. Barred as written the best reading is C major7 at 0.850; with the barline one quaver later it is D minor7 at 0.850. With no metre — every note weighted the same — 4 readings tie at 0.625 and the passage has no best analysis at all. The notes are identical in all three. What changed is where the bar starts, which is not a fact about harmony. Harmony and voice leading

Which notes are the chord

A progression is a list of chords, and before there is a list something has to decide which of the notes sounding are chord tones and which are passing. Take the eight notes of a scale as eight quavers and score every triad and seventh at every root: barred as written the best reading is C major seventh, with the barline moved by one quaver it is D minor seventh, and with no metre at all three readings tie exactly and the passage has no best analysis. Same eight notes in all three. Harmonic analysis is a function of a variable that is not harmony.

 One pattern, four metres. The same 12-step onset pattern read under 3 candidate metres, each scored by a preference rule set: 3 for a strong position that carries an onset, -2 for one that does not, -1 for an onset that lands off every strong position. The scores are 3/4 -4, 6/8 12, 2/4 -4, so 6/8 wins. The page prints 3/4, which the model does not prefer — it ranks 6/8 above it. Nothing about the sound differs between these readings; the bar line is supplied by the listener. Rhythm and metre

The time signature is a claim

A bar line is not a measurement. It is a claim about where the accents are, made before the sound exists, and there is a metre-induction model here that can be handed the same onsets and asked whether it agrees. On a hemiola it does not: the page says three and the model says six, by a margin of twelve against minus four. And on a bar of seven the signature is not even a candidate — what decides the reading is the beaming, which the signature does not contain.

 What a contour costs to remember. A melody of n notes over 8 degrees carries 3 bits a note. Its contour carries fewer, and fewer than the number of distinct contours suggests, because the contours are not equally likely: at 6 notes there are 243 of them but the entropy is 6.59 bits, an effective alphabet of 96. Each further note adds 1.28 bits of contour against three of melody, so the shape keeps a stable 37 per cent of what is there however long the tune. Perception and the listener

The part of the tune that is kept

Contour survives transposition, retuning, a change of instrument and a doubling of every interval, and the usual explanation is that it is what a listener retains. That can be counted rather than assumed. A six-note melody over eight degrees carries eighteen bits; its contour carries 6.59 — not the 7.92 the number of distinct shapes suggests, because the shapes are wildly unequal — and the effective alphabet is ninety-six out of two hundred and forty-three. Each further note adds 1.28 bits of shape against three of melody, and at about nine notes a contour is specific enough to pick one tune out of a thousand.

 The pitch moves and the repetition rate does not. Three partials around harmonic 10 of 200 hertz, shifted together by up to 200 hertz, with three curves. The flat line is the envelope repetition rate, which the shift cannot move at all. The rising line is the shift divided by the harmonic number — 200 hertz becoming 220.0 — which is what a harmonic template predicts and is what listeners report. The third curve is the next-best template, which overtakes the first partway along: the pitch is ambiguous, and it drops back rather than rising indefinitely. Perception and the listener

The pitch that moves the wrong distance

Take three partials two hundred hertz apart and move every one of them up by forty. The spacing has not changed, so anything reading the pitch off how often the waveform repeats must give the same answer as before. The pitch moves to 204 — the shift divided by the harmonic number — which is what a harmonic template predicts and what listeners report. Push the shift to a hundred and a second reading overtakes the first, so there are two pitches and neither is the spacing. This is the measurement that closes the question, and it closes it by ruling out one mechanism rather than by choosing between the two that are left.

 The mode is its tritone. Every set of degrees that lies inside at least one of the seven modes — 510 of them — scored by how many modes it leaves standing, with the 15 minimal sets that decide printed in full. All 15 have two members and every one of them either is a tritone or contains the tritone above the tonic. The only degree every mode has is the tonic itself; there is no degree belonging to just one mode. Averaged over every order the degrees could arrive in, 5.67 of the seven have to have been heard before the mode is settled — most of the scale, because the deciding pair arrives when it arrives. Scales and modes

The one note that decides the mode

A key signature names the seven and not the rotation, so the mode has to be heard. Enumerate every set of degrees that lies inside at least one of the seven modes — 510 of them — and count how many modes each leaves standing. Fifteen minimal sets decide, every one has two members, and every one is the mode's own tritone. The only degree all seven share is the tonic; no degree belongs to a single mode; and averaged over the orders the degrees could arrive in, 5.67 of the seven have to have been heard first.

 Where the page ends a phrase, and where the ear does. Twinkle, twinkle with two sets of phrase boundaries on it. The lower curve is a local boundary detector — a peak in how much the interval and the note length change from one to the next, with nothing in it about bar lines or harmony — and the marks above it are where the notation puts the phrase ends. It finds 100 per cent of them and 2 boundaries the page does not have. Where the two agree it is because a long note is sitting at the join; where they disagree the page is marking a grammatical unit and the detector is finding a perceptual one. Form and structure

Where a phrase ends

Run a boundary detector over the three tunes used throughout and it agrees with the notated phrasing on one of them perfectly and on another almost not at all. The reason is which cue each tune uses: Twinkle's phrases all end on a long note, so a duration-weighted detector finds five of five with no false alarms; Ode to Joy's run on in crotchets and its phrasing is in the intervals, where a duration detector finds one of three and a pitch detector finds all three and eight others. No fixed weighting serves both, and the published one is worse on each tune than the single cue that tune uses.

 note length against the onsets. Every candidate metre's fit to a 16-step pattern with 4 onsets, plotted against how strongly note length is weighted. At a gain of zero the scoring is the onset-only one every earlier model used, and the winner is step 1 and step 4. At a gain of 0.05 the answer becomes step 4. 2 candidates are exactly flat — step 2 and step 3 have no onset on any strong position, so there is no credit for the cue to multiply and no weighting of it can move the line. Form and structure

What the onsets left out

Eight essays induce a metre from a list of ones and zeros, and every failure they recorded was argued about as a failure of the rules. Two of the three are not. Note length is already in that list and the scoring throws it away: put it back and the son clave's two-way tie resolves to the notated downbeat. But the groove with its beat removed cannot be repaired by any cue at any strength, and the reason is arithmetic rather than empirical — the true phase has no onset on any of its strong positions, so there is no credit for a cue to multiply and its line is exactly flat.

 A model that cannot say two keys does not say it is unsure. I – IV – V – I played in C major and in a second major key at the same time, with the second key moved round the circle of fifths. For each separation: the correlation the standard key-finder gives its single best answer, and the correlation reached by the best PAIR of key profiles — a hypothesis the finder does not have. The pair recovers both keys that are sounding at every separation, 7 of 7. The single answer names neither of them at 4 of the 7, and its confidence does not fall when it is wrong: at four steps apart it reports E minor at r = 0.886, against 0.959 for the same progression in one key. Harmony and voice leading

Two keys at once

Every key figure so far assumes one key is sounding, and the standard key-finder has no value it can return that means two. Play one progression in C and the same progression a major third away at the same time, and the model does not report uncertainty: it reports E minor, at a correlation of 0.886, against 0.959 for the same progression in one key. It is as confident as it ever is, and neither key it names is being played. Give it the missing hypothesis — pairs of key profiles rather than single ones — and it recovers both keys at every separation, all seven of seven.

 Three ways a category boundary could move, and how far each moves it. The predicted shift of one boundary against how strong the context is, for three mechanisms. Expectation alone — a listener who thinks one category 20 times more likely than the other — moves the optimal boundary by σ²·ln(odds)/Δ, which with the eleven-cent noise used here is 2.8 cents at ten to one and 3.6 at 20. Re-learning the centres from a context 30 cents away moves it by half of that, 15 cents. Selective adaptation moves it the OTHER way. The two directions are what an experiment would separate, and no absolute calibration is needed to do it. Perception and the listener

The boundary that barely moves

Every identification figure here has fixed category centres, and the essay before this one ended by admitting that real boundaries are supposed to move with context. Three mechanisms could move one, and their predictions are an order of magnitude apart and in two different directions. Expectation on its own — a listener who thinks one interval twenty times more likely than the other — is worth three and a half cents.

 How fast two keys can alternate before the finder stops following. The share of bars a moving key-finder names correctly, once its reading is shifted back by its own lag, against how many bars each key holds for. One line per window. Below a block of three bars the second key is never named at all — 2 of the sweep's readings report a single key for the whole passage — and above about twice the window the tracking is over ninety per cent. The lag itself is about half the window: 0 bars at a window of 3, 0 bars at a window of 4, 3 bars at a window of 8. Harmony and voice leading

The alternation a key-finder cannot follow

Two keys sounding together are not in the key-finder's vocabulary, and an earlier essay ended by pointing at the other case and saying what was missing: a passage whose alternation rate can be varied while everything else is held still. Built, it gives a rule with three numbers in it — the second key is never named below a block of three bars, the tracking clears ninety per cent above twice the window, and the reading is late by half the window throughout.

 What the joint search changes, and what it never changes. Over 552 constructed passages of eight slots with rests, how often the joint reading differs from the pipeline's. The chord differs in 29 per cent and the barline in 31, with both differing in 20. The key differs in 0 per cent — never — because the key is read from a pitch-class histogram, which does not know where the bar starts or which notes are chord tones. Two of the three decisions are entangled and the third is not. Harmony and voice leading

Three decisions that constrain each other

Every model here decides one thing at a time — the key from the pitch classes, the metre from the onsets, the chords from the metre — and an earlier essay ended by saying a listener does all three at once. Resolving them jointly costs a hundred and fifty-seven times the search and changes the reading of two passages in five. It never once changes the key, and the reason it cannot is the reason the whole account is built the way it is.

 Which note a chord would rather have twice. Every complete four-part voicing of each chord inside the SATB ranges, grouped by which member sounds twice and scored for roughness — 480 voicings for a triad. The order for a major triad is root < fifth < third, which is the rule every part-writing treatise states. For a minor triad it is fifth < root < third, which is not. The numbers printed under each bar are the mean error, in cents, with which the four sounding notes fit a single harmonic series, and that measure separates the three far more sharply than roughness does. Intervals and chords

The note that sounds twice

A triad has three notes and a four-part texture has four voices, so one note is doubled — and the voicing model used here leaves the choice free because the rules have an opinion about it. Asked properly, the arithmetic agrees with the treatises for the first time in nine essays: root, then fifth, then third. For a minor triad it does not agree, and for a symmetric chord it correctly has nothing to say.

 One envelope, and the three places a listener might be said to hear it. The amplitude envelope of a note with a 90 millisecond exponential attack, with the three criteria the literature offers drawn across it. The heard moment is 8.0 ms at the detection criterion, 28 ms at the perceptual-onset criterion and 94 ms at the perceptual-attack criterion. The physical onset is at zero on this axis and no criterion puts the heard moment there. The buttons play this attack against a two-millisecond one, started at the same instant. Rhythm and metre

A note is heard after it starts

Every rhythm essay until now has treated a note's onset as the moment it happens. It is not: the instant a listener aligns a note with a beat is later than its physical start by an amount the note's own attack decides, and for a sung or bowed note that amount is about thirty milliseconds — the size of the whole quantity six essays on microtiming set out to measure.

 How much correlation it would take to matter. The limen of a 7-semitone interval at a note length of 0.25 seconds, against the correlation between the two notes' errors. The independent model at the left gives 9.44 cents. Halving that needs a correlation of 0.75; a fifth off it needs 0.31. The curve is √(1 − ρ) and nothing else, so the correlation required for a stated improvement is arithmetic — which turns the question from “does a key help?” into “by how much, and here is the number it must reach”. Intervals and chords

How much an anchor would have to be worth

Two pitch errors added in quadrature assume an independence nobody measured — a listener inside a key hears a note as a scale degree, and a shared reference is exactly a correlated error. Turning the dial is not evidence. What is evidence is that the dial is not free: a shared error cancels out of a difference completely, so a listener's single-note limen and their interval limen give the two components with nothing left over, and halving the interval limen needs a correlation of exactly 0.75.

 Every result so far, against the listener's own noise. Three findings drawn against the one parameter all of them assume: how finely the listener resolves a pitch. At 11 cents — a trained listener, and the value every earlier essay used — the octave holds 6 nameable categories, twelve equal ones are named right 91 per cent of the time, and a 20-to-one expectation moves a boundary by 3.6 cents. At 35 cents it is 2 categories, 72 per cent, and 37 cents. The capacity falls roughly as one over sigma and the shift rises as its square, so the three curves separate rather than moving together. Scales and modes

The listener the model was never run for

Five earlier essays rest on one number — how finely a listener resolves a pitch — and every one of them used a trained listener's eleven cents. The model's dependence on it is not gentle: the capacity goes as its reciprocal and the expectation shift as its square, so an untrained listener at thirty-five cents has two nameable categories per octave rather than six, and a foreign tuning system is not mis-transcribed by them but absorbed.

 How much of each spectrum a listener can assemble into one note. Each partial of each spectrum at the harmonic number it is nearest, against the whole-number series that fuses the most of them, with anything more than 1 per cent out marked as heard separately. an ideal string keeps 10 of 10; a piano string keeps 9 of 10; a bell keeps 7 of 8; a bar keeps 2 of 6; a kettledrum keeps 3 of 5. The fundamental is capped at a tenth of the top partial, and the cap is load-bearing rather than tidy: a bell's ratios are all whole multiples of a tenth, so an unconstrained search finds a fundamental twenty-five harmonics down, calls every partial exact, and reports that a bell fuses perfectly. Nothing that high is resolved and the low harmonics of it are not there. Perception and the listener

The spectrum that will not fuse

A partial about one per cent off its harmonic is heard as a sound of its own rather than as part of a note. Apply that criterion to a whole spectrum instead of to one mistuned component and it becomes a count: a piano string keeps nine of its ten partials, a bell keeps seven of eight, a bar keeps two of six. The physics of inharmonicity has had an essay here for a long time. This is what it sounds like.

 Adding parts adds power, and very little loudness. Each part is played at the same level, and the chord is realised every way its parts allow and averaged over them, so the quantity is a property of the texture rather than of one arrangement. Going from 3 parts to 8 adds 4.3 decibels of power and 0.1 decibels of loudness, because the extra parts land in bands that are already occupied — the count of occupied critical bands FALLS from 6.0 to 3.9 as the parts crowd into the same register. Form and structure

The dynamics are in the score already

Count the parts in each bar, realise them in their ranges, put every partial in its critical band, sum the loudnesses and run the result through the two smoothers built earlier. What comes out is a dynamic curve for a piece with no performance in it anywhere — and it says that doubling the number of parts inside a fixed register adds three decibels of power and about one of loudness, because the extra parts land in bands that were already occupied. Let the register widen with the parts and the same arithmetic gives eight phon, which is what a tutti actually is.

 Four endings, and the loudness each produces from the page alone. Short-term loudness through the closing 6 bars of a thirty-two bar scheme, computed from the part count of each bar with no performance data of any kind — the parts are realised every way their ranges allow, every partial is placed in its critical band, and the sum is run through the two loudness smoothers. thins to one arrives at 0.764 of the running impression; full final chord arrives at 0.952 of the running impression; unchanged arrives at 1.000 of the running impression; thins then full arrives at 0.929 of the running impression. The result worth the figure is that full final chord is not the loudest: adding parts to a final chord adds power and almost no loudness, because the extra parts land in critical bands the chord already occupies. An ending is made loud by contrast with what preceded it, not by thickness. Form and structure

A final chord is not made loud by adding to it

An earlier essay on closure said the loudest cue an ending has needs a corpus rather than an arithmetic. The arithmetic was built one essay ago, so it does not. Run four ending textures through it and two things come out backwards: a final chord three parts thicker than the rest arrives *quieter* against the running impression than the passage it ends, and a texture that drops a part a bar does not get quieter at all until the bar where there is one part left.

 What the notes in between do to the anchor the interval is measured against. How finely a 7-semitone interval can be judged when its two notes are separated by other notes rather than by silence, under the two published accounts. Confirming material restates the key and refreshes the shared reference, so the correlation climbs from 0.5 toward a ceiling and the limen falls to 4.81 cents. Overwriting material competes for the same memory, so the correlation decays to 0.04 and the limen rises to 9.28. By 8 notes the two accounts differ by 4.5 cents, which is 47 per cent of the limen with no anchor at all — and no experiment here distinguishes them. Intervals and chords

The notes in between

Every figure until now is about two notes with nothing between them, and a melody is notes with other notes between them. Two published accounts of what the intervening material does predict opposite signs — one says the key is restated and the shared reference is refreshed, the other says each note competes for the same memory and it decays. By eight notes they differ by four and a half cents, which is nearly half the limen the interval would have with no anchor at all.

 The one number the ordered key-finder was tuned on. For each rate of alternation between two keys, the cost of changing key at which the model stops hearing two keys and starts hearing borrowed chords in one. The threshold rises with the period — 0.95 at 1 bar, 0.95 at 2 bars, 0.95 at 4 bars, 2.00 at 8 bars, 3.50 at 16 bars — so the parameter and the rate trade off against each other exactly. The value tuned earlier, 2.2, sits above every threshold on this axis, which means its verdict about fast alternation was a consequence of the tuning rather than a finding about the music. Filled means the model names two keys; hollow means it names one and calls the rest borrowings. Scales and modes

A modulation and a borrowing are one number apart

The key-finder that keeps the order has one tuned parameter, and said so. Sweep it and the parameter turns out to be the whole verdict: below a threshold the model hears two keys alternating, above it one key with borrowed chords. The threshold rises with how slowly the keys alternate — and the value it chose sits above every threshold in range, so its finding about fast alternation was a consequence of the tuning.

 Two kinds of evidence, each measured against its own chance. Every key as a point: across, how many standard deviations its cadence count is above what the same bars resampled would give; up, the same for its pitch-class profile correlation. The winner on cadences is key C at 6.2 standard deviations and on profile C at 1.1, and they agree. Standard deviations above chance are the same unit whatever produced them, which is the commensuration this figure is for — and it costs something: it assumes the two chances are equally interesting, which is a weighting in disguise. The obvious null does not work at all for one of the two: shuffling the bars leaves a pitch-class histogram exactly as it was, so its spread is zero and every key scores nothing against it. Harmony and voice leading

A count and a correlation

Cadence evidence is a count of ordered pairs and profile evidence is a correlation with a template, and the cadence essay refused to total them because they are not in the same units. Score each against its own chance and they are — standard deviations above chance are the same unit whatever produced them. Then the trouble moves: the obvious null does nothing at all to one of the two, because shuffling the bars leaves a pitch-class histogram exactly as it was.

 How surprising each chord is, in bits. Each step's information content, −log₂ of the probability the root-motion weights used here give it. a perfect cadence totals 6.4 bits over 3 steps; a deceptive cadence totals 7.3 bits over 3 steps; I – IV – V – vi totals 7.3 bits over 3 steps. The single most surprising move drawn is IV to V at 2.7 bits, which is 42 per cent of everything its passage spends. The eight weights are ordinal and stipulated rather than counted, so these are the numbers that ordering implies and not a measurement of any repertoire. Harmony and voice leading

Surprise is a number

The chord that did not come was described rather than measured. Its measure is the information content of what did arrive, and a model of the probability has been to hand since the key-finding essays — eight root-motion weights, ordinal and stipulated. Reading them as a distribution prices a deceptive cadence at 2.71 bits against a perfect one's 1.85, and turns up the fact that the largest of the eight had never been read by anything.

 The fingerboard, with the hardest place on it. Every written pitch from G3 to G6 on every string that can reach it, shaded by the width of Schelleng's bow-force window there — dark is narrow, which is a note that is hard to start. Two effects are multiplied and neither earlier figure could show the other: the body's admittance, which depends on the frequency, and the bowing fraction and string impedance, which depend on where the hand is. The worst place is C♯4 on the G3 string, 6 semitones up it, at a window of 12.2 against 471 at the easiest — a factor of 39 across the instrument. It sits where the body's A0 resonance crosses the heaviest string played high, which is the compounding this figure was drawn to find: neither variable alone puts a minimum there. Instruments and their design

The hardest place on the fingerboard

Two things narrow a bow's window and each has been drawn alone. The bridge's admittance is a function of frequency; the bowing fraction is a function of where the left hand is. They meet on a real fingerboard, and multiplying the two curves gives a map with a worst place on it — C♯ on the G string, sixth position, where the body's air resonance crosses the heaviest string played short. The window there is twelve, against three hundred and eighty at the easiest.

 Two fusion cues, and they do not agree about a single spectrum. Each spectrum twice. Hollow is the harmonicity census — the fraction of partials near enough a whole multiple of one fundamental to fuse, which is harmonicity. Filled is the same fraction under common fate: how many partials decay at a rate within a factor of 2 of the strongest partial's. Ranked by harmonicity the order is an ideal string, a piano string, a bell, a kettledrum, a bar; ranked by common fate it is a kettledrum, a bar, an ideal string, a piano string, a bell. The two orderings are nearly reversed. An ideal string is perfect on the first cue and 20 per cent on the second, and a kettledrum — the worst spectrum in this collection for fitting a series — is the only one whose partials all die together. Perception and the listener

The partials that do not die together

The fusion census is a still photograph: it asks whether a set of partials fits one harmonic series and has no term for time. Put time in and the ranking reverses. An ideal string is perfect on harmonicity and holds a fifth of its partials together by decay; a kettledrum — the worst spectrum in this collection for fitting a series — is the only one whose modes all die at one rate.

 The same interval, started on each of the twelve. An interval of 7 semitones started on each pitch class of a major key, against how strongly the key specifies its two notes — the mean of the probe-tone profile at each. The interval account says the listener encodes a distance, so the key cannot enter and the prediction is a horizontal line at 5.4 cents. The degree account says the listener refers each note to the key, so its precision on a note falls as the key's specification of that note weakens; scaled to agree at the most stable start, it rises from 5.4 cents on C to 8.5 on E♭. Every earlier figure measures a quantity the second account says is not being formed at all. Intervals and chords

The quantity a rival account says is not there

Two earlier essays measure how much two notes' errors are correlated through a shared anchor, and price what that correlation would be worth. There is a rival account in which a listener refers each note to a key and never forms the distance at all — under which the correlation is not small, it is a description of something that is not happening. The two accounts agree on almost everything and disagree on one manipulation, and the manipulation costs an afternoon.

 How far the detector looks, note by note. The number of notes that fit inside a 3.5-second present at each point of the tune, once the performance has lengthened its phrase-final notes by 30 per cent. It runs from 5 to 10 notes against a constant 7 for the unperformed version, and it dips exactly where a boundary is, because a boundary is where the performance slows. Reading the boundary-strength curve with that width at every point instead of one width everywhere gives an agreement of 0.55 with the notated phrasing, against 0.36 for the fixed width the present dictates and 0.71 for a fixed width fitted to this tune. The dips are marked, and the notated boundaries are the vertical lines: the detector narrows itself at the places it is supposed to find, which is the circularity this figure has to be honest about — the lengthening was put there by the notation. Form and structure

A detector whose resolution the performance sets

The boundary detector lost its free parameter when the psychological present became a number of notes at a stated tempo, and what that held still was named at the time: a performance slows into a phrase end, so the number of notes inside the present is not the same everywhere in a tune — it falls exactly where a boundary is. Making the width follow the performance recovers half of what removing the parameter cost, and honestly leaves the other half.

 The surprise of each chord, against the uncertainty it arrived into. The information content of each step — minus the log of its probability under a distribution that multiplies the root-motion weight by how well the destination triad's notes fit the key — with the entropy of the moment before it drawn behind. I – IV – V – I: I→IV 1.71 bits, IV→V 2.80 bits, V→I 1.60 bits, against a mean uncertainty of 2.63; I – IV – V – vi: I→IV 1.71 bits, IV→V 2.80 bits, V→vi 2.61 bits, against a mean uncertainty of 2.63. A surprise larger than the entropy it arrived into is an outcome the model was not expecting even given how uncertain it was; one below it is an outcome the model had already mostly bet on. An earlier essay produced the first of those numbers and had no way to produce the second, because a set of preferences is not a distribution and only a distribution has an entropy. Harmony and voice leading

A chord, given a key and a predecessor

A chord's improbability has been priced from its root motion alone, which left one multiplication unmade: a chord is also improbable because its notes do not fit the key, and that number has been available since the probe-tone profile. Multiplied and renormalised, the two give a conditional distribution — and a distribution has an entropy, which is the quantity a surprise has to be read against and which a list of preferences cannot supply.

 A cycle whose position is in the instrumentation. 3 isochronous layers over a cycle of 16 steps, at periods 16, 8, 4. Every layer on its own is perfectly symmetric and tells a listener nothing about where they are; the combination gives 4 distinct signatures over 16 steps, and hearing one of them leaves 2.81 bits unknown. The information is in which instruments sound rather than in where the onsets fall, which is a different answer from the one a single timeline gives — and it needs no asymmetry anywhere. The cost is 7 strokes a cycle, 0.44 to the step, spread over 3 players. Rhythm and metre

A cycle that says where it is

Euclidean timelines were asked how quickly they tell a listener where in the cycle they are, and answered it with rotational asymmetry: a symmetric pattern never locates at all. A colotomic cycle answers the same question with nothing asymmetric in it. Several isochronous layers at nested periods — a gong every sixteen, a kempul every eight, a kenong every four — put the position in which instruments sound, and the position is legible from a single stroke.

 What a competition decides when the two cues do not agree. Each spectrum with its two cue readings and the grouping the competition chooses. Harmonicity asks whether a partial is near enough a whole multiple to belong; common fate asks whether it decays at the same rate as the rest. Where they disagree there is no rule in this collection, so the published apparatus is used instead: every way of splitting the partials into one stream or two is scored for the partials each cue says it has wrongly grouped and wrongly separated, and the cheapest wins. an ideal string — harmonicity 100 per cent, common fate 20, and the competition says one stream; a piano string — harmonicity 90 per cent, common fate 20, and the competition says a cut after partial 2; a bell — harmonicity 88 per cent, common fate 13, and the competition says a cut after partial 1; a bar — harmonicity 33 per cent, common fate 67, and the competition says a cut after partial 4; a kettledrum — harmonicity 60 per cent, common fate 100, and the competition says one stream. The exchange rate between the two cues is the number nobody here can supply, so what is reported beside each is how many decades of it leave the answer unchanged. Perception and the listener

The exchange rate nobody has

There are now two cues that disagree about how a spectrum divides, and every figure so far reports them separately because there is no principled way here to weigh one against the other. The published apparatus is a competition between grouping hypotheses with a cost per cue — and the useful thing it produces is not the winner but how much of the exchange rate the winner survives.

 Expectation as a curve, and what a change costs where it lands. Every quantity so far is attached to a chord change: a list of surprises, one per event. A listener's expectation is continuous — it sharpens through a bar and collapses when the change arrives — and the two ingredients for it are already here, the harmonic rhythm and the metrical beat weights. The curve is the hazard: given that the chord has not changed yet, the chance that it changes on this beat. It runs from 0.043 on the weakest beat to 0.290 on the downbeat, a ratio of 6.7, against 0.125 if every beat were alike. The marked beats are where the changes actually arrive, and their timing bill is 3.6 bits against 6.0 for a listener with no metre — so these changes are 1.7 times cheaper to expect than a metreless listener would find them. That term is new: an earlier essay prices which chord arrived and this prices when, and a listener meets the sum. Harmony and voice leading

Expectation is a curve, not a list

Every quantity so far is attached to a chord change: a list of surprises, one per event. A listener's expectation is continuous, sharpening through a bar and collapsing when the change arrives — and the two ingredients for it were already here, in two other accounts. What comes out is a second surprise, for when a chord arrives rather than for which one it is.

 The same tune read at six widths of the psychological present. A later essay made the detector's smoothing width a function of position, which removed its last free parameter but one — and the one it cannot remove is the width of the psychological present, because that is a fact about listeners rather than a choice. So the honest object is not a reading but a family of them, one per width. A short present finds 6 boundaries and a long one finds 2, and the family agrees on 0 of them. The fixed-width control, at its own best width, scores 0.67 against the adaptive readings' 0.67, 0.75, 0.33, 0.33, 0.40, 0.40 — so the adaptation does not win, which is what that essay reported too. What the family adds is the ordering: a boundary in every row is a different claim from one in a single row, and a single reading has no way to say so. Rhythm and metre

A family of readings

Removing the detector's free parameter, and then its constant tempo, cost persistence both times — the property that made its boundaries ordered rather than merely found. Recovering it means a family of adaptive readings rather than one, indexed by the width of the psychological present, which is the one parameter that cannot be removed, because it is a fact about listeners.

 An ensemble finding an asynchrony nobody told it about. An earlier essay produced a map of required leads — which notes of a scoring have to be played early, and by how much — and nothing tells the players those numbers, because they are a property of the instruments' attacks rather than of the music. So an ensemble has to find them, and the mechanism is already here: each player hears sounds rather than onsets and moves their next onset toward the mean of the others'. The spread of arrival times starts at 20 milliseconds and settles at 4, crossing 5 milliseconds after 5 beats — about 1.3 bars of four. The leads it converges on match that map to within 0.2 milliseconds, which is what makes this a convergence rather than a coincidence: the fixed point of players listening to each other is every player leading by their own attack. Rhythm and metre

How many bars an ensemble needs

The map of required leads is something nobody tells the players, because the leads are a property of the instruments' attacks. So an ensemble has to find them, and the mechanism is the one the microtiming essays describe: each player hears sounds rather than onsets and moves toward the others. It converges on the map to within a fifth of a millisecond, in five beats, and there is a best correction gain.

 How much of the reading comes from what has not happened yet. Every margin reported earlier is two-sided: the best path through a key at a bar is the best score into it plus the best score onward from it, and the second half uses bars a listener has not heard. Dropping that term is one line, because the dynamic program already had both halves separately. The mean margin falls from 3.90 bits with hindsight to 1.79 without it, so 54 per cent of this passage's certainty is retrospective. The two passes never disagree about which key is best here, so the hindsight buys confidence rather than a different answer. This is the quantity every earlier essay has assumed and none has measured. Scales and modes

How much of the reading arrives late

Every margin reported earlier is two-sided: the best path through a key at a bar is the score into it plus the score onward from it, and the second half uses bars a listener has not heard. Dropping that term is one line. On a thirty-two-bar song it removes more than half the certainty, and on a passage built to be ambiguous it changes the key named at nine bars out of eleven.

 The period is still there, and it is wider. The autocorrelation of a 12-partial complex on 220 hertz, drawn twice: steady, and averaged over one cycle of a 71-cent vibrato. A vibrato moves every partial by the same number of cents, so the complex is exactly harmonic at every instant and nothing is mistuned — what moves is the period the extractor is looking for. The peak survives. It loses 6 per cent of its height above the surrounding lags and gains 11 per cent in width, because the vibrato swings the period by 0.37 milliseconds against a peak 0.90 wide. Its maximum also moves, to 2.4 cents sharp of the still tone's, which is a prediction with a sign in it. Instruments and their design

The pitch that does not wobble

Three earlier essays have treated a vibrato as a modulation of roughness. The reason singers use one is what it does to the note, and there is an extractor here that turns a set of partials into a pitch and has never been asked what it does with partials that will not hold still. The period survives, at a cost that rises with the extent — and the practice stops within a hair of where the cost becomes total.

 One contrast survives every tempo anybody plays and the other does not. How much of each quantity's contrast between chords a listener still has at the end of each chord, against how long a chord lasts. The roughness curve is flat at one down to 45 milliseconds a chord and then falls off a cliff, because its window is 37 milliseconds and a boxcar either fits inside a chord or does not. The loudness curve is already losing at a second a chord and keeps 83 per cent at the slowest pace here, 39 at the fastest. Nothing in music is faster than the roughness window and a great deal of music is faster than the loudness one, so a passage delivers its dissonance and averages its dynamics. Form and structure

The dissonance arrives and the dynamic does not

A scoring decides two things at once and both of them have to be integrated by a listener before they exist. The loudness smoother's release is two seconds and the roughness window is thirty-seven milliseconds, and that ratio of fifty decides which of the two survives at the pace music is actually played. Nothing anybody performs is fast enough to blur a dissonance, and a great deal of it is fast enough to average a dynamic.

 The cue that settles it. Every spectrum to hand, arbitrated by the earlier competition and then again with the onset cue added at equal weight. 3 of the 5 change their verdict, and all 3 change the same way — from splitting into two streams to staying as one: a piano string, a bell, a bar. Nothing changes the other way, because the onset cue on a struck source votes for fusion on every partial and can only ever push toward one stream. The bell is the case worth naming: its partials are wildly inharmonic and it is heard as one sound, which is a fact the harmonicity cue alone cannot produce. Perception and the listener

The cue that settles it

Arbitrating between two grouping cues meant sweeping an exchange rate nobody could supply. The cue it had no term for at all is the one every account calls strongest, and its strength is computable: a struck string's partials start together to within a tenth of a millisecond against a threshold of twenty. Put that into the competition and three of five verdicts change, all the same way — and a bell becomes one sound.

 A page has two decibels and a player has sixty. Across, parts added to a final chord one at a time, each at the same level; up, the loudness that results, on a logarithmic scale. Going from one part to eight moves the total by 1.8 decibels and does not move it monotonically — four parts are louder than five and than eight. The faint line is what a naive power sum would give: 9.0 decibels. The band down the right is the same chord played by people, from forty to a hundred decibels, which spans 62. So a texture that thins from eight parts to one is not a diminuendo. It is a change of colour at constant loudness, and everything the closure figures call a dynamic belongs to the performance. Perception and the listener

A page has two decibels

The account of closure called its dynamic component a corpus debt: it had no model of dynamics in a form. Loudness supplies one, and applying it answers the debt by refusing it. Adding parts to a final chord one at a time, each at the same level, moves the loudness by under two decibels and not monotonically — while a player has sixty. A texture that thins is not a diminuendo.

 A listener has a quarter of an analyst's confidence and the same answer. The mean margin between a passage's best two key readings, against how many bars a listener's memory of the evidence takes to halve. The two-sided reading — the one that uses bars that have not happened yet — sits at 3.90; the forward pass with perfect recall at 1.79; a forward pass whose evidence halves every 6.6 bars at 1.00. The dots' size is how often that reading names the same key as the two-sided one: 100 per cent at perfect recall and 81 at a one-bar half-life. So forgetting costs a great deal of confidence and very little accuracy — the key is robust and the certainty is not. Harmony and voice leading

The listener who forgets

Setting an analyst's reading of a key against a listener's measures what arrives late. Both passes assume perfect recall of their own half — which is as wrong going forward as knowing the future is going back. Put a decay on the forward pass and a listener with a memory of a few bars keeps a quarter of the confidence and nine tenths of the answers.

 The number nobody has moves the size and not the order. The mean total surprise per chord change, against how much the two surprises share. At zero they are independent and the total is their sum; at one they are the same event and the total is the larger of the two. The mean falls by a factor of 1.53 across that whole range, which is the size of the thing a corpus would settle. The ordering of the events by total surprise does not move at all until the very end: 5 of the 6 correlations swept give exactly the ordering independence gives, and only perfect dependence changes it, by 3 places out of 7. The most surprising event in the passage is the same one at every correlation. So the corpus three separate accounts have recorded wanting would change what this figure reports and not what it concludes. Harmony and voice leading

Two surprises and one event

A chord change is surprising twice over — in which chord it is, and in when it comes — and a listener meets one event. Adding two surprises needs to know how much they share, which is a fact about a repertoire nobody has. Sweeping it instead: the total moves by half, the ordering does not move at all, and the most surprising moment in a passage is the same one whatever the answer turns out to be.

 The same eight notes are four times as much to read. How many bits each note of a line carries, taken as minus the log of the probability of the interval that reached it, under the distribution of melodic steps measured over the tunes used throughout. A scale costs 1.76 bits a note and a wide leaps costs 7.02 — a factor of 4.0 at the same number of notes on the page. Every quantity computed until now counts notes, and the page cannot tell these apart: eight quavers are eight quavers of horizontal space whichever line they spell. Scales and modes

A reader does not read notes

Eleven earlier essays count notes, and the page cannot tell one line of eight quavers from another. A reader can: a scale of eight is one object where eight leaps are eight. Measured against the melodic interval distribution, the same eight notes are four times as much to read — and the eye–hand span, the best-measured quantity in the reading literature, is four notes of a tune and one of a leaping line.

 A rest is a diminuendo, and a long one. How far a listener's running impression of loudness falls during a silence, converted into the diminuendo that would have taken it the same distance. Half a second of nothing is worth 3.2 decibels, a second and a bit is worth 8.1, and two and a half seconds is worth 17. The marked line is three and a half seconds, which is where a gap starts to be heard as an ending rather than as a pause: at that length the reference has fallen by 24 decibels, which is more than a fortissimo to a pianissimo. A tempo swept over a factor of ten, a deceleration over a factor of three and a gesture length over a factor of forty-eight all returned the same reading to four significant figures. This one moves it by twenty-four decibels. Form and structure

A rest is a diminuendo

Three parameters swept over factors of ten, three and forty-eight returned the same reading to four significant figures. The one manipulation left unswept moves it by twenty-four decibels: a silence. A listener's running impression decays at the loudness smoother's two-second release, so a general pause is a diminuendo nobody wrote, and at the length that makes a gap an ending it is worth more than any marking a composer has.

 A clarinet's partials, each on its own resonance. The 8 partials of the clarinet's chalumeau D that ride an impedance peak, each building toward its steady amplitude as 1 − exp(−t/τ) with τ = Q/πf from that peak's own Q. The time constants run from 11.9 milliseconds to 64.5, so the partials do not arrive at different times — they all begin the instant the reed does — and what differs is how fast each approaches its final level. The horizontal bars are how far apart the first and last are at three criteria: 5.5 ms at 10 per cent, 36.4 ms at 50 per cent, 121.0 ms at 90 per cent. A twenty-millisecond asynchrony is the threshold for hearing a partial out of a note, and this note crosses it at 32 per cent of steady amplitude — so whether a blown note's onset cue is unanimous or divided is decided entirely by how far along a partial has to be before it counts as having started. Perception and the listener

A blown note does not start late, it starts slowly

Computing the onset cue removed a free parameter and turned out to be unanimous, and it predicted that a wind instrument would put it back, because a blown note's partials arrive over tens of milliseconds. They do — 121 on a clarinet — and it is not an asynchrony: every partial begins the instant the reed does and they differ in rate, not in time. Read at a tenth of the steady amplitude the spread is 5.5 milliseconds against a threshold of twenty, so the cue is still unanimous, and the missing number is no longer the exchange rate but the criterion.

 Eleven partials is one partial too many. What fraction of a spectrum the harmonicity census finds fused, against how many partials it is asked to census. At ten a perfect harmonic series fuses 10 of 10 and the fundamental it finds is the right one. At eleven it fuses 5 of 11 and the fundamental jumps to exactly 2.00 — the octave above. The cause is the cap the census carries for a reason established earlier: without it a bell fuses perfectly at a fundamental nobody could hear, so the search refuses any fundamental more than about ten harmonics below the top partial. At eleven partials the first thing that cap excludes is the series' own fundamental, and the census then takes the octave and calls every odd partial inharmonic. So the number of partials and the cap are the same number, and nothing had ever said so, because every earlier figure censuses ten. Perception and the listener

Eleven partials is one too many

Six earlier essays census exactly ten partials and no figure has ever passed another number. At eleven, the harmonicity census stops finding a perfect harmonic series' own fundamental, takes the octave above it, calls every odd partial inharmonic, and the competition cuts an ideal string in two. It is not the arbitration — the cost of a second stream was swept over a factor of fifty and every verdict came back identical — it is a cap that exists for a good reason and turns out to be the same number as the count.

 Where a cycle of 16 at 16, 8, 4 outruns the listener's memory. The residual uncertainty a listener is left with once the evidence has stopped accumulating, against how long one turn of a 16-step cycle takes. The listener's memory of a step halves after 3.5 seconds throughout; what changes is how many steps that is. At a cycle of 1.6 seconds it is 35 steps and every design reaches certainty, which is the regime a clave is played in. At 60 seconds it is 0.93 steps and none of them does: layers at 16, 8, 4 settles at 1.89 bits, son clave settles at 2.27 bits, the bossa-nova pattern settles at 2.38 bits, the best single line of 7 settles at 1.85 bits. That is the range a gong cycle occupies, and it is the design that wins there. Rhythm and metre

The cycle that outruns the memory

A timeline and a colotomy were compared at equal strokes and the comparison had no clock in it. A memory span is a number of seconds and a cycle is a number of steps, so the two only meet through a tempo — and at a clave's two seconds a listener's memory covers twenty-eight steps and forgets nothing, while at a gong cycle's forty it covers 1.4 and forgets almost everything. The single line is the better locator up to twenty-three seconds a cycle and the layered code is better after it, which is very close to where each is actually used.

 An entering part is worth 0.9 phons, in the middle of its range. A texture of 5 parts at 62 decibels each, with one more part added at the same level, tried at every semitone from C2 to C7. The vertical axis is what the addition is worth in phons, and a phon is a decibel here; the shaded strip is the difference limen for loudness, so an entry inside it is not heard as a change of level. The median entry is 0.89 phons and only 29 of 61 clear the limen — the lowest of them at A♭4, 415 hertz. The best available, at B♭6, is worth 4.2. The two lines are the two loudness models to hand: they agree everywhere above the tenor register and part company below it, where the greedy critical-band grouping reports 24 entries that make the texture quieter and the excitation pattern reports none. Perception and the listener

A part entering is not a change of level

Five earlier essays measure a sonority that is already sounding. The one thing an orchestrator actually controls is the entry, and priced at every semitone it comes to 0.89 phons — under the difference limen — with only 29 of 61 available entries clearing it and none below A♭4. What an entry is instead is an object arriving, and the reason is that its own rise time is faster than the listener's.

 Which bars the key is decided by, and which bars it is believed on. Every bar of a 32-bar scheme removed in turn, with what its absence costs. The column is how far the passage's mean margin falls without that bar — how much of the model's certainty it supplies. The dot is how many bars are then read as a different key — how much of the answer it supplies. The 18 bars an analysis would point at — a section opening or closing, a dominant, the chord a dominant resolves to — average 0.079 bits of certainty and 0.50 bars moved; the 14 ordinary bars average 0.047 and 0.07. So the structural bars carry 1.7 times as much certainty as the ordinary ones, and 7.0 times of the answer. Those are different quantities, and the second is the one an analysis is about: an ordinary bar can carry a great deal of a passage's certainty and none of its reading. Harmony and voice leading

The bars a key is made of

A discount that treats every bar alike is the wrong shape for a memory, so take each bar away in turn and see what it was worth. The bars an analysis points at carry seven times as much of the answer as the ordinary ones and slightly less of the certainty — and a memory built only of them reads the passage worse than a memory with no structure in it at all.

 A third of the timing surprise is paid for by nothing happening. Every beat of a bar of 4/4 at 1 chord change a bar, with what it costs a listener in expectation. The lower block is the arrival — the chance the chord changes here times what that change costs to be surprised by. The upper block is the hold, which is paid when the chord does not change and is charged at every beat rather than at every event. Over the bar the arrivals come to 2.61 bits and the non-arrivals to 1.29, so the term never spent before is 33 per cent of the total rather than most of it. The two together are exactly the binary entropy of each beat's own hazard, which is why the bar's whole timing bill is 3.90 bits and cannot be raised by rearranging where the changes fall. Harmony and voice leading

The surprise of nothing happening

A hazard charges a listener twice — once when the chord changes and once, quietly, at every beat it does not. The second term was expected to dominate, because there are more beats than changes. It is a third of the bill at one chord a bar and never reaches a half at any rate a metre survives, because the cost of a beat where nothing happens is second-order small.

 What a chord change costs where it actually lands. Every beat of a bar of 4/4 at 1 change a bar, priced as a listener meets it: the pale block is what has already been paid waiting through the beats the chord did not come on, and the dark block is the arrival itself. Their sum is what it costs to be surprised by a change here. The last column is the remaining case, never priced before: no change in the bar at all, at 1.61 bits and a probability of 0.33. The 9 costs are a proper distribution — 1.000 — which is the check that this is one model rather than two. And the spread is the finding: the arrival term alone puts a factor of 6.7 between the best and worst beat, and counting the waiting makes it 19. Harmony and voice leading

Where the chord actually lands

Every timing surprise so far is evaluated on a downbeat: the curve spans a factor of 6.7 and every number is read at its peak. Charge the waiting as well as the arrival and the costs over a bar become a proper distribution, the spread between the best and worst beat rises to a factor of 19.5, and a third of the probability sits on a bar in which nothing changes at all.

 How much of A4's pitch error a key could possibly remove. The largest correlation two notes' pitch errors can have at A4, against how long each note lasts. A note's error has two parts and only one of them is the listener's: the steady-tone limen of 4.04 cents, which context might reduce, and the bound a note of length T puts on its own frequency, which context cannot touch. Taking them in quadrature, the shareable fraction is the curve. At a quarter-second note it is 0.21, so the one-half priced earlier is not available at all until each note lasts 486 milliseconds — which is exactly the crossover found by a different route, because a correlation of a half is the two parts being equal. The step is the convention used here, which takes the larger of the two rather than their sum and therefore says the shareable fraction below the crossover is zero. Intervals and chords

The part of the error a key cannot touch

Four earlier essays turn one dial — the correlation between two notes' pitch errors — and apply it to the whole of a note's limen. Half of that limen is not the listener's: a note of finite length does not carry its frequency more finely than 1/2T, and no context can put information into a signal that is not there. So the correlation has a ceiling, it is 0.21 at a quarter-second note at A4 and 0.07 at A2, and the figure that prices a correlation of one half is drawn where one half is unavailable.

 A struck note's two ends are the same for every loss law. The partial levels of a string spectrum struck at 80 decibels on 130.8 hertz, and what is left of it when the fundamental itself falls under the threshold of hearing, for three laws relating a partial's decay rate to its number. The left panel is every one of them: a loss law cannot change the spectrum at the instant of the strike, because no time has passed. The other three are every one of them too: whatever the law, the note ends with nothing above the threshold. So both ends of the slide are shared, and everything that distinguishes an exponent of 0.5 from an exponent of 1 from an exponent of 2 is in the middle. Timbre and acoustics

The middle nobody could have guessed

A struck note has no steady state, only a slide from one spectrum to another — so the question is what the middle carries that the ends do not. The answer is exact rather than statistical: every loss law in the family leaves the strike with the same spectrum and ends in the same silence, so both endpoints carry precisely nothing about which of them it is. The whole difference is 41.3 decibels, and it peaks 0.38 seconds in, seven per cent of the way through the note.

 Where a section's fluctuation stops being a beat, on 220 hertz. Two rates up the spectrum of a 220 hertz note sung by a section whose voices are spread by 15 cents. The rising line is the beat rate between a typical pair of them, which grows with the partial because a mistuning in cents is a difference in hertz that scales with frequency; it reaches the 15 hertz at which a beat stops being a beat by partial 5.5, at 1217 hertz. The flat line is the amplitude modulation the vibrato imposes through the formants, which is 6.0 hertz at every partial because the vibrato modulates every partial by the same number of cents at the same rate. The two are equal at 487 hertz. Above 1217 hertz the beating has become roughness and the only fluctuation left is the vibrato's — and that frequency is the same one an octave up, where it is partial 2.8 instead. Instruments and their design

The rate that does not rise with the partial

Twelve earlier essays give every vibrato the same six hertz, and the measured spread is 5.5 to 7.5. Putting the two fluctuations a choir contains on one axis shows why the rate matters: the beating between mistuned voices rises with the partial and leaves the range a listener follows as fluctuation at 1,217 hertz, while the vibrato's own modulation is six hertz at every partial. Above that frequency a section fluctuates by vibrato alone — and if every singer had the same rate, it would barely fluctuate at all.

 A notehead in four parts costs 1.30 bits and one in two parts costs 1.89. What one notehead asks of a reader, against how many parts are on the page, for a progression realised by the voice-leading solver used here at 2 semitones of motion a voice a chord. The horizontal rule is a note of a single melody under the measure established earlier, 1.89 bits, which is what a texture costs when its parts have to be read one at a time. The bars are the harmonic reading: the chord, charged at the worst case of 2.81 bits for one of seven diatonic degrees, plus the logarithm of how many voicings of it the previous chord could legally have moved to. At two parts there is no bar, because a duet has no complete voicing of any triad — it cannot state the harmony and has to be read as 3.79 bits of two independent lines. Every thicker texture is cheaper a notehead than the thin one, and the four-part figure is an upper bound. Harmony and voice leading

Four parts are easier to read than two

Twelve earlier essays read one line, and a score is several at once. Measured through the voice-leading model, a notehead of a four-part chorale asks a reader for 1.30 bits and a note of an independent line asks 1.89 — so twice the ink is less than three quarters of the load. The reason is a boundary those essays already established: a duet has no complete voicing of any triad at all, so two parts cannot be read from their harmony and have to be read as two melodies.

 A head model a few millimetres out reads every azimuth but the front. The azimuth a listener reports against the azimuth a source is at, for internal head radii from 8.22 to 9.28 centimetres against a true radius of 8.75. The delay a source produces is (r/c)(θ + sin θ) and the listener inverts it with the radius they believe they have, so their answer solves θ̂ + sin θ̂ = (r/r̂)(θ + sin θ). Every curve passes exactly through the origin: on the median plane there is no delay and therefore no error, whatever the head model is. The error grows with azimuth and is largest at the side. An internal head 5.3 millimetres too small runs out of azimuth at 82 degrees: beyond that the world is delivering a delay larger than any its owner's model can produce, and every source out there collapses onto the side. Perception and the listener

Where a wrong head gives itself away

Every claim so far maps a delay to a direction through one fixed geometry, and the listener acquires that map while the geometry grows under them by seventy per cent. So the map can be wrong — and the essay before this one said the error would be largest on the median plane, where the delay curve is steepest. It is exactly zero there. The steepness is in the error and in the threshold and cancels between them, which leaves a listener whose internal head is 1.3 millimetres out with one place to catch it: hard to the side, where nobody localises well.

 Thirty cents out of tune is heard as 8 on a 125 ms note and 25 on a 1 s one. How far out of tune a note sounds against how far out of tune it is, at 4 note lengths, at 440 hertz. The key is treated as a prior over pitch: a mixture of Gaussians on the twelve scale degrees, weighted by Krumhansl and Kessler's probe-tone profile and given the width the degree account already uses. The likelihood is the note's own effective limen, which for a short note is the Fourier bound 1/2T. The estimate is the posterior mean, and the shrinkage toward a prior is one line of arithmetic. At 1 s a thirty-cent mistuning is heard as 25.0 cents and at 125 ms as 7.8. Every curve turns back up near the middle of the semitone, because past there the nearest degree is the other one and the pull reverses. The buttons sound at A4, which is the pitch the figure is computed at, at the shortest note length it draws. Intervals and chords

A short note is heard more in tune than it is

Four earlier essays treat a key as something that reduces the noise in a pitch judgement. Treat it instead as a prior and the prediction changes kind: not a smaller error but a systematic bias, pulling a short note toward the nearest scale degree by an amount the Fourier bound sets. Thirty cents out of tune on an eighth-of-a-second note is heard as eight. And the part the debt got wrong is the part that matters — the bias does not vanish on a long note. It stops at 17 per cent at A4 and at 48 per cent at A2, because the likelihood's width has a floor that no duration removes.

 The mixture at which the chain acquires a tonic. Eight turns of i–iv–v–i in A minor, natural throughout, read at every mixture of the two emissions: nought is the original set overlap against the triad on each degree, one is the measured probe-tone profile rotated to each candidate key. The blocks along the top are the name the model gives, the line below is how far ahead of its best rival that name is. Below a mixture of 0.55 every bar is called C major, which is the collection and not the key; at and above it every bar is called A minor. The margin collapses to 0.33 bits at the crossing and recovers to 4.56 — higher than the 3.38 it started at, because a profile has an opinion about this passage and an overlap does not. Harmony and voice leading

A tonic bought with the function

Putting the measured probe-tone profile inside the ordered key-finder is one term, and it does what was predicted: the natural-minor passage is named A minor at every bar instead of C major. It also does two things nobody predicted. It renames a scheme that has been read in the wrong key at every bar since the day it was written, and it destroys the chord's function while it is buying the key.

 The reading the joint search was never offered. The best chord at each mixture of the two segmentation cues, and what the same weighting gives the same notes shuffled into a different order. Both fall along the axis, and most of the fall is the ruler rather than the music: a metrical weighting over a bar of eight spans a factor of eight and a three-to-one duration spans three, so the weighted note mass is 2.1 times more concentrated at the left of the figure than at the right, and a concentrated mass is easier for four notes to cover. What is not the ruler is the gap. It is widest at a mixture of 0.75, where the reading is D minor7 at 2.15 standard deviations above its own null, against 1.12 for C major7 at a mixture of nought. The joint search holds this axis at nought, so D minor7 is not among the hypotheses it considers. Harmony and voice leading

A fourth decision, and two that were never made

The joint search resolves key, metre and segmentation together and holds the segmentation's cue mixture at zero. Adding the mixture is one loop, and reading the search in order to add it turns up something worse than a missing axis: on the passages it is drawn on, the key it reads is the same key at all forty-eight of its hypotheses and the metre scores every barline identically. The fourth axis then cannot be ranked at all until each reading is measured against its own null, because a mixture changes the ruler and not only the answer.

 Eighty-one chords the expectation model cannot tell apart. Every voicing of a dominant seventh on G inside the three octaves above its own root, placed by how rough it is and how far its outer voices are apart, and coloured by which member of the chord is at the bottom. The roughness runs from 0.269 to 1.449, a factor of 5.4, computed from each voicing's own spectrum under Plomp and Levelt's roughness model. The identity surprise the expectation model assigns is 3.51 bits for every one of the 81, because it is a function of a scale degree and its predecessor and there is no register anywhere in it. What separates them is spacing rather than inversion: roughness falls as the outer voices spread apart, correlating -0.42 with the span, and is indifferent to which member of the chord is at the bottom at 0.02. The seventh in the bass is not what makes a chord rough; a fourth and a third packed together at the bottom of the range is. Harmony and voice leading

Eighty-one chords, one number

A dominant seventh has eighty-one arrangements inside three octaves and their roughness spans a factor of five and a half. The tonal-expectation model gives every one of them the same 3.51 bits, because its states are scale degrees and there is no register anywhere in them. Conditioning the surprise on the voicing costs no corpus — and the arithmetic says the conditioning belongs beside the probability rather than inside it, for three reasons that can each be computed.

 Where a bar of 25 and a bar of 9 first disagree. The additive metre 2+2+2+3+2+2+2+3+2+2+3 — 25 units, an onset at the head of every group — with the accents it predicts drawn above the accents predicted by reading it as a repeating bar of 9, which is the cut of it that agrees longest. The two rows are identical for 24 consecutive steps and differ for the first time at step 25, where the shorter reading expects an accent and the metre does not supply one. Nothing before that step distinguishes the two hypotheses, so a listener who has not heard 25 consecutive steps has no evidence either way — whatever they are disposed to hear. Rhythm and metre

A twenty-five is a nine until its last unit

Every account of long additive metres says they are heard as groups of shorter ones, and the metre-induction model had never been pointed at the claim. Pointed at it, the model does not prefer the group — it prefers the long bar outright, and would go on preferring it more the longer anybody listened. What it cannot do is start: the evidence that separates a bar of twenty-five from a bar of nine does not exist until the whole bar has been heard, and at the tempo an unequal metre is best played at the psychological present holds sixteen units.

 A bar of silence is worth 13.1 decibels written last and 6.0 written first. Four closing gestures, drawn against how many seconds of silence each contains. All four hold the same 12-part texture, write the same 6.0-decibel diminuendo over 4 seconds, and differ only in what order the diminuendo and the silence are written in. The quantity is how far the listener's running impression has fallen when the final chord arrives, in decibels of equivalent diminuendo. Written with the silence last, a general pause of a bar is worth 13.06 decibels and one of three and a half seconds is worth 28.3. Written with the silence first, both are worth 6.00 — exactly the diminuendo's own depth, because the music resuming after the silence puts the reference back at its own level. The silence written first is worth less than the silence written with no diminuendo at all, which reads 8.12 at a bar: a diminuendo placed after a general pause takes 2.12 decibels away and adds nothing. Form and structure

A general pause is spent by the note after it

Whether a composer should write the pause before the diminuendo or after it looks like a question about how big the ensemble is. It is not. Forty decibels of ensemble are worth one decibel of silence, and the order is worth seven — because a running impression rises twenty times faster than it falls, so half a general pause is spent by ninety-seven milliseconds of sound.

 A note on the downbeat costs 1.89 bits and one on the offbeat 5.70. What each position in a bar of 4/4 asks of a reader, by two routes. The solid bar counts where the notes of this collection's own three tunes actually fall — 28, 2, 26, 4, 28, 0, 16, 0 notes at the 8 positions — and takes minus the log of the frequency. The rule across each bar is the same quantity from the stated metrical weights, 1, 0.15, 0.5, 0.15, 0.85, 0.15, 0.5, 0.15, normalised and logged the same way. Nothing makes the two agree. They put the eight positions in the same order, and they price the tunes' own rhythm a fifth of a bit apart — while differing by more than a whole bit about the quaver after the downbeat, which two notes in a hundred and four ever use. The two positions these tunes never touch at all are drawn at the floor, which is the same floor the melodic measure gives an interval nobody plays. A weight was always a probability waiting to be read as one. Rhythm and metre

Where the note is costs more than which note it is

Thirteen earlier essays measure a page, and the three that price a reader price only its pitches — every line they measure is a run of equal notes. A metrical weight normalised by its own sum is a probability, and minus its logarithm is bits — the same substitution made earlier for intervals. Measured over the tunes used throughout it comes out at 2.23 bits a note against the pitches' 1.89, so the larger half of a reader's load is where the note is.

 The same interval, mistuned by the same amount, at each of its two ends. A C to G in the major key, played 25 cents wrong, with the departure carried by the lower note, split between the two, and carried by the upper note. All three are the same interval size; what differs is which note is off the scale. The share of the departure that survives into what a listener hears is 32 per cent when the lower note carries it and 56 when the upper does. The middle bar is the mean of the other two to within a hundredth, so the averaging is linear and the asymmetry is the whole of the effect. Two things produce it: the prior is 9.0 cents wide at the C and 10.0 at the G, and the likelihood is 13.2 cents wide at the lower pitch and 8.8 at the higher. Intervals and chords

An interval is two posteriors subtracted

Treating a key as a prior over one note predicts that an interval's pull is not the single-note pull doubled, because the two degrees are not equally weighted. Half of that is wrong: splitting a mistuning between the two notes gives exactly the mean of what each end gives alone, to a thousandth, at every one of the twenty-one intervals in the scale. What is not the mean is which end carries it — and the pull turns out to be largest not on the shortest notes but on notes of about an eighth of a second, where the likelihood is a quarter of a semitone wide.

 Which intervals in a key can be mistuned invisibly, and from which end. Every interval between two degrees of the major scale, ranked by how differently its two ends treat a 25-cent departure on notes of 0.25 seconds. The C to B is the most lopsided, at 47 points: a mistuning on its upper note reaches the listener nearly 2.5 times as strongly as the same mistuning on its lower one. The D to E is the most even, at -0. A negative bar is an interval whose LOWER note is the one that carries a mistuning into the listener, which happens whenever the lower degree is the less specified of the two. No account of interval perception predicts a table like this, because an interval is usually treated as one quantity rather than as a difference of two estimates. Intervals and chords

Which end the mistuning is on

Twenty-one intervals in the major scale, each with two ends, and the same twenty-five cents reaches a listener at anywhere between 32 and 79 per cent of its size depending on which of the two notes carries it. The most lopsided is the tonic to the leading note, where a departure on the upper note arrives two and a half times as strongly as the same departure on the lower. Two mechanisms produce it and they can be separated by one flag: two thirds of the asymmetry is register and one third is the key.

 Every product of a just interval is a harmonic of the note it implies. Each interval drawn as two harmonics of a fundamental it does not contain — the lower note is harmonic q and the upper harmonic p — with its three combination tones placed on the same numbering: the difference tone at p − q, the cubic product below the pair at 2q − p, and the one above at 2p − q. minor second 16:15: 1, 14, 17; major second 9:8: 1, 7, 10; minor third 6:5: 1, 4, 7; major third 5:4: 1, 3, 6; fourth 4:3: 1, 2, 5; fifth 3:2: 1, 1, 4; minor sixth 8:5: 3, 2, 11; major sixth 5:3: 2, 1, 7; minor seventh 9:5: 4, 1, 13; major seventh 15:8: 7, 1, 22. The shaded column is the fundamental itself. The difference tone sits on it for every interval up to the fifth, the cubic product for the fifth and every interval above except the minor sixth, whose products are its fundamental's octave and twelfth. Intervals and chords

The tone on the root changes hands at the fifth

Every combination tone of a just interval is a harmonic of a fundamental neither note contains, and which harmonic is fixed by the ratio. The difference tone lands on that fundamental for every interval up to the fifth; the cubic product lands on it for the fifth and every interval above except the minor sixth. So the loud product names the root of a narrow interval and the quiet one names the root of a wide one — and a just major seventh's difference tone is a note seven harmonics up that no keyboard has.

 A major triad's combination tones, against its own notes. The three notes of a major triad on C4 in root position, voiced C4–E4–G4, as tall lines, and every combination tone its pairs make, as short ones: difference tones lowest, cubic products taller. In just intonation 2 cubic products land exactly on a note of the chord, and none comes within forty hertz of one. In equal temperament no cubic product lands on a note of the chord, and the nearest miss is 5.63 hertz. Intervals and chords

A major triad's combination tones are its own notes

Play a just major triad of pure tones and two of the ear's cubic products land exactly on its root and its fifth. The reason is a condition rather than a coincidence — a chord's cubic products fall on its own notes when its middle note is the mean of the outer two in hertz — and it holds for the major triad in root position and in the six-four, and for no minor triad in any position or tuning. Equal temperament misses the landing by one number, 5.6 hertz on middle C, which is a beat that belongs to no pair of notes in the chord.

 A scale in parallel thirds has a line underneath it that nobody plays. A major scale on C4 harmonised in parallel diatonic thirds, with the difference tone f₂ − f₁ of each pair drawn as a third line. In five-limit just intonation that line is C2, A1, C2, F2, G2, F2, G2, C3. In equal temperament it moves to C♯2, A1, B1, F♯2, A♭2, E2, F♯2, C♯3, departing from the just line by +67, +33, −82, +69, +65, −80, −84, +67 cents. Intervals and chords

The bass line under a passage in thirds

A major scale harmonised in parallel thirds gives the ear a difference tone under every pair, and in five-limit just intonation those tones are a diatonic bass line — C, A, C, F, G, F, G, C — made of the scale's own notes. Tempered, the same line moves only by whole tones, a neutral third and a fourth stretched to 650 cents, and wobbles by up to 84 cents from note to note. In sixths the bass is drawn by the other product, because the cubic product of a pair is the difference tone of the same pair inverted.

 Every question is answered in a corner, not on a ridge. The degree share times the minor share, over the plane of two cues: how much of the emission is the probe-tone profile, across, and how much a bass note is worth, down, with every chord's root in the bass and a bass rule that rewards the triad rooted on the bass. It runs from 0% at a profile share of 0 and a bass weight of 0 to 87% at 0.6 and 1.5. Bass 0: 0%, 0%, 0%, 0%, 41%, 32%, 46%. Bass 0.25: 0%, 0%, 0%, 0%, 52%, 55%, 44%. Bass 0.5: 0%, 0%, 0%, 0%, 68%, 62%, 51%. Bass 1: 0%, 0%, 0%, 0%, 78%, 77%, 77%. Bass 1.5: 0%, 0%, 0%, 0%, 87%, 80%, 86%. Bass 2: 0%, 0%, 0%, 0%, 87%, 87%, 87%. Scales and modes

Two cues meet in a corner

The profile finds a key's tonic and the bass finds a chord's degree, and until now each was swept with the other held at nothing. Swept together across 42 settings, the plane they make is not the ridge that was predicted. The tonic is a step in one direction, at a profile share of 0.55, and the bass cannot move it; the degree is a slope in the other, rising to 89 per cent as the bass is weighted, and the profile barely touches it. Every question is answered only in a corner of the plane — and the one place the two cues overlap is the one piece of music both can rescue.

 The ninety per cent was a ceiling. The share of 188 scheme bars read right on both key and degree with a bass note worth 1.5, for three bass lines — every chord's root, a line moving to the nearest chord tone, every chord's fifth — and two ways of using the bass: rewarding the triad rooted on it, or any triad containing it. Wide bars are with no profile in the emission, narrow bars with the profile alone. root line, root rule: 89% and 86%; root line, member rule: 60% and 29%; smooth line, root rule: 68% and 46%; smooth line, member rule: 61% and 47%; sixfour line, root rule: 3% and 17%; sixfour line, member rule: 52% and 30%. The reading with no bass at all is 47%. Scales and modes

A bass line is not a list of roots

Every bass note the key-finder has been given was its chord's root, and under that line a bass cue reads 89 per cent of scheme bars on the right degree. Give the same chords an economical bass that moves to the nearest chord tone, as a keyboard reduction would, and nearly half of them are inverted. The cue that rewards the triad rooted on the bass then reads 68 per cent at best and worse as it is trusted more; the cue that rewards any triad containing the bass cannot be fooled and stops at 61. The same inverted line does one thing the roots never did: it puts the leading note of each new key at the bottom, and finds the rondo's modulations.

 How often the metre, the chords and their product find the barline, chords at 1. Constructed passages of four bars of eight quavers, 100 at each setting, with the barline at the first slot. Rhythm regularity is how much likelier a note is on a strong slot than a weak one; chord regularity is how much likelier a note is to be a tone of its bar's chord than a random scale tone. rhythm 0: metre finds it 10%, chords find it 41%, product finds it 16%; rhythm 0.25: metre finds it 34%, chords find it 35%, product finds it 56%; rhythm 0.5: metre finds it 49%, chords find it 21%, product finds it 66%; rhythm 0.75: metre finds it 50%, chords find it 17%, product finds it 56%; rhythm 1: metre finds it 50%, chords find it 11%, product finds it 45%. Harmony and voice leading

The chords never move the barline

Every hypothesis the joint search had drawn was one bar long, and on one bar with a note in every slot the metre cannot choose a barline at all. Four bars with rests in them make the barline a decision the metre and the chords both have an opinion about, and the prediction was that the chords would move the barline more often than the barline moves the chords. It is the other way round, completely: whenever the two prefer different barlines the search takes the metre's, on up to 72 per cent of passages, and in fifteen hundred passages the chords never once move it. What the chords decide is the one thing the metre cannot see — whether the bar starts on the downbeat or half a bar later — and they decide it right a little over two times in three at best.

 The product and the sum of standard scores, finding the barline, chords at 1. Constructed passages of four bars of eight quavers, 100 at each setting, with the barline at the first slot. Rhythm regularity is how much likelier a note is on a strong slot than a weak one; chord regularity is how much likelier a note is to be a tone of its bar's chord than a random scale tone. rhythm 0: product finds it 16%, sum of z finds it 22%, metre finds it 10%, chords, z 35%; rhythm 0.25: product finds it 56%, sum of z finds it 58%, metre finds it 34%, chords, z 38%; rhythm 0.5: product finds it 66%, sum of z finds it 61%, metre finds it 49%, chords, z 19%; rhythm 0.75: product finds it 56%, sum of z finds it 58%, metre finds it 50%, chords, z 14%; rhythm 1: product finds it 45%, sum of z finds it 48%, metre finds it 50%, chords, z 12%. Harmony and voice leading

The chords are a weak witness to the barline

Scaled by its own range, the metre overrules the chords every time the two disagree about where a bar begins. The obvious repair is to score each reading against its own chance — the metre against the same number of notes placed at random, the chords against the passage's notes shuffled across its bars — and add the standard scores. It changes very little: the search finds the barline within six points of where the product found it, and the chords gain the power to move the barline only on passages whose rhythm says nothing, where random notes move it nearly as often. The null's real result is the size of the two witnesses. At the written barline the metre stands up to 5.9 standard deviations above chance, and the chords, with every note a tone of its bar's chord, stand 1.55 above it at best.

 An accent moves the longest separable bar from 16 units to 18, and no further. For every bar length from 9 to 25 units, over all 1820 arrangements of twos and threes that are not a repeat of a shorter bar, the fewest and the most consecutive steps before the whole bar beats every shorter cut of it. 9: onsets alone 9 to 14, long beat predicted 7 to 12, downbeat predicted 7 to 8; 10: onsets alone 10 to 13, long beat predicted 8 to 11, downbeat predicted 8 to 9; 11: onsets alone 11 to 18, long beat predicted 9 to 16, downbeat predicted 9 to 10; 12: onsets alone 12 to 17, long beat predicted 10 to 15, downbeat predicted 10 to 11; 13: onsets alone 13 to 22, long beat predicted 11 to 20, downbeat predicted 11 to 12; 14: onsets alone 14 to 23, long beat predicted 12 to 21, downbeat predicted 12 to 13; 15: onsets alone 15 to 26, long beat predicted 13 to 24, downbeat predicted 13 to 14; 16: onsets alone 16 to 25, long beat predicted 14 to 23, downbeat predicted 14 to 15; 17: onsets alone 17 to 30, long beat predicted 15 to 28, downbeat predicted 15 to 16; 18: onsets alone 18 to 29, long beat predicted 16 to 27, downbeat predicted 16 to 17; 19: onsets alone 19 to 34, long beat predicted 17 to 32, downbeat predicted 17 to 18; 20: onsets alone 20 to 35, long beat predicted 18 to 33, downbeat predicted 18 to 19; 21: onsets alone 21 to 38, long beat predicted 19 to 36, downbeat predicted 19 to 20; 22: onsets alone 22 to 37, long beat predicted 20 to 35, downbeat predicted 20 to 21; 23: onsets alone 23 to 42, long beat predicted 21 to 40, downbeat predicted 21 to 22; 24: onsets alone 24 to 41, long beat predicted 22 to 39, downbeat predicted 22 to 23; 25: onsets alone 25 to 46, long beat predicted 23 to 44, downbeat predicted 23 to 24. A present of 3.5 seconds holds 16.0 steps at 218 milliseconds a step, so the longest bar some arrangement of which separates inside it is 16 units on onsets alone, 18 with the long beat predicted and 18 with the downbeat predicted. Rhythm and metre

The accent buys two units, however loud it is

A twenty-five cannot be told from a group of shorter bars on its onsets until more steps have gone by than a listener's present holds, and the obvious objection is that nobody plays an aksak bar as bare onsets: the long beat is louder, and the bar's first beat is marked. So how loud does an accent have to be? The question has a surprising answer. Loudness is not the variable. The existing accent cue changes nothing, and delays the answer where it changes anything. An accent that a reading has to predict works at any strength at all, and at no strength does more than a fixed amount: on the long beat it buys the two steps of a short beat, and on the downbeat it takes every arrangement to one floor — the bar less its last beat — which no cue carried by the notes can break. The longest bar that can be heard as one moves from sixteen units to eighteen.

 Six named proportions, as blurred as the durations that make them. Six proportions between two parts of a piece — 1 : 1, 4 : 3, 3 : 2, golden section, 2 : 1, 3 : 1 — placed on one axis by the logarithm of the ratio of the longer part to the shorter, and drawn as bars one criterion wide (d′ = 1) for a listener timing both parts with a Weber fraction of 7%, 15%, 35%. Bars that overlap are proportions that listener cannot tell apart. At 7%, 4 of 5 neighbouring pairs stay apart; at 15%, 3 of 5 neighbouring pairs stay apart; at 35%, 0 of 5 neighbouring pairs stay apart. Form and structure

A proportion is only as fine as its two durations

Analyses of form measure proportions in bars and report them to three figures — a climax at 0.618, a section in the ratio 3 : 2. A listener has each part only as an estimate of how long it lasted, and a ratio of two estimates is blurred by both. Timed as well as anyone times a single second, eleven proportions fit between 1 : 1 and 3 : 1; timed from memory over minutes, two do. The golden section is told from 3 : 2 only below a Weber fraction of 5.4 per cent.

 The same forms by the clock and by what is stored. Six forms, each drawn twice: its sections sized by their share of the bars, and sized by their share of what a listener has to store when a bar counts only if it is recognised from 1 bar of context. Returns are drawn pale with a dashed edge. twelve-bar blues: returns take 67 per cent of the clock and 13 per cent of the storage; thirty-two-bar AABA: returns take 25 per cent of the clock and 5 per cent of the storage; rondo, ABACA: returns take 40 per cent of the clock and 10 per cent of the storage; verse and chorus: returns take 50 per cent of the clock and 13 per cent of the storage; two eight-bar phrases: returns take 0 per cent of the clock and 0 per cent of the storage; a four-bar ostinato: returns take 88 per cent of the clock and 0 per cent of the storage. Form and structure

A return is shorter than its first hearing

A rondo's refrain takes three fifths of the clock and a verse-and-chorus song is balanced to the bar. Count instead the bars a listener could not have predicted when they arrived, and the returns shrink to between a tenth and a quarter of what is kept — so a song equal by the clock is between three and seven times heavier in its first half. A coder that learns repeats one bar at a time says the halves are equal, and the two memories disagree by more than any proportion a listener could confuse.

 A final chord stands above the impression for a fraction of a second. How far a final chord at the tutti's own level stands above the listener's running impression at the instant it is released, against how long it lasts, for four ways of arriving at it. Straight out of the tutti it stands above nothing at any length; after 1.2 s of silence the impression is 8.1 dB down, and the chord stands highest, 5.14 dB, when it lasts 54 ms; after 3.5 s of silence the impression is 23.7 dB down, and the chord stands highest, 14.42 dB, when it lasts 28 ms; after a 6 dB diminuendo the impression is 5.0 dB down, and the chord stands highest, 3.18 dB, when it lasts 42 ms. Every curve is level again by half a second, because the impression's attack of 99 ms catches the note's attack of 22 ms, so a held chord is released at the impression's level whatever preceded it. Form and structure

A final chord stands out for a twentieth of a second

A general pause drives a listener's running impression down, and the final chord that follows is supposed to cash the fall in. It cashes in at most two thirds of it. The note's own loudness rises with a 22-millisecond constant and the impression with a 99-millisecond one, so after a bar of silence the chord stands furthest above the impression 54 milliseconds in, by 5.1 of the 8.1 decibels the silence bought, and after 206 milliseconds the two are within a phon of each other. A short stamp spends most of its life standing out; a chord held a second and a half spends a seventh of it.

 Four bars read by where the chords change, barline by barline. A constructed passage of four bars of eight quavers, its barline at the first slot and its chords C, F, Em, Dm. Notes: slot 1 C, slot 3 E, slot 4 G, slot 5 E, slot 6 C, slot 9 C, slot 10 C, slot 11 F, slot 13 A, slot 16 C, slot 17 B, slot 19 B, slot 20 E, slot 21 E, slot 24 G, slot 25 F, slot 29 F. For each of the eight places the barline could fall: as written metre, z 3.97, chords, z 2.00, change, z 4.30, metre + change, z 8.27; 1 quaver late metre, z -2.45, chords, z 2.14, change, z -0.87, metre + change, z -3.32; 2 quavers late metre, z -1.38, chords, z 2.15, change, z -0.54, metre + change, z -1.93; 3 quavers late metre, z -0.31, chords, z -0.27, change, z -2.64, metre + change, z -2.95; 4 quavers late metre, z 3.97, chords, z -1.63, change, z -4.30, metre + change, z -0.33; 5 quavers late metre, z -2.45, chords, z 1.27, change, z 0.87, metre + change, z -1.58; 6 quavers late metre, z -1.38, chords, z 1.18, change, z 0.54, metre + change, z -0.84; 7 quavers late metre, z -0.31, chords, z 1.05, change, z 2.64, metre + change, z 2.32. Best metre, z: as written and 4 late. Best chords, z: 2 late. Best change, z: as written. Best metre + change, z: as written. Harmony and voice leading

The chords mark the barline by changing there

Read bar by bar, the chords stood barely above chance at the barline and broke the metre's half-bar tie two times in three at best. Read instead by where they change — how different the chords are across a candidate's barlines against how different they are across the middle of its bars — the same notes break the tie right on 81 to 96 per cent of passages, and added to the metre they find the barline on up to 89 per cent against 61. The weakness was the question the old reading asked, not the harmony.

 Level does not dilute the register's roughness, it multiplies it. The mean roughness of the I – vi – IV – V – I arrivals at 4 registers, each relative to the register as written, read three ways. Level-free, the bass is 8.6 times rougher than the treble. With every note at 70 dB it is 8.6 times, the same factor, because one level rescales every pair alike. With each chord played at the level that makes it as loud as the written register's chords — 82.8 dB −2 octaves, 75.5 dB −1 octave, 70.0 dB as written, 66.8 dB +1 octave — the bass is 343 times rougher than the treble, because roughness grows with the square of the pressure and the bass needs more of it to be heard at the same loudness. Harmony and voice leading

A rough arrival is rough because of its spacing

The pair the expectation essays report for every chord — how surprising it was, how rough its voicing is — has no level in it. Putting level back in answers the question it left open, and not the way it was framed. At one written dynamic the arrivals keep their order from 40 to 90 dB at three registers of four, and the bass stays 8.6 times rougher than the treble. Made equally loud, the bass has to be played 12.8 dB harder, and it is 343 times rougher: level does not explain the register's roughness away, it multiplies it.

 The thirds' difference-tone line, note by note, against the dynamic. Each note of the line the difference tone f₂ − f₁ draws under a scale in just thirds, as its level above the higher of the threshold of hearing and the primaries' masking, against the level of the primaries. The product sits 50 dB below the primaries at 60 dB and grows twice as fast as they do. C2 above the limit from 73.5 dB; A1 above the limit from 76 dB; C2 above the limit from 73.5 dB; F2 above the limit from 70 dB; G2 above the limit from 68.5 dB; F2 above the limit from 70 dB; G2 above the limit from 68.5 dB; C3 above the limit from 66 dB. Intervals and chords

A combination-tone bass needs a forte

A scale in just thirds draws a diatonic bass line through its difference tones, and in sixths the cubic product draws one. Given the two published level laws, with their constants swept, the thirds' bass is not heard at all below primaries of about 66 dB and is heard whole only from 71 to 81. The cubic products are a different kind of object: the primaries mask them decibel for decibel as they rise, so no dynamic changes whether they are heard. Most of the thirds' inner line never is, and the sixths' bass needs a forte and a gentle law.

 A listener who knows every metre recognises none of them inside the present. For every bar length from nine units to twenty-five, the fewest and the most steps from the downbeat before every other one of the 1820 arrangements of twos and threes has been contradicted by the stream, on onsets alone, with the long beats accented and with the downbeat accented, against the 16 steps a present of 3.5 seconds holds. onsets alone: recognised within the present for 0 of 1820; long-beat accent: recognised within the present for 0 of 1820; downbeat accent: recognised within the present for 85 of 1820. The dashed line is the present. Rhythm and metre

Knowing every metre is slower than knowing none

A long aksak bar cannot be told from its shorter cuts by induction before one step into its last beat, and no accent carried by the notes moves that floor. The obvious escape is a listener who knows the repertoire and recognises the metre instead. Recognition among all 1,820 arrangements of twos and threes never beats the floor, is never quicker than induction, and is slower for half the metres: a nine induced in 9 steps is recognised in 27. What breaks the floor is a small repertoire that leaves out the metre's own longest cut — with the cut known, no repertoire of any size does.

 With a fading memory the slowest order of the gaps 1 1 2 2 2 2 2 is not the one a perfect memory finds. For each of the 3 cyclic orders of the gaps 1, 1, 2, 2, 2, 2, 2 in 12 steps, the bits of position still unknown once listening has settled, against how many steps a listener's memory of a step takes to halve, with a mismatch costing 6. 2 2 2 1 2 1 2: 12 → 4e-9, 6 → 4e-5, 4 → 0.007, 3 → 0.070, 2 → 0.540, 1.5 → 1.130, 1 → 1.786; 2 2 2 2 1 1 2: 12 → 2e-5, 6 → 0.035, 4 → 0.335, 3 → 0.795, 2 → 1.405, 1.5 → 1.694, 1 → 2.005; 2 2 1 2 2 1 2 (the standard bell pattern): 12 → 2e-5, 6 → 0.027, 4 → 0.231, 3 → 0.535, 2 → 1.028, 1.5 → 1.368, 1 → 1.819. With perfect memory the slowest to locate is the standard bell pattern; at a half-life of 6 steps the highest floor is the order 2 2 2 2 1 1 2, and at 1 it is the order 2 2 2 2 1 1 2. Rhythm and metre

The bell pattern is slowest only to a perfect memory

Among the orders of its own gaps, a named timeline is usually both the most even and the slowest to locate — for a listener who never forgets. Give the listener a memory that halves and the result comes apart. Of six timelines slowest among their orders with perfect memory, only the fume-fume stays slowest for every forgetting listener, and the standard bell pattern, which is the fume-fume with onsets and rests exchanged and settles at exactly the same floors, is second of its three orders for every memory of half its cycle or less. The census ranking survives better, and in fourteen of twenty-one censuses it was the arithmetic of a pattern that repeats.

 Come in part-way with the downbeat accented, and no bar of sixteen units or more is recognised inside the present. For every bar length from nine units to twenty-five, the fewest and the most steps a listener who knows every arrangement of twos and threes needs to recognise the metre and where its bar begins, with the downbeat accented: coming in at a sample of steps inside the bar, against hearing it from its written downbeat. On onsets alone, or with the long beats accented, a metre entered part-way is never told from its rotations. 9: from inside the bar 10 to 17, 16 of 20 inside the present; from the downbeat 10 to 10; 10: from inside the bar 11 to 19, 12 of 20 inside the present; from the downbeat 11 to 11; 11: from inside the bar 12 to 21, 9 of 18 inside the present; from the downbeat 12 to 12; 12: from inside the bar 13 to 23, 8 of 24 inside the present; from the downbeat 13 to 13; 13: from inside the bar 14 to 25, 8 of 28 inside the present; from the downbeat 14 to 14; 14: from inside the bar 15 to 27, 4 of 28 inside the present; from the downbeat 15 to 15; 15: from inside the bar 16 to 29, 4 of 32 inside the present; from the downbeat 16 to 16; 16: from inside the bar 17 to 31, 0 of 32 inside the present; from the downbeat 17 to 17; 17: from inside the bar 18 to 32, 0 of 36 inside the present; from the downbeat 18 to 18; 18: from inside the bar 19 to 32, 0 of 36 inside the present; from the downbeat 19 to 19; 19: from inside the bar 20 to 37, 0 of 40 inside the present; from the downbeat 20 to 20; 20: from inside the bar 21 to 35, 0 of 40 inside the present; from the downbeat 21 to 21; 21: from inside the bar 22 to 40, 0 of 44 inside the present; from the downbeat 22 to 22; 22: from inside the bar 23 to 39, 0 of 44 inside the present; from the downbeat 23 to 23; 23: from inside the bar 24 to 44, 0 of 48 inside the present; from the downbeat 23 to 24; 24: from inside the bar 23 to 39, 0 of 48 inside the present; from the downbeat 23 to 25; 25: from inside the bar 22 to 44, 0 of 52 inside the present; from the downbeat 23 to 25. In all, 61 of 590 entries are recognised within the 16 steps of a 3.5-second present. Rhythm and metre

A dancer who comes in late needs the downbeat marked

Every window for recognising an aksak metre so far started at its written downbeat. A dancer joining a dance already going has not heard the downbeat, and the arithmetic of that is blunt: a metre entered part-way is, onset for onset, each of its own rotations heard from their downbeats, and the rotations are metres too — 2+2+3 and 3+2+2 are counted differently. So on onsets, and with the long beats accented, no metre is ever told from its rotations. Only an accented downbeat tells them apart, and with it a listener who knows thirty metres recognises 54 per cent of them inside the present from a random entry, against 1 per cent without.

 Every product of every pair of partials is a harmonic of one absent note. A 4 : 5 interval on 261.6 and 327.0 hertz, just, each note carrying 6 partials at one over n, with the fundamental at 80 decibels. Every pair of partials makes its own difference tone, and there are 19 distinct frequencies among them. All of them are exact multiples of 65.41 hertz — the note neither instrument is playing — because partial j of a p·f₀ note and partial k of a q·f₀ note differ by (kq − jp)·f₀ whatever j and k are. The heavy line is what each product has to clear — the threshold of hearing at its own frequency, or the masked threshold the two primaries cast there, whichever is higher. 11 of the 19 get through. Intervals and chords

Three harmonics of the bass arrive before the bass

Every product priced until now was between two pure tones, and nothing that plays thirds is pure. Give each note a spectrum and the ear receives every pair of partials — and for a just interval p:q every one of their products is an exact multiple of the same absent fundamental. That crowd lands where the threshold of hearing is tens of decibels cheaper, so it names the bass at 71 decibels where the component at the bass's own frequency needs 74, and at 75 against 85 an octave lower. Tempered, the crowd still forms and names a note seventy cents flat.

 The ghost bass drops when the passage gets louder. The note the whole crowd of products names, as a multiple of the fundamental the interval implies, against how loudly the interval is played. a major third: 2.9999999999999996 times the fundamental below 70 decibels and 1 times above it, a drop of 19 semitones; a minor third: 4 times the fundamental below 62 decibels and 2 times above it, a drop of 12 semitones; a fourth: 2 times the fundamental below 72 decibels and 1 times above it, a drop of 12 semitones. Softly, only the cubic products clear their thresholds, and they are an exact series on (2p − q) times the fundamental with no gaps in it. Loudly, the difference tones fill in the low harmonics, no template on the higher note can explain them, and the fit falls. Nothing about the interval has changed; the listener is simply being given a different subset of the same harmonic series. Intervals and chords

The ghost bass drops a twelfth at a forte

Both crowds arrive at once and every member of both is a multiple of the same absent fundamental, so a listener is never given a choice between them — only a different subset of one harmonic series at every dynamic. Softly, the subset is an exact gapless series on three times the fundamental. Loudly, the difference tones fill in the low harmonics and no template on the higher note survives them. Between 62 and 72 decibels, depending on the interval, the note the crowd names falls by an octave or a twelfth, and the two qualities of third cross at different levels.

 A count is least reliable at both ends and best in the middle. How precisely a listener knows the length of a section they are counting, against how many units long it is, at three kinds of timing judgement. A slip — one unit miscounted, at 2% a unit — accumulates as a random walk, so its relative cost FALLS as the section lengthens. A lapse — the count lost altogether, at 1% a unit — compounds, so the chance of still having the count falls geometrically and a long enough section is certain to lose it. A listener who has lost the count is back to timing, so the two failures mix into a floor. Against a Weber fraction of 7.5% the count is worth most at 21 units, where it is 1.7 times finer than timing, and falls back under a quarter better by 98 units; Against a Weber fraction of 15% the count is worth most at 10 units, where it is 2.4 times finer than timing, and falls back under a quarter better by 101 units; Against a Weber fraction of 38% the count is worth most at 4 units, where it is 3.7 times finer than timing, and falls back under a quarter better by 102 units. The length at which it stops being worth much is nearly the same in all three, because it is set by the lapse rate alone. Form and structure

A count is not an estimate

Both established routes to a proportion are estimates — a duration timed, blurred by a Weber fraction, and a duration stored, biased by what was new. A listener who has induced a hypermetre has a third, and it is exact until it fails. It fails two ways that pull opposite: a slip miscounts one unit and its relative cost falls as the section lengthens, while a lapse loses the count entirely and its chance compounds. The mixture has a floor at about four units, where counting is 3.7 times finer than timing, and it is worth almost nothing past a hundred.

 Given the bar in octaves, the cue gets the degree back. How often the reading names the right scale degree, against how much the bass cue is worth, under three rules for what a bass note rewards. the bass is the root: 68 per cent at best; the bass is some chord tone: 67 per cent at best; the bar in octaves: 92 per cent at best; the bass is the root, on a root line: 89 per cent at best. The published cue rewards the triad rooted on the bass, which is right only when the chord is in root position; rewarding any triad containing the bass is right always and rewards three chords a bar instead of one. The third rule is not a bass cue at all — given the bar voiced in octaves a listener knows which three pitch classes are sounding, and that names the chord outright. It reaches 92 per cent on a real line against the published cue's 68 on the same line, and the dashed curve is what that cue manages on a line made entirely of roots — 89 per cent. Scales and modes

Given the bar in octaves, the degree comes back

A bass note is a bare pitch class in the usual figure, and a register in any realisation anybody plays. Voice each bar in octaves and a listener knows which three pitch classes are sounding and which is at the bottom, which names the chord outright — and the rule that uses it reads the right scale degree in 92 per cent of bars on a real bass line, against 68 for the published cue on the same line and 89 for that cue on a line made entirely of roots. The question was whether the register recovers the 89. It recovers it and passes it.

 A displaced map is displaced by the same amount everywhere. How far a listener's heard direction is displaced, in units of the smallest angular change they could detect at that azimuth, for four constant offsets added to every interaural delay. Each curve is flat. an offset of 5 microseconds is worth 0.33 just-noticeable steps at every azimuth; an offset of 10 microseconds is worth 0.67 just-noticeable steps at every azimuth, and past 88° hands the listener a delay their own head cannot produce; an offset of 20 microseconds is worth 1.33 just-noticeable steps at every azimuth, and past 86° hands the listener a delay their own head cannot produce; an offset of 40 microseconds is worth 2.67 just-noticeable steps at every azimuth, and past 82° hands the listener a delay their own head cannot produce. The reason is exact: differentiating Woodworth's curve gives a slope proportional to (1 + cos θ), so the angular displacement a fixed offset produces carries a factor of 1/(1 + cos θ) — and so does the smallest detectable angle, so the ratio has no azimuth in it. That is the opposite of a wrong head radius, whose displacement is zero on the median plane and grows toward the side. Perception and the listener

The error that moves straight ahead

The essay before this one found that a listener whose internal head is the wrong size makes no error at all on the median plane, and has to look hard to the side to catch it. Every head drawn here has its ears at equal radii, which makes the delay curve odd and every error a factor — and a factor cannot move a zero. Real heads are not symmetric. A constant offset of twenty microseconds displaces a listener's straight ahead by two and a quarter degrees, and it displaces every other direction by the same number of just-noticeable steps, exactly.

 With 4 of six required at the end, the best schedule hands parts over. Six parts entering, leaving and re-entering a five-chord passage over four sounding players, in the walk through all 64 sets of sounding parts that makes the least audible entrance as audible as possible, with no memory of the chord before, at least 4 of the six sounding at the last chord. I, open: clarinet on G4 enters at 0.6 dB; vi, close below: oboe on E4 enters at 0.3 dB, and clarinet leaves; IV, close above: brass on E3 enters at -0.7 dB, voice on C5 enters at 2.2 dB, and oboe leaves; V, bracketing: violin on G5 enters at 7.2 dB; I, hollow: flue pipe on C6 enters at 4.0 dB. The least audible entrance is -0.72 dB against -2.65 for the best schedule in which nobody leaves; the walk has 2 exits and 6 entrances. Form and structure

An exit is worth nothing until the tutti is given up

Six parts entering a five-chord passage have a best schedule when each enters once and stays, and letting parts leave and come back was supposed to improve it. Searched over every set of sounding parts at every chord, it improves it by exactly nothing, with or without the chord before still masking — as long as all six must be playing at the end. Let one part be missing from the final chord and the weakest entrance gains 1.4 decibels; let two be missing and it gains 1.9, by a relay in which the parts with least room come in, are heard for one chord, and give way.

 Checkpoints sharpen the middle of a form and leave its top vague. A piece of 480 seconds divided 7 times, with how many proportions between 1 : 1 and 3 : 1 a listener tells apart at each level: timed, counted in 2-second units, and counted with a second count of 16-second phrases that can mend a lapse in the first. 480 s: 2.1 timed, 2.2 counted, 3.9 with 0 per cent of lapses shared and 2.5 with 50 per cent of lapses shared; 240 s: 2.1 timed, 2.5 counted, 7.1 with 0 per cent of lapses shared and 3.1 with 50 per cent of lapses shared; 120 s: 2.1 timed, 3.1 counted, 13.2 with 0 per cent of lapses shared and 4.0 with 50 per cent of lapses shared; 60 s: 2.1 timed, 4.1 counted, 19.4 with 0 per cent of lapses shared and 5.4 with 50 per cent of lapses shared; 30 s: 2.1 timed, 5.4 counted, 18.1 with 0 per cent of lapses shared and 7.2 with 50 per cent of lapses shared; 15 s: 5.2 timed, 12.2 counted, 12.2 with 0 per cent of lapses shared and 12.2 with 50 per cent of lapses shared; 7.5 s: 5.2 timed, 10.3 counted, 10.3 with 0 per cent of lapses shared and 10.3 with 50 per cent of lapses shared. With the two counts failing independently, the level of 60 seconds goes from 4.1 to 19.4, and the whole piece only from 2.2 to 3.9. Form and structure

Checkpoints sharpen the middle of a form, not its top

A listener who counts bars loses the count somewhere in a long section and is thrown back on timing the whole of it. A listener who also counts phrases can mend the lapse at the last phrase. If the two counts fail independently, the level a minute long goes from four distinguishable proportions to nineteen; the whole eight-minute piece goes only from two to four, because thirty phrases are long enough to lose a count as well. And if a fifth of lapses take both counts at once, three quarters of the gain is gone.

 The content errs slow and the bass errs fast. Passages of eight bars built at three harmonic rhythms, each with a bass that states every new chord's root and moves to another chord tone on a beat 50 per cent of the time. Two readings are asked the chord rate: one from how much the pitch-class content changes across a grid, one from how completely the bass's moves land on it. At two chords a bar the content reading names the rate 63 per cent of the time, too fast 0 and too slow 37; the bass reading 100, 0 and 0. At one chord a bar the content reading names the rate 67 per cent of the time, too fast 0 and too slow 33; the bass reading 18, 80 and 2. At a chord every two bars the content reading names the rate 98 per cent of the time, too fast 2 and too slow 0; the bass reading 0, 100 and 0. The two readings miss in opposite directions and at opposite ends of the tempo range: the content reading at fast harmonic rhythms, by naming a multiple, and the bass reading at slow ones, by naming its own arpeggiation. Harmony and voice leading

The bass errs fast where the content errs slow

Asked how often the chords change, a reading built on pitch-class content names a slower multiple and never a faster rate. Give the passage a bass that states each new root and moves between chord tones inside a chord, and a reading built on the bass's moves errs the other way: it names a faster grid and never a slower one. At two chords a bar the bass is right every time; at a chord every two bars it is never right. Six ways of combining the two readings each trade one end of the range for the other.

 Put back beside its notes, the crowd names the bass at every dynamic. A just major third on complex tones, drawn on one axis of harmonic numbers of the fundamental its ratio implies: the partials of the two played notes, and the products of those partials that clear threshold, at 55 and 80 dB. At 55 dB the products alone name 3 times the fundamental with 0 empty slots; the products and the notes together name 1 times it with 4 empty slots, and the notes alone name it with 9. At 80 dB the products alone name 1 times the fundamental with 0 empty slots; the products and the notes together name 1 times it with 0 empty slots, and the notes alone name it with 9. The soft reading on a higher note exists only when the loud notes are set aside. Taken together, the products do not decide which fundamental is named; they decide how many holes its template has. Intervals and chords

The played notes already name the ghost bass

The products of a just third's partials, fitted on their own, name a note a twelfth above the bass when the interval is soft and drop to the bass when it is loud. Put the two played notes back beside them and the drop disappears: the notes and their products name the bass at every dynamic, because the notes' own partials are harmonics of it already. What the dynamic changes is not which note is implied but how complete its harmonic series is — nine holes from the notes alone, four when soft, none when loud.

 The ceiling is thirteen bars, and every one of them has a name. Every bar of the six schemes read by the key-finder at a key cost of 3 and a bass weight of 3, with the chord in every bar given exactly — no segmentation is involved. 175 of 188 bars have both the key and the degree right, 93.1 per cent. The misses: 7 in the thirty-two-bar song's bridge, where the chain III7 is read in E major, III7 is read in E major, VI7 is read in E major, VI7 is read in D major, II7 is read in D major, II7 is read in D major, V7 is read in D major; 4 bars of the rondo's A minor episode read as C major; 2 next to a change of key; and 0 of any other kind. Scales and modes

The ceiling is thirteen bars with names

The key-finder's two cues stop buying anything at about 93 per cent of scheme bars read right, and the obvious suspect was the chord segmentation feeding it. The reading was never given a segmentation: every bar arrives with its true chord. What the ceiling is made of can be listed instead, and it is thirteen bars of 188 — seven in a bridge of secondary dominants, four in a minor episode whose chords C major also owns, and two at the edges of a modulation. No weight of any cue moves one of them.

 A listener who hears only the landmarks recognises the metre sooner. The share of metres recognised within the 16 steps of a 3.5-second present after coming in at a random step, against how many metres the listener knows, for five listeners: every onset with nothing marked, every onset with the long beats accented, only the downbeats and long beats with nothing marked, every onset with the downbeat accented, and only the downbeats and long beats with the downbeat marked. Every onset, nothing marked: 5 known, 49% within the present, 3% never; 10 known, 19% within the present, 8% never; 30 known, 1% within the present, 17% never; 100 known, 0% within the present, 34% never. Every onset, long beats accented: 5 known, 68% within the present, 3% never; 10 known, 39% within the present, 8% never; 30 known, 6% within the present, 17% never; 100 known, 0% within the present, 34% never. Landmarks only, nothing marked: 5 known, 69% within the present, 0% never; 10 known, 61% within the present, 3% never; 30 known, 26% within the present, 3% never; 100 known, 9% within the present, 9% never. Every onset, downbeat accented: 5 known, 88% within the present, 0% never; 10 known, 74% within the present, 0% never; 30 known, 54% within the present, 0% never; 100 known, 14% within the present, 0% never. Landmarks only, downbeat marked: 5 known, 91% within the present, 0% never; 10 known, 82% within the present, 0% never; 30 known, 56% within the present, 0% never; 100 known, 23% within the present, 0% never. Rhythm and metre

A late dancer needs the landmarks, not the rhythm

A listener who joins an additive-metre dance part-way recognises it far more reliably with the downbeat accented. Strip the stream down to its landmarks — the onsets that begin a bar or a long beat, with every other onset removed — and the listener does as well or better: knowing a hundred metres, 23 per cent are recognised within the present against 14 with every onset. Unmarked, the landmarks still beat every onset with the long beats accented. Neither half does it alone; what identifies a metre from a late entry is where its long beats sit relative to its bar.

 Against a pulse, the bell pattern is the quickest of its orders to place. The bits of position a listener is still missing, averaged over the first cycle heard, for each cyclic order of the gaps 1 1 2 2 2 2 2 in 12 steps, heard alone, against a pulse every three steps and against a pulse every four, at perfect memory, half-life 3 steps, half-life 1.5 steps. 2 2 2 1 2 1 2: alone 1.08, 1.23, 1.73; against a pulse every 3 steps 0.69, 0.75, 0.96; against a pulse every 4 steps 0.63, 0.67, 0.85. 2 2 2 2 1 1 2: alone 1.22, 1.63, 2.02; against a pulse every 3 steps 0.60, 0.73, 0.92; against a pulse every 4 steps 0.83, 1.07, 1.36. 2 2 1 2 2 1 2 (the standard bell pattern): alone 1.25, 1.45, 1.82; against a pulse every 3 steps 0.54, 0.56, 0.67; against a pulse every 4 steps 0.55, 0.57, 0.68. Alone, the bell pattern is not the quickest order to place at any memory. Against either pulse it is the quickest at every memory. Rhythm and metre

Against a pulse the bell pattern is the easiest to place

Heard alone, the standard bell pattern is not the quickest order of its own gaps to place in its cycle, for a listener with any memory. Heard against a pulse every three steps or every four — which is how anyone hears it — it is the quickest, at every memory and at every alignment of pulse and bell, and by a wide margin: at a memory of a quarter of the cycle, 0.56 bits unplaced over the first cycle against 0.73 for either rival against a pulse in threes. Six of eight named timelines do the same. A timeline's order of gaps looks chosen for how it sits against the beat, not for how it sounds alone.

 Counted over what arrives, the balanced bass is not the roughest register. The mean roughness of the I – vi – IV – V – I arrivals with each chord played as loud as the written register's, relative to the written register, counted over every partial and over the partials that stand above what the rest of the chord masks. Every partial: 70 −2 octaves, 7.48 −1 octave, 1.00 as written, 0.20 +1 octave. Delivered partials only: 6e-9 −2 octaves, 2.83 −1 octave, 1.00 as written, 0.19 +1 octave. Over every partial the lowest register is 343 times rougher than the highest; over what arrives it is the smoothest of the four, and the roughest is −1 octave, 2.8 times the written register. Perception and the listener

A bass chord low enough to balance has already hidden its tenor

Played as loud as the written register, a progression two octaves down is 343 times rougher than the same progression an octave up — if every partial on the page is counted. Count only the partials that stand above what the rest of the chord masks and that register is the smoothest of the four, with nothing left that beats. The balance is not what does it: the extra thirteen decibels move no voice by more than two partials. The register had already buried the tenor at the written dynamic.

 Holding the bass through one mid-bar change in six finds the barline five times in six. Passages of eight bars at two chords a bar, with the bass arpeggiating on 50 per cent of its beats, and each barline reading's share of passages it places correctly, against the share of mid-bar chord changes voiced over the bass already sounding. 0% held (convention strength 0): metre then bass 50%, bass alone 13%, metre alone 50%, chord changes alone 10%; 4% held (convention strength 0.1): metre then bass 62%, bass alone 33%, metre alone 50%, chord changes alone 10%; 6% held (convention strength 0.2): metre then bass 68%, bass alone 41%, metre alone 50%, chord changes alone 15%; 9% held (convention strength 0.3): metre then bass 75%, bass alone 50%, metre alone 50%, chord changes alone 18%; 15% held (convention strength 0.4): metre then bass 81%, bass alone 66%, metre alone 50%, chord changes alone 13%; 16% held (convention strength 0.5): metre then bass 85%, bass alone 71%, metre alone 50%, chord changes alone 14%; 20% held (convention strength 0.6): metre then bass 86%, bass alone 74%, metre alone 50%, chord changes alone 11%; 26% held (convention strength 0.7): metre then bass 90%, bass alone 82%, metre alone 50%, chord changes alone 13%; 27% held (convention strength 0.8): metre then bass 91%, bass alone 82%, metre alone 50%, chord changes alone 13%; 28% held (convention strength 0.9): metre then bass 95%, bass alone 89%, metre alone 50%, chord changes alone 10%; 33% held (convention strength 1): metre then bass 96%, bass alone 90%, metre alone 50%, chord changes alone 10%. The metre ties the barline with the half-bar and the chord changes are at chance, since the chords change at both; the bass's holds are the only evidence that separates them. Harmony and voice leading

A bass that holds through a change marks the barline

At two chords a bar the chords change on the barline and on the half-bar alike, so a reading of where they change is at chance, and the metre ties the two. The bass has one more piece of evidence: a change inside the bar can be voiced over the note already sounding, and a change on the barline is voiced over its root. Hold the bass through one mid-bar change in six and the metre and bass together place the barline in 85 per cent of eight-bar passages; one in three, 97. The convention cannot be stronger than that, and the chord changes, asked first, only get in the way.

 As the bass stops standing still, what marks the barline stops being its motion and becomes its note. Passages of eight bars at one chord a bar, and how often each reading places the barline, against how often the bass moves to another tone of the chord on a crotchet. Every reading is the metre's own candidates with one witness breaking its tie. bass moving on 0% of crotchets: moves 100%, root 50%, surer of the two 100%, metre alone 50%; bass moving on 13% of crotchets: moves 100%, root 94%, surer of the two 100%, metre alone 50%; bass moving on 25% of crotchets: moves 100%, root 99%, surer of the two 100%, metre alone 50%; bass moving on 38% of crotchets: moves 99%, root 100%, surer of the two 100%, metre alone 50%; bass moving on 50% of crotchets: moves 95%, root 100%, surer of the two 100%, metre alone 50%; bass moving on 63% of crotchets: moves 88%, root 100%, surer of the two 100%, metre alone 50%; bass moving on 75% of crotchets: moves 77%, root 100%, surer of the two 100%, metre alone 50%; bass moving on 88% of crotchets: moves 54%, root 100%, surer of the two 100%, metre alone 50%; bass moving on 100% of crotchets: moves 20%, root 100%, surer of the two 94%, metre alone 50%. A bass that never moves inside the chord moves only at chord changes, which at this rate are barlines, so the move reading is certain and the root sounds throughout the bar and says nothing; a bass that moves on every crotchet moves everywhere and sounds the root only where the chord has just arrived. Harmony and voice leading

A walking bass marks the bar by the note it is on

The reading that finds a barline from where the bass moves collapses to twenty per cent when the bass moves on every crotchet, and the bass the last essay called busiest moved twice a bar, because its beat was the half-bar. Given one that really walks, the same computation with sounds the chord's root in place of moves rises from nothing to a whole note a bar exactly as the other falls, and the two are equal at a bass moving on three crotchets in eight. Neither places a barline alone: the root is still sounding on the quaver after it. The metre, which cannot tell a barline from a half-bar, tells it from the quaver after it at once.

 Two chords a bar over a moving bass: every witness there is, at chance together. Passages of eight bars at two chords a bar, with the chords always voiced over their own roots, and how often each reading places the barline against how often the bass moves inside the chord. 0%: metre alone 50%, with the moves 50%, with the root 50%, with both 50%; 13%: metre alone 50%, with the moves 49%, with the root 50%, with both 49%; 25%: metre alone 50%, with the moves 49%, with the root 50%, with both 49%; 38%: metre alone 50%, with the moves 49%, with the root 50%, with both 49%; 50%: metre alone 49%, with the moves 53%, with the root 49%, with both 54%; 63%: metre alone 50%, with the moves 54%, with the root 50%, with both 54%; 75%: metre alone 50%, with the moves 48%, with the root 50%, with both 48%; 88%: metre alone 50%, with the moves 51%, with the root 50%, with both 51%; 100%: metre alone 50%, with the moves 51%, with the root 50%, with both 51%. The metre ties the barline with the half-bar because both carry the same strong slots; the chords change at both; the bass states a root at both and moves between both. Nothing in the passage distinguishes them. Harmony and voice leading

A texture in which nothing says where the bar begins

Two chords a bar over a bass that moves inside the chord is the one arrangement in which every witness this account has stands together at the metre's coin toss. The metre ties the barline with the half-bar, the chords change at both, and the bass states a root at both and walks between both — at every busyness, and whichever question the bass is asked. The only thing that breaks the symmetry is the voicing convention, and what it is worth falls from 98 per cent to 90 as the bass fills in, not because fewer changes can be held but because the motion around them is evidence for nothing. Chasing five points of residue above the toss found a defect that had been flattering the reading since it was written.

 The leap at the barline is not in the harmony: it is in which octave the root is played in. How often the size of the bass's motion places the barline, at one chord a bar over eight bars, against how often the bass moves inside the chord, under two realisations of the same pitch classes in a register. Roots at the bottom of the compass: 100% at 0%, 100% at 13%, 100% at 25%, 99% at 38%, 96% at 50%, 95% at 63%, 93% at 75%, 92% at 88%, 90% at 100%. Least motion: 100%, 100%, 96%, 91%, 83%, 69%, 52%, 32%, 20%. Also drawn, the reading of whether the bass moved at all: 100%, 100%, 99%, 99%, 93%, 89%, 70%, 54%, 20%. The two realisations use the same notes in the same order and differ only in which octave each is played in. Harmony and voice leading

The leap at the barline is a fact about register

A bass leaps to a root at the barline and steps inside the bar, and no reading of pitch classes can see it — so the essay that named it recorded it as a limit. Given octaves and the size of every motion, it is worth nothing: realised at least motion the reading tracks whether the bass moved at all and ends 0.87 semitones on the wrong side of zero, because a line that always takes the nearest octave has no leaps to take. Play a stated root at the bottom of the compass instead, and the same sixteen notes in the same order turn a margin of minus 1.67 semitones into plus 1.67 and hold at 90 per cent where the move reading has fallen to 20. The barline is not where the bass moves furthest. It is where the bass goes back down.

 Relabelling the bridge does not rescue it. The thirty-two-bar song's eight bridge bars at a key cost of 3 and a bass weight of 3: the chord in each, the key the scheme labels it in, the key its dominant points at, and what the reading names. III7: scheme says C major, the tonicization says A minor, the reading says E major degree 1 — the chord's own root as a tonic; III7: scheme says C major, the tonicization says A minor, the reading says E major degree 1 — the chord's own root as a tonic; VI7: scheme says C major, the tonicization says D minor, the reading says E major degree 4; VI7: scheme says C major, the tonicization says D minor, the reading says D major degree 5; II7: scheme says C major, the tonicization says G major, the reading says D major degree 1 — the chord's own root as a tonic; II7: scheme says C major, the tonicization says G major, the reading says D major degree 1 — the chord's own root as a tonic; V7: scheme says C major, the tonicization says C major, the reading says D major degree 4; V7: scheme says C major, the tonicization says C major, the reading says C major degree 5. Right under the scheme's label: 1 of 8. Under the tonicization: 1. Under the tonicization with the mode set aside: 2. Read as its own tonic: 4. Scales and modes

The bridge is not a labelling problem

Seven of the key-finder's thirteen remaining misses are the thirty-two-bar song's bridge, and the essay before this one proposed that they might be an argument about labels rather than a failure: relabel the eight bars with the keys their dominants point at, and the reading's E and D majors might turn out to be right. They do not. The reading gets one bar of eight under either label, because four of the eight are read as the chord's own root taken for a tonic — a dominant seventh heard as a chord in its own right. And one of the bars the relabelling could have bought is unwinnable in principle: an applied dominant is the same four notes in the major and the minor of the key it points at.

 A count at every level turns a form's vague top into its sharpest level. A piece of 480 seconds divided 7 times, with how many proportions between 1 : 1 and 3 : 1 a listener tells apart at each level, counted three ways: in 2.0 s units alone, with one checkpoint count every 16 s, and with a count at every level of 2.0 s, 16 s, 64 s, 128 s, each mending the one below, every count failing independently. 480 s: 2.2, 3.9 and 76.4; 240 s: 2.5, 7.1 and 54.0; 120 s: 3.1, 13.2 and 38.1; 60 s: 4.1, 19.4 and 19.4; 30 s: 5.4, 18.1 and 18.1; 15 s: 12.2, 12.2 and 12.2; 7.5 s: 10.3, 10.3 and 10.3. The whole piece goes from 2.2 with one count and 3.9 with one checkpoint to 76.4 with a count at every level. Form and structure

Only a lapse of attention keeps a form's top vague

A listener who keeps a count at every level of a form — bars, phrases, periods, sections — each mending the one below when it lapses, does not have a vague top at all. The whole eight-minute piece becomes the sharpest level, at seventy-six distinguishable proportions against two. Failures shared between neighbouring counts barely dent it, because a shared failure has to climb every level to reach the top. What puts the vagueness back is a lapse that takes every count at once, and one lapse in a hundred of that kind is enough.

 A lapse of attention that ends costs a form nothing unless a section is heard twice in a row. How many proportions between 1 : 1 and 3 : 1 a listener who knows the form tells apart across a whole 480-second piece, counted at every level of 2.0 s, 16 s, 64 s, 128 s, with 10 per cent of lapses taking every count at once and lasting 10 seconds on average, after which the listener re-anchors at the section the music says they are in. through-composed, ABCDEFGH: 77.1, 0.0 per cent of returns misplaced; rondo, ABACABAD: 76.4, 0.0 per cent of returns misplaced; sonata, exposition repeated, ABABCDAB: 76.6, 0.0 per cent of returns misplaced; minuet and trio, halves repeated, AABBCCDDAB: 50.2, 5.7 per cent of returns misplaced; thirty-two-bar song, twice, AABAAABA: 47.7, 4.3 per cent of returns misplaced; strophic, AAAAAAAA: 35.4, 10.1 per cent of returns misplaced. With lapses that never end every form reads 4.5, and with every returning listener placed exactly 77.2. The forms that lose most are the ones that play a section twice in a row, not the ones with the most repetition. Form and structure

A repeat sign loses a listener that a refrain does not

A lapse of attention loses every count a listener is keeping, and when it ends the listener has only the music to say where they are. A listener who knows the form comes back to a rondo's refrain and is almost never misplaced, because the episodes between refrains make only one of them reachable. They come back near the join of a section repeated immediately and are misplaced nearly half the time, and a form built of such repeats loses between a third and a half of the precision a count at every level had restored. One of the most familiar signs in notation does more damage than the most repetitive form.

 Timed from every layer, a slow colotomic cycle is placed almost at once. Bits of position a listener with a 3.5-second memory is still missing over the first cycle, against how long the cycle takes, for a 16-step colotomy with layers every 16, 8, 4 steps: a newcomer, a listener timing from the last gong, a listener timing from the last stroke of every layer, and a listener timing from the gong to ten per cent; with son clave, timed from its own cycle, for comparison. a colotomy, newcomer: 1.6 s 0.71, 2 s 0.71, 3 s 0.72, 4 s 0.73, 6 s 0.75, 8 s 0.78, 12 s 0.85, 16 s 0.93, 24 s 1.14, 32 s 1.37, 40 s 1.60, 48 s 1.83, 60 s 2.14; a colotomy, timing a whole cycle: 1.6 s 0.27, 2 s 0.45, 3 s 0.45, 4 s 0.45, 6 s 0.46, 8 s 0.46, 12 s 0.47, 16 s 0.48, 24 s 0.54, 32 s 1.21, 40 s 1.45, 48 s 1.70, 60 s 2.02; a colotomy, timing from every layer: 1.6 s 0.00, 2 s 0.00, 3 s 0.00, 4 s 0.00, 6 s 0.00, 8 s 0.00, 12 s 0.01, 16 s 0.01, 24 s 0.01, 32 s 0.03, 40 s 0.11, 48 s 0.17, 60 s 0.20; a colotomy, from the gong to ten per cent: 1.6 s 0.34, 2 s 0.34, 3 s 0.34, 4 s 0.34, 6 s 0.34, 8 s 0.35, 12 s 0.35, 16 s 0.35, 24 s 0.37, 32 s 0.42, 40 s 0.49, 48 s 0.60, 60 s 0.80; son clave, timing the cycle: 1.6 s 0.32, 2 s 0.68, 3 s 0.68, 4 s 0.68, 6 s 0.68, 8 s 0.69, 12 s 0.70, 16 s 0.73, 24 s 0.85, 32 s 1.63, 40 s 1.90, 48 s 2.12, 60 s 2.38. At 40 seconds the clock leaves 0.11 bits, against 1.45 timing from the gong and 0.49 for a ten-per-cent timer. Rhythm and metre

The strokes between the gongs are a clock

An experienced listener arriving in a forty-second gong cycle was expected to be no better placed than a newcomer, because a whole cycle is timed in the band where a Weber fraction is near forty per cent. But nobody times a gong cycle as one interval. A listener who resets at the last gong, the last kempul and the last kenong times three short intervals instead of one long one, and arrives knowing their position to a tenth of a bit — better than a listener timing the whole cycle to ten per cent. A single timeline, whose strokes are all one instrument, offers nothing to reset on.

 With both pulses sounding, the bell pattern's lead over the other orders of its gaps disappears. The bits of position a listener with a memory half-life of 3 steps is still missing over the first cycle, averaged over every way each order can sit against the pulses, for each order of the gaps 1 1 2 2 2 2 2 in 12 steps, heard alone, against a pulse every three steps, against a pulse every four and against both; the dashed line is the two pulses alone, with no bell, at 0.53. 2 2 2 1 2 1 2: alone 1.234, every 3 steps 0.738, every 4 steps 0.672, both pulses 0.437. 2 2 2 2 1 1 2: alone 1.630, every 3 steps 0.734, every 4 steps 1.037, both pulses 0.422. 2 2 1 2 2 1 2 (the standard bell pattern): alone 1.452, every 3 steps 0.560, every 4 steps 0.564, both pulses 0.425. The standard bell pattern leads its nearest rival by 0.174 bits against the pulse in threes and 0.107 against the pulse in fours; against both, the three orders are within 0.015 bits and its lead is -0.003. Rhythm and metre

Against both feels, the bell pattern's order stops mattering

Against a pulse in threes, or a pulse in fours, the standard bell pattern is placed a sixth or a tenth of a bit sooner than either other order of its own gaps. A twelve-step cycle in the ensembles that play it is heard against both feels at once, and the two pulses together are a position code of their own: they coincide once a cycle and place a listener within three steps. Heard against both, the three orders are within a seventieth of a bit of one another and the bell's lead is gone — the advantage its order was shaped for lives only where one feel is dropped.

 An octave down, the ghost bass needs notes twice as long before it is there at all. The share of each note of a C major scale in just thirds that has its difference-tone bass, against how long each note lasts, for the passage starting on C3, C4 and C5, with the bass heard only after 4 of its own cycles. Thirds from C3: 40 ms 0%, 60 ms 0%, 80 ms 3%, 100 ms 12%, 125 ms 22%, 150 ms 33%, 200 ms 50%, 250 ms 60%, 300 ms 67%, 400 ms 75%, 500 ms 80%, 750 ms 87%, 1000 ms 90%; no note has its bass below 147 ms. Thirds from C4: 40 ms 3%, 60 ms 20%, 80 ms 38%, 100 ms 50%, 125 ms 60%, 150 ms 67%, 200 ms 75%, 250 ms 80%, 300 ms 83%, 400 ms 88%, 500 ms 90%, 750 ms 93%, 1000 ms 95%; no note has its bass below 73 ms. Thirds from C5: 40 ms 38%, 60 ms 58%, 80 ms 69%, 100 ms 75%, 125 ms 80%, 150 ms 83%, 200 ms 88%, 250 ms 90%, 300 ms 92%, 400 ms 94%, 500 ms 95%, 750 ms 97%, 1000 ms 98%; no note has its bass below 37 ms. Every step down an octave halves the bass and doubles the note length it needs. Intervals and chords

A ghost bass an octave lower needs notes twice as long

A scale played in just thirds has a bass underneath it that nobody plays — the difference tones of its thirds spell a diatonic line. Every essay on that bass has held its intervals still. A moving passage gives the bass a new pitch every note, and a low pitch is not there until it has run a few of its own cycles. With four cycles, a passage in thirds from C4 keeps its bass for 83 per cent of each note at three notes a second and loses it completely below 73 milliseconds a note. An octave down, every number doubles: the passage from C3 has no bass at all in its lowest note at semiquavers, and it is the scale's one Pythagorean third that sets the limit.

 Re-struck under a chord that has been decaying, a buried tenor gets its partials back. How many of the tenor's eight partials stand above the masking of the other three voices, for the first chord of I–vi–IV–V–I balanced for loudness in four registers, on a struck instrument whose held voices decay while the tenor is re-struck a step away at a stated time after the chord; beside it, the same tenor held from the start. Two octaves down: 0.25 s 1 moving, 0 held; 0.5 s 3 moving, 0 held; 1 s 6 moving, 0 held; 2 s 8 moving, 0 held; 4 s 8 moving, 0 held. An octave down: 0.25 s 6 moving, 1 held; 0.5 s 8 moving, 1 held; 1 s 8 moving, 1 held; 2 s 8 moving, 1 held; 4 s 8 moving, 0 held. As written: 0.25 s 8 moving, 4 held; 0.5 s 8 moving, 5 held; 1 s 8 moving, 5 held; 2 s 8 moving, 2 held; 4 s 8 moving, 1 held. An octave up: 0.25 s 8 moving, 3 held; 0.5 s 8 moving, 4 held; 1 s 8 moving, 3 held; 2 s 8 moving, 1 held; 4 s 8 moving, 0 held. Two octaves down, where the held tenor keeps nothing at any time, the moving one has six partials back by a second and all eight by two. Perception and the listener

A moving tenor is heard because it is struck again

A bass-heavy chord balanced for loudness hides its tenor under the bass's upper skirt, and arrangers keep inner voices moving in the low register as though motion were the cure. Counted as masking, motion alone is worth nothing: the note the tenor leaves stops masking the new one within ten milliseconds, and on a sustained instrument the moving tenor keeps exactly what the held one keeps. On a struck instrument everything changes, and for a reason that is not the motion. The held voices have been decaying since they were struck and the tenor's new note is struck now, so an octave down a tenor re-struck half a second into the chord has all eight of its partials back where the held tenor has one.

 Told where each dominant goes, the reading splits every bridge chord between arrival and departure. The thirty-two-bar song's eight bridge bars — two each of E7, A7, D7 and G7 — at a key cost of 3 and a bass weight of 3, read without the resolution cue and with it at a weight of 3, where a bar read as the dominant of a key earns the weight when the next bar is rooted on that key's tonic. Without: E I, E I, E IV, D V, D I, D I, D IV, C V. With: A V, A V, A I, D V, D I, G V, G I, C V. Read as the key its dominant points at, root only: 2 of 8 without and 5 with; read as a dominant at all, 2 and 5. Scales and modes

Looking ahead rescues the relative minor, not the bridge

A key-finding reading that scores each bar by its own notes cannot hear an applied dominant, whose meaning is in the bar after it. Given one cue about a pair of bars — a dominant earns credit when the next bar is rooted on its key's tonic — the thirty-two-bar song's bridge is read as the chain of dominants it is, each chord both the arrival of the last and the departure to the next. But the same cue hears a blues's opening tonic seventh as the dominant of its own IV, and at the setting that read the listing best it costs five bars for three. The cue earns its place only in a reading reluctant to change key, and there what it rescues is the rondo's episode in the relative minor.

 A listener whose ears differ by a decibel or two has a straight ahead that moves with the dynamic. Where the median plane sits for a listener whose two ears' thresholds differ, against the level of a 500 Hz tone, if the whole of the latency difference the published brainstem latency law gives reaches the comparator that reads interaural delay. 1 dB worse: 19.5° at 30 dB (172 µs), 12.0° at 60 dB (106 µs), 7.4° at 90 dB (65 µs); 2 dB worse: 40.6° at 30 dB (347 µs), 24.4° at 60 dB (214 µs), 14.9° at 90 dB (132 µs); 3 dB worse: 65.6° at 30 dB (524 µs), 37.7° at 60 dB (323 µs), 22.7° at 90 dB (200 µs). Every curve halves its offset every 43 dB and none reaches zero at any level a listener hears; the dashed line is one detectable step on the median plane, 1.68°. Perception and the listener

A quiet bass is further off-centre than a loud violin

An ear whose threshold is a decibel worse than the other is a little later to report every sound, and the published brainstem latency law says how much later: a gap that halves every 43 decibels and grows toward the bass. Taken at full size, a one-decibel asymmetry would put straight ahead twelve degrees round at a moderate level — seven times what the ear's own two thresholds say a decibel is worth, so most of that latency cannot be reaching the comparator that reads direction. What survives any scaling is the shape. The displacement is not one number but a curve in level and in pitch, and it deforms an orchestra rather than moving it.

 The landmark stream loses its lead when the listener allows for a lost landmark, before any is lost. The share of metres recognised within the 16-step present from a random entry, 10 metres known, against the rate at which played onsets go unheard, for the landmark stream with its downbeat marked and every onset with the downbeat accented. The listener asks for odds of 19 to one against every rival; a silence where a rival predicts an onset is worth 1/p, so one silence settles a rival below p = 0.053 and two are needed up to 0.229. landmarks, downbeat marked, onsets lost: 82% at 0, 77% at 0.05, 47% at 0.1, 15% at 0.4; landmarks, downbeat marked, caution only: 82% at 0, 82% at 0.05, 57% at 0.1, 46% at 0.4; every onset, downbeat accented, onsets lost: 74% at 0, 70% at 0.05, 58% at 0.1, 22% at 0.4; every onset, downbeat accented, caution only: 74% at 0, 74% at 0.05, 69% at 0.1, 69% at 0.4. Rhythm and metre

A crowd costs a late dancer the silences, not the steps

A dancer who hears only the landmarks of an aksak rhythm — the heads of bars and long beats — recognises it from a late entry better than one who hears every onset, as long as everything played is heard. Let a share of what is played be lost in a crowd and the lead goes, but not for the reason expected. Most of the cost arrives before a single landmark is missed: a listener who allows that one might be can no longer treat a silence as proof, and the landmark stream is built almost entirely of silences. At one lost onset in ten, the caution alone takes 25 points from the landmarks and 5 from the full stream.

 The notes name their ghost bass long before the bass has run a single cycle's worth of its count. A just major third, C4 against E4, with every component — each note's first six partials, the difference tone and the cubic tone — entering a harmonic fit once it has run 4 of its own cycles. at 2.0 ms, 1 component: nothing fitted; at 2.4 ms, 2 components: nothing fitted; at 2.5 ms, 3 components: names 103.3 Hz; at 3.1 ms, 4 components: names 82.0 Hz; at 3.8 ms, 5 components: names 86.3 Hz; at 4.1 ms, 6 components: names 81.8 Hz; at 5.1 ms, 6 components: names 65.4 Hz, the ghost bass; at 6.1 ms, 6 components: names 65.4 Hz, the ghost bass; at 7.6 ms, 6 components: names 65.4 Hz, the ghost bass; at 12.2 ms, 6 components: names 65.4 Hz, the ghost bass; at 15.3 ms, 6 components: names 65.4 Hz, the ghost bass; at 20.4 ms, 6 components: names 65.4 Hz, the ghost bass; at 61.2 ms, 6 components: names 65.4 Hz, the ghost bass. The fit names the 65.4 Hz ghost from 5.1 ms; the bass's own 4 cycles take 61.2 ms. Intervals and chords

The notes name the ghost bass before it has a cycle

A ghost bass that had to wait for four of its own cycles needed 61 milliseconds under a third from C4, and so vanished from fast passages. A pitch mechanism fitting the whole object — both notes' partials, which run their cycles many times faster, and the products — names the same bass in five. Seven of the eight thirds of a just scale have their bass for most of every note at any tempo a player reaches. The eighth, the scale's Pythagorean D–F, never has its own: the object hears it as a 6 : 5 and supplies the bass a D a comma lower would give, a semitone above the difference tone, until the difference tone arrives and breaks the fit.

 A melody finds the barline the harmony cannot, and only by preferring the downbeat for its long notes. At two chords a bar over a walking bass, eight bars, how often the length of the melody's notes breaks the metre's tie and places the barline, against the share of long notes that fall on the downbeat rather than the half-bar, for melodies with a long note in a quarter, a half and all of their bars. 25% of bars long: 46% at 50%, 53% at 60%, 61% at 70%, 71% at 80%, 78% at 90%, 85% at 100%; 50% of bars long: 48% at 50%, 61% at 60%, 76% at 70%, 91% at 80%, 95% at 90%, 97% at 100%; 100% of bars long: 47% at 50%, 74% at 60%, 85% at 70%, 97% at 80%, 100% at 90%, 100% at 100%. The bass's readings, whether it moves and whether it sounds the root, average 50% over the same passages. At even odds the melody is no better than the metre alone, whatever the number of long notes. Harmony and voice leading

A melody marks the barline by where it lingers

At two chords a bar over a walking bass, every harmonic witness stands at the metre's coin toss about where the bar begins. Give the passage a melody with a habit — long notes that prefer the downbeat — and the barline comes back: with a long note in half the bars and four in five of them on the downbeat, in 91 per cent of passages. The count of long notes barely matters; the preference is everything, and at even odds no number of long notes helps. A phrase end held through its bar is the strongest single cue. And a listener who adds the bass's readings to the melody's does worse than one who ignores the bass.

 Quietly, or quickly, the scale's Pythagorean third is heard with the bass of the just third a comma away. The third D4–F4 of a just C major scale, a Pythagorean 32 : 27, fitted as a whole object with every component that has run 4 of its cycles and every product only if it is audible at the level played: what the fit names at the end of a note of each length, at each dynamic. 40 dB: 20 ms 58.4 Hz, 40 ms 58.4 Hz, 60 ms 58.4 Hz, 80 ms 58.4 Hz, 100 ms 58.4 Hz, 150 ms 58.4 Hz, 200 ms 58.4 Hz, 300 ms 58.4 Hz, 500 ms 58.4 Hz; 50 dB: 20 ms 58.4 Hz, 40 ms 58.4 Hz, 60 ms 58.4 Hz, 80 ms 58.4 Hz, 100 ms 58.4 Hz, 150 ms 58.4 Hz, 200 ms 58.4 Hz, 300 ms 58.4 Hz, 500 ms 58.4 Hz; 60 dB: 20 ms 58.4 Hz, 40 ms 58.4 Hz, 60 ms 58.4 Hz, 80 ms 58.4 Hz, 100 ms 58.4 Hz, 150 ms 58.4 Hz, 200 ms 58.4 Hz, 300 ms 58.4 Hz, 500 ms 58.4 Hz; 66 dB: 20 ms 58.4 Hz, 40 ms 58.4 Hz, 60 ms 58.4 Hz, 80 ms 58.4 Hz, 100 ms 58.4 Hz, 150 ms 58.4 Hz, 200 ms 58.4 Hz, 300 ms 58.4 Hz, 500 ms 58.4 Hz; 70 dB: 20 ms 58.4 Hz, 40 ms 58.4 Hz, 60 ms 58.4 Hz, 80 ms 58.4 Hz, 100 ms 58.4 Hz, 150 ms 58.4 Hz, 200 ms 58.4 Hz, 300 ms 58.4 Hz, 500 ms 58.4 Hz; 74 dB: 20 ms 58.4 Hz, 40 ms 58.4 Hz, 60 ms 58.4 Hz, 80 ms 58.4 Hz, 100 ms 58.4 Hz, 150 ms 58.4 Hz, 200 ms 58.4 Hz, 300 ms 58.4 Hz, 500 ms 58.4 Hz; 76 dB: 20 ms 58.4 Hz, 40 ms 58.4 Hz, 60 ms 58.4 Hz, 80 ms 27.1 Hz, 100 ms 27.1 Hz, 150 ms 27.1 Hz, 200 ms 27.1 Hz, 300 ms 27.1 Hz, 500 ms 27.1 Hz; 80 dB: 20 ms 58.4 Hz, 40 ms 58.4 Hz, 60 ms 58.4 Hz, 80 ms 27.1 Hz, 100 ms 27.1 Hz, 150 ms 27.1 Hz, 200 ms 27.1 Hz, 300 ms 27.1 Hz, 500 ms 27.1 Hz; 85 dB: 20 ms 58.4 Hz, 40 ms 58.4 Hz, 60 ms 58.4 Hz, 80 ms 27.1 Hz, 100 ms 27.1 Hz, 150 ms 27.1 Hz, 200 ms 27.1 Hz, 300 ms 27.1 Hz, 500 ms 27.1 Hz; 90 dB: 20 ms 58.4 Hz, 40 ms 58.4 Hz, 60 ms 58.4 Hz, 80 ms 27.1 Hz, 100 ms 27.1 Hz, 150 ms 27.1 Hz, 200 ms 27.1 Hz, 300 ms 27.1 Hz, 500 ms 27.1 Hz. 58.1 Hz is the bass of D–F with D a syntonic comma lower, a just 6 : 5; the Pythagorean third's own difference tone is 54.5 Hz and is audible from 76 dB, and once it has run its cycles the fit breaks. Intervals and chords

The dynamic chooses which D the scale has

A just C major scale has one third that is not 5 : 4 or 6 : 5 — D–F, a Pythagorean 32 : 27 — and a pitch mechanism fitting the notes' partials hears it as the just minor third a comma away, with that third's bass, B♭1, beneath it. Only the Pythagorean third's own difference tone can contradict that, and it is audible only from 76 dB. So below a forte the ghost line under a scale in just thirds is C, B♭, C, F, G, F, G, C for the whole of every note, and the syntonic comma is resolved by the listener rather than the players; above it, and only in notes longer than 73 milliseconds, the B♭ is taken away.

 No setting of the two costs reads both the bridge's chain and the rondo's episodes. The six schemes read by a key-finder with a second slot for an applied dominant, at a bass weight of 3 and a resolution cue of 3, over every key cost and every cost of entering a tonicization: in each cell, the bridge's bars read as the scheme labels them (of 8), the rondo's (of 40) and the whole listing's (of 188). key cost 2: 0.5 4/24/163, 1 4/24/156, 2 1/24/153, 3 1/28/154, 4 1/34/160, 5 1/34/160, 6 1/34/160, 8 1/34/160, 12 1/34/160; key cost 3: 0.5 8/24/172, 1 8/24/172, 2 4/24/163, 3 1/24/160, 4 1/34/170, 5 1/34/170, 6 1/34/170, 8 1/34/170, 12 1/34/170; key cost 4: 0.5 8/24/172, 1 8/24/172, 2 4/24/168, 3 3/24/162, 4 1/24/160, 5 1/35/171, 6 1/35/171, 8 1/35/171, 12 1/35/171; key cost 6: 0.5 8/24/172, 1 8/24/172, 2 8/24/172, 3 8/24/172, 4 3/24/167, 5 2/24/166, 6 2/30/172, 8 2/35/177, 12 2/35/177; key cost 8: 0.5 8/24/172, 1 8/24/172, 2 8/24/172, 3 8/24/172, 4 3/24/167, 5 2/24/166, 6 2/24/166, 8 2/30/172, 12 2/34/176; key cost 12: 0.5 8/24/172, 1 8/24/172, 2 8/24/172, 3 8/24/172, 4 7/24/171, 5 5/24/169, 6 5/24/169, 8 2/24/166, 12 2/24/166. Every cell that reads the whole bridge reads the rondo at 24 of 40, its two eight-bar episodes heard as tonicizations of V and vi inside C; every cell that reads the episodes as modulations reads at most two bridge bars. The one-slot reading at a key cost of 6 reads 177. Scales and modes

A second slot hears every modulation as a tonicization

A key-finder that names one key per bar reads a chain of applied dominants by moving the key with every chord. Give its state a second slot — the key a bar is in, and the key its dominant points at — and the thirty-two-bar song's bridge is read exactly as the scheme labels it, every bar, for the first time in these essays, and the blues's opening stays in its tonic. At every setting that does that, the rondo's two eight-bar episodes are heard as tonicizations inside C, and no setting of the three costs reads both. A modulation to a closely related key and a long tonicization are the same chords; what separates them is how long they last, and a cost for entering a state cannot count.

 A syncopated melody is heard against the bar only if its phrases land on it. At one chord a bar over a walking bass, a melody whose long notes all fall on the half-bar, with a long note in half its bars and in all of them: how often the melody and bass together place the barline (solid) and the melody alone (dashed), against the share of two-bar phrase ends held through their last bar from its downbeat. a long note in 50% of bars: 0% held, together 21%, melody 0%; 10% held, together 36%, melody 20%; 25% held, together 67%, melody 58%; 50% held, together 86%, melody 84%; 75% held, together 97%, melody 97%; 100% held, together 100%, melody 100%. a long note in 100% of bars: 0% held, together 21%, melody 0%; 10% held, together 23%, melody 4%; 25% held, together 45%, melody 29%; 50% held, together 77%, melody 70%; 75% held, together 96%, melody 95%; 100% held, together 100%, melody 100%. Harmony and voice leading

A syncopated melody needs its phrases to land

Put a melody whose long notes all fall on the half-bar over a harmony that changes once a bar and a walking bass, and a listener adding up the evidence hears the bar begin in the wrong place three times in four. The harmony knows where the bar is, but only through one witness: the bass's root votes for the barline in every passage, and the bass's motion never does — a walking bass lands on each new root by step, so it moves less at the barline than inside the bar. One syncopated voice ties the root, and the motion breaks the tie its way. What rescues the barline is not the harmony but the melody's own phrase ends: hold half of them from the downbeat and the reading finds the bar in 86 per cent of passages.

 Made to stay 2 chords, the parts change over in groups rather than one by one. Six parts entering, leaving and re-entering a five-chord passage over four sounding players, in the walk through all 64 sets of sounding parts that makes the least audible entrance as audible as possible, with no memory of the chord before, at least 2 of the six sounding at the last chord, and every part that enters staying at least 2 chords before it may leave. I, open: clarinet on G4 enters at 0.6 dB; vi, close below: nobody enters; IV, close above: brass on E3 enters at -0.7 dB, voice on C5 enters at 2.2 dB, and clarinet leaves; V, bracketing: violin on G5 enters at 7.2 dB; I, hollow: oboe on E4 enters at -1.1 dB, flue pipe on C6 enters at 4.7 dB, and brass and voice leave. The least audible entrance is -1.07 dB against -2.65 for the best schedule in which nobody leaves; the walk has 3 exits and 6 entrances. Form and structure

A part that must stay changes over in shifts

The best way to keep every entrance audible, when the passage need not end with everybody in, was a relay: a part enters, is heard for one chord, and gives way. Make every part stay at least two chords once it is in and the relay does not collapse into overlaps, as predicted. It changes over in shifts: two parts enter together, one chord passes in which nobody enters, and both leave on the chord their successors arrive. Half of what exits were worth survives a two-chord stay and nearly two fifths a three-chord one — and with the chord before still masking, the stay costs less, not more, because it was the one-chord handover that memory had already made leaky.

 A moving tenor on a sustained instrument can hand over only the partials the chord leaves it. A four-voice arrival balanced for loudness, played on a sustained instrument, with the tenor moving a second after the chord: how many of the tenor's partials stand above the masking of the other three voices, on average over four arrivals, against how far the chord is transposed, for a string, an organ and a clarinet spectrum. The held tenor keeps exactly the same partials. string: 24 semitones down 0.0, 21 semitones down 1.0, 18 semitones down 1.0, 15 semitones down 1.0, 12 semitones down 1.0, 9 semitones down 2.5, 6 semitones down 1.8, 3 semitones down 2.3, as written 3.5, 3 semitones up 4.3, 6 semitones up 2.5, 9 semitones up 3.8, 12 semitones up 3.8; organ: 24 semitones down 0.0, 21 semitones down 1.0, 18 semitones down 1.0, 15 semitones down 1.8, 12 semitones down 2.3, 9 semitones down 2.3, 6 semitones down 1.0, 3 semitones down 2.8, as written 3.0, 3 semitones up 3.3, 6 semitones up 1.3, 9 semitones up 3.0, 12 semitones up 3.0; clarinet: 24 semitones down 0.5, 21 semitones down 2.5, 18 semitones down 1.0, 15 semitones down 2.5, 12 semitones down 3.5, 9 semitones down 3.0, 6 semitones down 3.3, 3 semitones down 3.3, as written 3.8, 3 semitones up 3.3, 6 semitones up 3.5, 9 semitones up 3.5, 12 semitones up 3.5. Whatever the onset does, it can only do it to these. Perception and the listener

An onset can only hand over what the chord leaves

On an organ, a tenor that moves while the chord holds keeps exactly the partials a held tenor keeps, so the masking count cannot tell them apart. What can is the onset: a component that starts thirty milliseconds after the sound around it is heard out as something new. A moving tenor starts hundreds of milliseconds after the chord and gets that for every partial it has left — and what it has left, an octave down, is one partial on a string and three on an organ, which still name the tenor's own pitch. Two octaves down there is nothing to hand over. And a tenor that moves at the chord change, with the other voices, gets nothing, which is why the motion that carries an inner voice is the motion between the harmony's changes.

 Charged by the bar, the second slot gives up the bridge before it gives back the rondo. The share of the song's bridge (of 8) and of the rondo (of 40) read as the schemes label them, against the price of every bar spent in an applied-dominant state, with a tonicization costing 0.5 to enter, a resolution cue of 3 and a bass weight of 3, at key costs of 3 and 6. key cost 3: 0: bridge 8, rondo 24; 0.25: bridge 8, rondo 24; 0.5: bridge 8, rondo 24; 0.75: bridge 4, rondo 24; 1: bridge 4, rondo 24; 1.5: bridge 4, rondo 24; 2: bridge 4, rondo 24; 2.5: bridge 1, rondo 24; 3: bridge 1, rondo 29; 4: bridge 1, rondo 34; 5: bridge 1, rondo 34; 6: bridge 1, rondo 34. key cost 6: 0: bridge 8, rondo 24; 0.25: bridge 8, rondo 24; 0.5: bridge 8, rondo 24; 0.75: bridge 8, rondo 24; 1: bridge 8, rondo 24; 1.5: bridge 8, rondo 24; 2: bridge 8, rondo 24; 2.5: bridge 6, rondo 24; 3: bridge 3, rondo 24; 4: bridge 3, rondo 24; 5: bridge 2, rondo 24; 6: bridge 2, rondo 35. At no price does one reading hold both. Scales and modes

A modulation is a dominant that keeps coming back

A key-finder with a slot for an applied dominant read the thirty-two-bar song's bridge as its scheme labels it and heard the rondo's two episodes as tonicizations inside C, and no setting read both. The obvious repair — charge a tonicization by the bar, so a long one costs more than a short one — makes it worse: the reading gives up the bridge before it gives back a single episode, because the bridge spends six bars in applied chords and the episodes three and two. By the bar, the chain is the longest tonicization in the listing. What the episodes have and the bridge does not is a dominant that resolves to the same tonic again and again.

 Scored as one arrival, a pair enters as a loud part carrying one that cannot be heard. Six parts entering, leaving and re-entering a five-chord passage over four sounding players, in the walk through all 64 sets of sounding parts that makes the least audible entrance as audible as possible, with no memory of the chord before, at least 2 of the six sounding at the last chord. I, open: clarinet on G4 enters at -10.9 dB, violin on G5 enters at 19.6 dB; vi, close below: oboe on E4 enters at -1.5 dB, voice on C5 enters at 14.5 dB, and clarinet and violin leave; IV, close above: nobody enters, and oboe and voice leave; V, bracketing: brass on E3 enters at -0.0 dB, flue pipe on C6 enters at 25.0 dB; I, hollow: nobody enters. Scored with the parts entering on a chord taken together, at most 2 at a time, as one arrival against what was already sounding: the weakest arrival is 10.93 dB, and the most buried single entrant, measured as itself, is -10.94 dB. Form and structure

A pair entering as one hides the part that cannot be heard

Scored as two entrances, a pair of parts coming in together is charged for masking itself; scored as one arrival — a new colour against what was already sounding — it is not, and the best schedules change completely. Allowed two at a time, the weakest arrival rises from 0.6 dB above the mask to 10.9, and every pair the search chooses is a loud part carrying a quiet one: a violin 19.6 dB clear beside a clarinet 10.9 under. Allowed any number, the six parts enter at once. The staggered build-up of every earlier schedule exists only because each part was meant to be heard as itself.

 The three orders of the bell's gaps with the pulse in fours heard 50 per cent of the time. Bits of position unknown over the first cycle, averaged over alignments, for each order of the gaps 1 1 2 2 2 2 2 against a pulse in threes heard in full and a pulse in fours of which 50 per cent of strokes are heard, by a listener who knows the share; memory half-life 3 steps. 2 2 2 1 2 1 2: 0.613; 2 2 2 2 1 1 2: 0.593; 2 2 1 2 2 1 2 (the standard bell pattern): 0.513. The bell pattern leads by 0.080 bits. The expectation over which strokes are lost is exact; drawn instead from 150 random patterns of loss it comes to 0.509 against 0.510 at one alignment. Rhythm and metre

A feel heard half the time gives the bell half its lead

Against one feel the standard bell pattern is the quickest order of its gaps to place; against both feels at once its lead is gone. An ensemble's texture is neither: each feel is heard some of the time. Given a listener who knows how often a pulse's strokes are lost, the lead comes back in proportion to the loss — half the strokes missing restores half of it, at every memory from a step and a half to a whole cycle. There is no threshold at either end, and a listener who treats every silence as proof comes out more certain and worse placed.

 With a drum in the introduction, 2 bars of ground are enough — for a listener who remembers. At one chord a bar over a walking bass, a melody whose long notes all fall on the half-bar (50 per cent of its bars hold one), preceded by a stated number of bars of the harmony and bass alone with a drum on every downbeat of the introduction: the share of the melody's eight bars that a listener reading bar by bar hears against the written barline, for five listeners who differ in how much confidence they can bank. no memory: 0 bars 47%, 1 bars 46%, 2 bars 46%, 3 bars 46%, 4 bars 45%, 6 bars 46%, 8 bars 45%, 12 bars 45%, 16 bars 46%. banks 2 votes: 0 bars 45%, 1 bars 57%, 2 bars 58%, 3 bars 57%, 4 bars 56%, 6 bars 56%, 8 bars 55%, 12 bars 56%, 16 bars 58%. banks 4 votes: 0 bars 44%, 1 bars 68%, 2 bars 79%, 3 bars 79%, 4 bars 76%, 6 bars 77%, 8 bars 77%, 12 bars 76%, 16 bars 79%. banks 8 votes: 0 bars 44%, 1 bars 71%, 2 bars 90%, 3 bars 95%, 4 bars 95%, 6 bars 95%, 8 bars 96%, 12 bars 95%, 16 bars 97%. never forgets: 0 bars 44%, 1 bars 71%, 2 bars 90%, 3 bars 96%, 4 bars 99%, 6 bars 100%, 8 bars 100%, 12 bars 100%, 16 bars 100%. A listener with no memory stays near 47 per cent however much ground is heard; one who never forgets goes from 44 with none to 90 after two bars and 100 after twelve. Harmony and voice leading

Two bars of drum hold the barline; a dozen bars of bass do not

A listener who settles where the bar begins and then holds it hears a syncopated tune against the bar only if the ground before it built enough confidence to spend. Read bar by bar, a walking bass under one chord a bar is worth less than half a vote a bar for the barline, and a syncopated bar takes an eighth of a vote back, so a tune needs about twelve bars of that ground to be heard against the bar nine times in ten. A drum on the downbeat makes each bar of ground worth a vote and a half, and two bars are then enough — but only for a listener who can bank more than four votes of confidence.

 A slow attack is heard where a much quieter note would be. How far a listener whose ears differ by 2 dB hears a 98 Hz note displaced toward the better ear, in detectable steps of 15 µs, against its dynamic, if 14 per cent of the brainstem latency law's difference reaches the delay comparator and the note's position is set by its onset, weighted over 3 ms from the moment it becomes audible. The dashed line is the note treated as a steady tone. struck (2 ms rise): 40 dB 5.73, 50 dB 5.00, 60 dB 4.32, 70 dB 3.70, 80 dB 3.16, 90 dB 2.69. plucked (5 ms rise): 40 dB 5.95, 50 dB 5.32, 60 dB 4.66, 70 dB 4.02, 80 dB 3.45, 90 dB 2.94. tongued (25 ms rise): 40 dB 6.46, 50 dB 6.07, 60 dB 5.52, 70 dB 4.87, 80 dB 4.23, 90 dB 3.63. bowed (100 ms rise): 40 dB 6.69, 50 dB 6.51, 60 dB 6.16, 70 dB 5.64, 80 dB 5.01, 90 dB 4.36. a soft bowed entrance (250 ms rise): 40 dB 6.76, 50 dB 6.67, 60 dB 6.46, 70 dB 6.07, 80 dB 5.52, 90 dB 4.87. Steady: 5.52, 4.70, 4.00, 3.40, 2.90, 2.47. At 90 dB a 250 ms entrance is displaced 1.98 times as far as the steady tone. Perception and the listener

A bowed note is placed as if it were quieter

An ear a decibel or two worse than the other is later to report every sound, by an amount that grows as the sound gets quieter, and a listener takes a note's direction from its onset. A plucked or struck note reaches its dynamic before the onset is over and is placed almost where a steady tone would be. A bowed note is still near the threshold of hearing while its onset is being read, so it is placed as though it were 27 decibels quieter than it is. The prediction was that slow attacks would soften the displacement; they sharpen it, and an asymmetric listener's orchestra is deformed by its articulation as much as by its register.

 At the same share of time lost, a crowd that hides stretches costs less than one that hides onsets. The share of metres recognised within the 16-step present from a random entry, 10 metres known, with a stated share of the time lost to noise, against the mean length of each stretch lost, in steps — "one by one" is every step lost independently, as in the essay on lost onsets. The listener asks for odds of 19 to one and allows for losses at the true rate. The dashed lines are the same listeners with nothing lost. landmarks, downbeat marked, 10 per cent lost: one by one 46%, 2 46%, 4 46%, 8 47%, 16 48%, 32 49%, 64 50%; nothing lost 56%. landmarks, downbeat marked, 25 per cent lost: one by one 27%, 2 27%, 4 28%, 8 30%, 16 32%, 32 34%, 64 35%; nothing lost 48%. every onset, downbeat accented, 10 per cent lost: one by one 61%, 2 60%, 4 61%, 8 62%, 16 64%, 32 65%, 64 66%; nothing lost 74%. every onset, downbeat accented, 25 per cent lost: one by one 40%, 2 39%, 4 41%, 8 44%, 16 48%, 32 51%, 64 53%; nothing lost 74%. Rhythm and metre

A crowd that hides bars is kinder than one that hides onsets

A dancer joining an additive-metre dance in a noisy room loses some of what is played, and the essay before this one lost it one onset at a time. A crowd does not. It hides stretches — a burst of laughter covers a bar. The prediction was that a stretch as long as a bar would hurt the stream of landmarks most, since it could remove every landmark in a cycle, and that long bars would survive best. Holding the share of time lost fixed, longer stretches help every stream, help the full stream more than the landmarks, and leave long bars as hard to place as ever.

 Three ways of remembering where the dominants went, and none reads both the bridge and the episodes. At a key cost of 6, a resolution cue of 3 and a tonicization cost of 0.5: the song's bridge (of 8 bars), the rondo (of 40) and the whole listing (of 188) read as the schemes label them, for a reading that charges an applied dominant resolving again to the target the last one reached — with that memory lasting the whole home key, or four bars — and for one that instead pays a bonus whenever a key's own dominant resolves to its own tonic. a return charged, remembered through the key: 0: bridge 8, rondo 24, all 172; 2: bridge 8, rondo 24, all 136; 4: bridge 8, rondo 24, all 155; 8: bridge 8, rondo 16, all 147; 12: bridge 8, rondo 16, all 147. a return charged, remembered four bars: 0: bridge 8, rondo 24, all 172; 2: bridge 8, rondo 24, all 172; 4: bridge 8, rondo 24, all 172; 8: bridge 8, rondo 24, all 172; 12: bridge 8, rondo 24, all 172. a confirmation paid: 0: bridge 8, rondo 24, all 172; 2: bridge 8, rondo 24, all 149; 4: bridge 8, rondo 24, all 116; 6: bridge 2, rondo 29, all 115; 8: bridge 2, rondo 28, all 104. The rondo's episodes are recovered only once, at a confirmation of 6, and by then the bridge has fallen to 2. Scales and modes

A return is cheaper to avoid than to pay for

The essay before this one found what separates a modulation from a tonicization: an episode's dominant keeps resolving to the same target, and the bridge's dominants never do. So the key-finder was given a memory of where its last dominant went, and charged for a return — the prediction being that one setting would at last hear the episodes as keys and the bridge as a chain. None does. Every return the charge is aimed at is avoided for less than a nat, by hearing a D7 as a minor chord or by moving the passage into another key, while the reading the charge was meant to force costs twelve.

 Every decibel taken off the passenger comes off the arrival. The best walk with parts entering together scored as one arrival, at most two at a time, against a floor under each member: no entrant, heard as itself, more than the floor below its mask. Solid lines are the weakest arrival, dashed the most buried member, for two parts required at the end and for all six. 2 at the end: floor 0: arrival 0.59, most buried 0.59; floor 1: arrival 0.59, most buried -0.86; floor 2: arrival 0.59, most buried -1.57; floor 3: arrival 0.59, most buried -2.87; floor 4: arrival 0.59, most buried -3.47; floor 5: arrival 0.59, most buried -3.47; floor 6: arrival 4.50, most buried -5.42; floor 8: arrival 4.50, most buried -6.16; floor 10: arrival 10.93, most buried -9.90; floor 12: arrival 10.93, most buried -10.94; every part heard as itself, the earlier walks' score, 0.59. 6 at the end: floor 0: no walk; floor 1: no walk; floor 2: no walk; floor 3: arrival -2.65, most buried -2.65; floor 4: arrival 0.50, most buried -3.27; floor 5: arrival 0.59, most buried -4.24; floor 6: arrival 4.50, most buried -5.42; floor 8: arrival 4.50, most buried -6.13; floor 10: arrival 4.50, most buried -6.13; floor 12: arrival 4.50, most buried -6.13; every part heard as itself, the earlier walks' score, -2.65. Form and structure

A floor under each member is paid for in steps

Score parts entering together as one arrival and the best pairs are a loud part carrying one that cannot be heard. Put a floor under each member — no part more than so many decibels under its own mask — and the prediction was a schedule between the two, neither staggered nor buried. There is none. The arrival's margin climbs in three steps as the floor is loosened, each step bought with a passenger buried about as deep as the step is high, and below a floor of five decibels nothing beats bringing the parts in one at a time, because the clarinet can clear its mask only alone.

 The tempered seventh chord repeats where the tuned one does, only less exactly. The autocorrelation of the sound reaching the ear, for a dominant seventh over C3 at its longest-lived spacing — C3, then E4, G4 and B♭4 — tuned 2 : 5 : 6 : 7 and tempered, struck at 80 dB, partials above 5 kHz removed, across every lag that corresponds to a pitch above 30 Hz. The tuned chord's highest peak is 1.000 at 15.29 ms, 65.41 Hz, C2, the series' fundamental. The tempered chord's highest is 0.824 at 15.21 ms, 65.75 Hz, C2 +9¢. At the bass note's own period, 7.64 ms, the two reach only 0.24 and 0.25: neither chord repeats strongly at the pitch of its lowest note. Intervals and chords

A tempered seventh chord already has the tuned one's low note

A dominant seventh tuned 4 : 5 : 6 : 7 is one periodic sound, and its period is a note below the chord that nobody plays. The question left open was whether tuning adds that note. Run through the temporal account of residue pitch, the tempered chord turns out to repeat at the same place, a few cents sharp, at 0.82 of the tuned chord's exactness at the strike and 0.96 a second later. Tuning does not create the low note. It makes it exact, and it gets there by a route on which every step helps, which is not true of the beating.

 A listener who weights its witnesses by their record needs 8 bars of walking bass, not 12. A syncopated melody over one chord a bar and a walking bass, preceded by a stated number of bars of the harmony and bass alone: the share of the melody's eight bars heard against the written barline by a listener who never forgets, counting every witness's vote alike, and by one who weights each witness by how often it has agreed with the reading it held. Solid: a long note in half the melody's bars; dashed: in every bar. Half the bars, votes alike: 0 bars 40%, 1 bar 49%, 2 bars 57%, 3 bars 63%, 4 bars 71%, 6 bars 77%, 8 bars 83%, 12 bars 93%, 16 bars 97%. Weighted: 0 39%, 1 59%, 2 71%, 3 76%, 4 83%, 6 88%, 8 92%, 12 95%, 16 94%. Every bar, votes alike: 0 6%, 1 13%, 2 19%, 3 27%, 4 34%, 6 48%, 8 56%, 12 72%, 16 85%. Weighted: 0 6%, 1 38%, 2 54%, 3 66%, 4 74%, 6 85%, 8 90%, 12 93%, 16 98%. Nine bars in ten are reached after 12 bars of ground counting alike and 8 weighting; for the tune that syncopates every bar, after more than sixteen and 8. Reading a reliably wrong witness backwards instead of ignoring it changes no value by more than two points. Harmony and voice leading

A listener who weights its witnesses needs eight bars of bass, not twelve

A syncopated tune over a walking bass needed about twelve bars of ground before a listener with a memory heard it against the bar nine times in ten, with every witness given one vote a bar. Let the listener weight each witness by how often it has agreed with the reading it holds and the root earns a second vote, the bass's motion loses half of one, and the syncopation loses almost all of its. Eight bars are then enough, a tune that syncopates every bar becomes hearable against the bar at all, and reading the syncopation backwards buys nothing that ignoring it does not.

 In the bass a bowed step keeps the entrance's displacement; in the treble every note is placed afresh. A listener whose ears differ by 2 dB, with 14 per cent of the brainstem latency difference reaching the delay comparator: how far the second note of a bowed legato pair at 60 dB is displaced, in detectable steps of 15 µs, against the interval it moves by, in three registers. A note with no onset of its own keeps the first note's position, drawn at the entrance's displacement; the dotted lines are each register's entrance from silence and the dashed its steady tone. G2, a double bass (entrance 6.16, steady 4.00): -12 6.19, -7 5.02, -4 4.35, -2 4.06, -1 no onset, +1 no onset, +2 no onset, +4 4.08, +7 4.26, +12 4.53. A4, a violin (entrance 3.73, steady 2.15): -12 3.72, -7 3.56, -4 2.82, -2 2.35, -1 2.19, +1 2.15, +2 2.22, +4 2.39, +7 2.67, +12 3.07. A6, a violin's top (entrance 2.13, steady 1.21): -12 2.13, -7 2.13, -4 1.77, -2 1.40, -1 1.26, +1 1.22, +2 1.29, +4 1.43, +7 1.66, +12 1.93. At the dial's up 2 semitones: G2 no onset, keeps 6.16; A4 2.22; A6 1.29. Perception and the listener

A bowed bass line is placed by its entrance; a violin line by every note

A listener whose ears differ slightly places a note by its onset, and a bowed entrance is placed as though it were far quieter than it is. A legato line's later notes do not start from silence. Whether one of them has an onset at all depends on how much of the note before is already in its own auditory filter — and in the bass, where the filters are wide and their lower skirts shallow, a step of a tone or less has none, so the note keeps the entrance's displacement. In the treble nearly every note has its own onset and is placed near its steady position, except where the line leaps down.

 An elaboration's most useful boundary is one the skeleton does not have. Bits of position unknown over the first cycle of a sixteen-step cycle with gong, kempul and kenong, timing from every layer, when an elaborating part adds a figure a whole cycle long whose one recognisable boundary falls at the step drawn, at 40, 80, 160 seconds a cycle. Dashed: the skeleton alone. Dotted: the skeleton with a kethuk every two steps. 40 s: skeleton 0.112, with a kethuk 0.104; boundary at 2 0.029, 4 0.004, 6 0.004, 8 0.006, 10 0.005, 12 0.004, 14 0.028. 80 s: skeleton 0.329, with a kethuk 0.301; boundary at 2 0.125, 4 0.094, 6 0.067, 8 0.087, 10 0.057, 12 0.072, 14 0.089. 160 s: skeleton 0.646, with a kethuk 0.457; boundary at 2 0.360, 4 0.340, 6 0.256, 8 0.323, 10 0.251, 12 0.295, 14 0.292. A boundary on the gong is the gong's own timer and adds nothing. Anywhere in the middle of the cycle it re-times the digit the gong reads worst, and does more than a whole layer of kethuk strokes. Rhythm and metre

An elaboration is a clock hand where the skeleton has none

A colotomic cycle is a clock whose hands are its punctuating layers, and the finest hand a slow cycle can use runs out as the cycle slows. The elaborating parts that fill the cycle were predicted to supply that finer hand. They can, and a two-step figure recognised at every boundary is a kethuk exactly. But the figure that does most is not a fine one. It is a long one with a single recognisable point in the middle of the cycle, where no skeleton stroke falls: at eighty seconds a cycle it leaves a listener a fifth as uncertain as the skeleton alone, where a kethuk takes off a tenth — and it still does more than a kethuk when a listener catches that point one time in ten.

 Read in its partials' filters, every step of a bowed bass line is an onset. The second note of a bowed legato pair on the double bass's G2 at 60 dB, for a listener whose ears differ by 2 dB: its displacement in detectable steps against the interval, placed by the filter at its fundamental alone (dashed, with pale dots where that filter sees no onset and the note keeps the entrance's position) and by the first of its eight partials' filters to see an onset (solid). By partials: -12 6.19 in partial 1, -7 5.02 in partial 1, -4 3.68 in partial 2, -2 3.12 in partial 3, -1 2.76 in partial 4, +1 2.61 in partial 5, +2 3.10 in partial 3, +4 3.56 in partial 2, +7 3.15 in partial 3, +12 3.51 in partial 2. The entrance, placed the same way, is 4.49 steps by its partial 4, against 6.16 by its fundamental. At the dial's +2: 3.10 steps, first in partial 3. Perception and the listener

A bowed bass line has an onset at every step, in its upper partials

A legato note's onset was read in the auditory filter at its fundamental, and in the bass a step of a tone or less left that filter nearly as full as before: no onset, and the note kept its entrance's displacement. A bowed note is not a fundamental. Read in every filter its partials reach, every step of a bowed bass line has an onset, first seen in the third to fifth partial, where a tone moves a partial right out of its filter. The line is re-placed at every note after all — and nearer the centre, because a partial's filter is at a frequency where the ears' latency difference is smaller. The bowed bass entrance itself is displaced a quarter less than its fundamental said.

 A listener who arrives knowing the style needs no introduction. A syncopated melody over one chord a bar and a walking bass, after a stated number of bars of the harmony and bass alone: the share of the melody's eight bars heard against the barline by a listener who never forgets and counts every vote alike, by one who learns each witness's reliability from nothing, and by one who arrives already crediting the bass's motion and the melody's long notes at a coin toss — the habits of a style in which the bass walks and the tune syncopates — and learns from there. every vote alike: 0 bars 40%, 1 bar 49%, 2 bars 57%, 3 bars 63%, 4 bars 71%, 6 bars 77%, 8 bars 83%, 12 bars 93%, 16 bars 97%. learning from no habits: 0 bars 39%, 1 bar 59%, 2 bars 71%, 3 bars 76%, 4 bars 83%, 6 bars 88%, 8 bars 92%, 12 bars 95%, 16 bars 94%. arriving with the style's habits: 0 bars 93%, 1 bar 97%, 2 bars 99%, 3 bars 99%, 4 bars 100%, 6 bars 100%, 8 bars 100%, 12 bars 100%, 16 bars 100%. Harmony and voice leading

A listener who knows the style needs no introduction

A listener who weights each witness to the barline by its record needed eight bars of walking bass before a syncopated tune was heard against the bar nine times in ten, and with no introduction at all it learned the syncopation's lesson instead. Finer records — one for each side a witness takes, one for each section — change nothing. What changes everything is arriving with the style's habits: a listener who already gives the bass's motion and the melody's long notes no vote hears the tune against the bar 93 per cent of the time from its first bar, and 94 per cent for a tune that syncopates every bar. The root need not be trusted more. The others need to be trusted less.

 A second recognisable point is worth a great deal if it can be told from the first. Bits of position unknown over the first cycle of a sixteen-step cycle with gong, kempul and kenong, when the elaborating parts add none, one, two or three recognisable points, each set at its best placement, read by a listener who can tell the points apart and by one who cannot, at 80 and 160 seconds a cycle. 80 s: none 0.329; 1 told apart 0.057 at 10, alike 0.057 at 10; 2 told apart 0.001 at 6/13, alike 0.005 at 6/12; 3 told apart 0.000 at 6/10/14, alike 0.007 at 4/9/14. 160 s: none 0.646; 1 told apart 0.242 at 11, alike 0.242 at 11; 2 told apart 0.094 at 6/13, alike 0.180 at 7/13; 3 told apart 0.038 at 5/11/15, alike 0.117 at 7/10/13. Points told apart add like checkpoints, each taking away most of what is left. Points alike add a little, and at eighty seconds a third alike point leaves the listener worse placed than two. Rhythm and metre

Two landmarks in a cycle help only if they can be told apart

One recognisable point inside a slow gong cycle, placed where the skeleton has no stroke, did more for a listener than a whole layer of kethuk strokes. A second point does as much again, a third nearly as much again — provided the listener can tell which point they have just heard. If the points sound alike, a second one is worth a fraction of that, seventy of the hundred and five ways of placing it leave the listener worse off than one point would, and two alike points half a cycle apart undo most of what the first one gave. Whether the listener keeps one timer or several turns out not to matter at all.

 A first hearing costs the through-composed piece most and the rondo almost nothing. How many proportions between 1 : 1 and 3 : 1 a listener tells apart across a whole 480-second piece of eight one-minute sections, counted at every level, with a tenth of lapses taking every count at once and lasting 20 seconds on average, for a listener who knows the piece, one who knows only its plan of sections, one who knows the style's five forms, and one who knows nothing of the style. through-composed, ABCDEFGH: knows the piece 75.7, knows only the plan 62.7, knows the style 61.4, knows nothing 59.3; rondo, ABACABAD: knows the piece 56.2, knows only the plan 56.0, knows the style 54.6, knows nothing 51.6; sonata, ABABCDAB: knows the piece 61.0, knows only the plan 59.7, knows the style 57.8, knows nothing 53.9; song form, AABAAABA: knows the piece 35.3, knows only the plan 35.3, knows the style 34.5, knows nothing 32.8; strophic, AAAAAAAA: knows the piece 25.9, knows only the plan 25.9, knows the style 25.5, knows nothing 25.0. What a first hearing lacks is not the plan but the ability to recognise a section never heard: new music says only that it is new. Form and structure

New music says only that it is new

A listener who knows a piece comes back from a lapse of attention and is placed at once by whatever they hear, and a through-composed piece, which never repeats anything, is the easiest of all forms to come back to. A listener hearing it for the first time is not so lucky. What they hear on returning is either something they have heard before, which names itself, or something they have not, which says only that it is new. For a rondo that costs almost nothing. For the through-composed piece it costs its whole lead: at lapses of half a minute or more, a first-time listener is placed no better in it than in a rondo — and being told the piece's plan in advance gives back almost none of the difference.

 Only the D a just ii wants gives the chord a bass that survives its voicing. The bass a pitch mechanism fitting the whole chord — every note's partials and every audible difference and cubic tone — names under the ii chord of C major, with its D tuned three ways, in root position and both inversions, at 50 and 80 dB. the scale's D, 9 : 8: D F A at 50 dB G ±0, D F A at 80 dB none, F A D at 50 dB B +24, F A D at 80 dB F −2, A D F at 50 dB A −30, A D F at 80 dB none. a just ii's D, 10 : 9: D F A at 50 dB B♭ −4, D F A at 80 dB B♭ −4, F A D at 50 dB B♭ −4, F A D at 80 dB B♭ −4, A D F at 50 dB F +13, A D F at 80 dB B♭ −4. equal temperament: D F A at 50 dB G +7, D F A at 80 dB D −1, F A D at 50 dB G +7, F A D at 80 dB D −1, A D F at 50 dB A −23, A D F at 80 dB D ±0. B♭ is the bass the 10 : 9 D implies, since D–F–A is then 10 : 12 : 15 over it. Intervals and chords

The scale's ii chord borrows the dominant's bass

Heard quietly, the pair D–F of a just C major scale brings the bass of the other D with it — B♭, the bass the 10 : 9 D implies — so the comma is resolved by the listener rather than by the players. Add the A that makes the pair a ii chord and that stops being true. The whole-object fit now hears the chord over G, as the top of a dominant ninth, and it does so because the template prefers simple harmonics to accurate ones; a different bass is heard in each inversion, and at a forte none at all. Only a ii built on the 10 : 9 D has one bass in every voicing and at every dynamic. Played quietly from a fixed scale, I–vi–ii–V–I is heard with the bass line C, F, G, G, C: the ii merged into the chord after it, and the progression home.

 A bell's tierce is best placed at the first instant of the strike, and only for an instant. The tierce's margin above its masked threshold, in decibels, through the first 400 milliseconds after the strike, for bells on G1, C2, E2, G2, C3, with every mode starting at full amplitude and built up in the listener's own auditory filters. G1: -1.3 at 0.5 ms, -3.9 at 400 ms, never above threshold; C2: 0.3 at 0.5 ms, -2.5 at 400 ms, above threshold from 0.5 to 0.7 ms; E2: 1.6 at 0.5 ms, -1.2 at 400 ms, above threshold from 0.5 to 4.3 ms; G2: 2.7 at 0.5 ms, -0.2 at 400 ms, above threshold from 0.5 to 104.2 ms; C3: 4.7 at 0.5 ms, 1.7 at 400 ms, above threshold from 0.5 to 400.0 ms. Every curve falls from its first moment: there is no peak inside the window. Perception and the listener

A bell's strike gives its tierce a millisecond

A low bell's tierce is masked from below by its prime and hum, so it is present at the strike or never — and the strike was the one moment the count treated as a steady state. Give the modes a rise as well as a fall and the prediction was that the tierce's margin would peak somewhere inside the first two hundred milliseconds. It peaks at the very first instant and falls from there, because the shadow a masker casts upward grows faster than the masker. On a C2 bell the tierce is above its threshold for less than a millisecond, and under the most generous rise anybody could propose for twenty-five: ten to thirty decibels too brief to be heard. What the onset visits for an instant, a softer stroke keeps for the whole note.

 A listener choosing between two styles hears the syncopated tune later than one who never chooses. A syncopated melody over one chord a bar and a walking bass, after a stated number of bars of the harmony and bass alone: the share of the melody's eight bars heard against the barline by a learning listener with no habits (every witness credited at three in four, which is also the habit of a style whose tunes mark the bar), by one arriving with the habits of a style whose tunes syncopate (the bass's motion and the melody's long notes credited at a coin toss), and by one who keeps both sets of habits and weighs them by how well each has predicted the votes heard so far. no habits: 0 bars 39%, 1 bar 59%, 2 bars 71%, 3 bars 76%, 4 bars 83%, 6 bars 88%, 8 bars 92%, 12 bars 95%, 16 bars 94%. the syncopating style's habits: 0 bars 93%, 1 bar 97%, 2 bars 99%, 3 bars 99%, 4 bars 100%, 6 bars 100%, 8 bars 100%, 12 bars 100%, 16 bars 100%. both, weighed by their record: 0 bars 67%, 1 bar 82%, 2 bars 91%, 3 bars 92%, 4 bars 95%, 6 bars 97%, 8 bars 98%, 12 bars 99%, 16 bars 98%. Harmony and voice leading

The habit worth bringing is to distrust the tune

A listener who arrives with a style's habits — the bass's motion and the melody's long notes credited at a coin toss — hears a syncopated tune against the bar from its first bar. But a listener does not know the style in advance. Give it two sets of habits, one for tunes that mark the bar and one for tunes that syncopate, and let it weigh them by how well each predicts what it hears: it does better than a listener with no habits and far worse than one who simply knows, and it becomes sure of the style in fewer than one passage in ten. The reason is that the evidence runs one way. A tune that marks the bar never testifies against the habit of distrusting tunes, so the habit that serves every style is the syncopating one, kept always.

 The bell pattern's lead is for arriving. The standard bell pattern's lead over the nearest other order of its own gaps, in bits of surprise a step, cycle by cycle, with the pulse in threes heard in full and the pulse in fours half the time. A listener who told how reliable the feel is: 0.043, 0.000, 0.000, 0.000, 0.000, 0.000. A listener who learns it from the strokes: 0.063, 0.001, 0.000, 0.000, 0.000, 0.000. A listener who takes every silence as proof: 0.144, 0.006, 0.005, 0.005, 0.004, 0.004. The lead is all in the first cycle, for every listener; after it the three orders keep a listener placed equally well. The listener who has to learn the reliability gets more from the bell than one who is told it and less than one who never learns. Rhythm and metre

The bell pattern's order is for arriving

Against a feel that comes and goes, the standard bell pattern is the quickest order of its gaps to place, and the lead grows the less a listener knows about the drummer. Followed past the first cycle, the lead is gone, whether or not the listener ever learns how often strokes are lost. And when both feels thin, a different order keeps a listener placed better than the bell does, because its two short gaps make one figure where the bell's make two that look alike.

 One rate in every filter, and whether the filters keep time. The envelopes of four auditory filters — centred at 93, 170, 385, 822 hertz — over the first quarter-second of the dominant seventh in close position over C2, tuned 4 : 5 : 6 : 7. On the left every partial of every note starts together, as in a chord struck at one instant: all four filters pulse once every 61.2 milliseconds and their pulses line up. On the right each note is given its own starting phase, as four voices or four organ pipes have: every filter still pulses at the same rate, but each at its own moment in the period, and the lowest and highest filters' envelopes correlate at -0.34 rather than 0.56. Timbre and acoustics

A tuned chord's flutter binds it only if its notes start together

A close seventh chord tuned 4 : 5 : 6 : 7 low on the keyboard fluctuates at one rate in every auditory filter, and one rate everywhere looked like the ear's evidence that a sound is one thing. It is only half of that evidence. The filters also have to rise and fall at the same moments, and whether they do depends on when each note starts — which tuning does not decide. Spread the notes' onsets by two milliseconds and the tuned chord's filters move together no more than a tempered chord's.

 How far under its mask a passenger is, and how much it still changes. Every pairing of two parts entering together on any chord of the passage in which the quieter part — the passenger — sits under its mask on average: 71 of them. Across, the passenger's mean margin, the measure every earlier walk used; up, how far its partials together change the arrival above the threshold of detection. Every one of the 71 is above zero: the lowest is 0.11 dB. 35 of the 41 buried more than 3 dB by the mean still change the arrival by more than 3 dB. The few near the zero line are the clarinet on the bracketing dominant, where every one of its partials is just under. Form and structure

A passenger under its mask is still heard in the arrival

The best way to bring parts in together was a loud part carrying a quiet one, and the quiet one was scored as buried — ten decibels under its mask, on average over its partials. An average is the wrong question. A masked threshold is the level at which a sound becomes a detectable change, and nearly every buried passenger has partials well clear of it. Scored as a change to the arrival rather than as a line of its own, every one of seventy-one passengers is heard, and the carriers' walk passes almost untouched.

 Which witness lies when the bass moves its root. Over a tune whose long notes mark the bar, the share of the melody's bars in which each witness votes for the true barline (up) or for the half-bar (down), under a walking bass that states each root on the barline, one that anticipates it by a quaver and ties it over, and one that puts the fifth on the barline and the root at the half-bar. a bass stating the root: root 51 for, 0 against; motion 0 for, 13 against; melody 51 for, 0 against. the root anticipated: root 50 for, 0 against; motion 0 for, 100 against; melody 51 for, 0 against. the root arriving late: root 0 for, 100 against; motion 0 for, 10 against; melody 51 for, 0 against. Anticipated, the root still sounds on the barline, tied over, and still votes for it; the bass's motion, which now lands a quaver early and moves again at the half-bar, votes against it in every bar. Late, the root votes against the bar every time. Harmony and voice leading

A late root is believed because the record is kept by the reading

The habit worth bringing to a tune turned out to be trusting the bass's root and doubting everything else. A bass that anticipates the next chord does not break that habit — the root is still tied across the barline, and it is the bass's motion that lies. A bass that puts the root at the half-bar does break it, and nothing a listener learns can repair it, because each witness is judged by whether it agrees with the reading, and the reading was made by the liar.

 The chromatic notes the reading already hears, and the ones it does not. Every note in the listing's six schemes that lies outside the key the reading hears it in, counted once a bar, by scheme, and how many of them are notes of the chord the reading hears in that bar: 32-bar song 6 of 6, rondo 5 of 5, blues 0 of 30 — 11 of 41 in all. Every chromatic note of the song's bridge and the rondo's episodes is the third of an applied dominant the reading already hears, and so already accounted for. All 30 it leaves unaccounted for are the blues's sevenths. Scales and modes

The leading note is already inside the chord that hides the key

A key-finder kept hearing a rondo's episodes as passages inside C, and the escape it used was to hear a D7 as a D minor chord, throwing away its F♯. Charge a chromatic note by how strongly it points at another key, and that escape closes. Nothing else moves, because every F♯ and G♯ in the episodes was already inside the chord the reading heard — as the third of V of V. The charge takes a chromatic note seriously everywhere except where it was needed, and in the blues it mistakes a flattened seventh for a leading note.

 Two honest bars are enough for a listener that can change its habits. Two bars of ground and eight of a tune that marks the bar, over a walking bass whose root arrives late in every bar, with a drum on the downbeat of the passage's first bars and silent after: the share of the tune's bars heard against the barline by four listeners, against how many bars the drum plays. A listener who no habits: 0 0, 1 0, 2 5, 3 37, 4 58, 6 72, 8 74, 10 74. A listener who trusts the root, doubts the rest: 0 0, 1 0, 2 0, 3 6, 4 12, 6 25, 8 38, 10 50. A listener who doubts the root: 0 74, 1 98, 2 100, 3 100, 4 100, 6 100, 8 100, 10 100. A listener who chooses among all three: 0 0, 1 10, 2 93, 3 100, 4 100, 6 100, 8 100, 10 100. A record scores each vote against the reading held when it is cast. One stroke forms a reading that the next bar of root erases; two leave it standing when the tune enters, and from then on every bar scores the root as the witness that disagrees. A listener crediting the root and the drum equally is held at a tie by any amount of drum alone. Harmony and voice leading

Two honest bars convict a lying root

A bass that states its root late defeats every listener who learns from agreement, because the reading forms before any honest witness speaks. Give the passage a drum on its first downbeats and the same record that ratified the liar convicts it — but only a listener holding a habit that doubts the root can act on the verdict. Two strokes are enough for that listener, for as long as the tune lasts. One is not, and no amount of drum is enough for a listener that credits the root and the drum alike.

 A vibrato hides a section's beats above about four hundred hertz. Two voices of a section whose tuning is spread by 15 cents, singing one note: how many nulls of their beating a listener can follow each second — counted only while the beat is at or below fifteen a second — at the frequency of the partial compared, sung straight and with a 71-cent vibrato at about six hertz. 220 Hz 2.2 straight, 5.3 with vibrato; 330 Hz 3.3 straight, 4.4 with vibrato; 440 Hz 4.2 straight, 3.9 with vibrato; 550 Hz 5.0 straight, 3.3 with vibrato; 660 Hz 5.1 straight, 2.8 with vibrato; 880 Hz 5.5 straight, 2.4 with vibrato; 1100 Hz 4.8 straight, 2.0 with vibrato; 1320 Hz 4.6 straight, 1.5 with vibrato; 1760 Hz 3.9 straight, 1.2 with vibrato; 2200 Hz 3.5 straight, 1.1 with vibrato; 2640 Hz 3.2 straight, 0.9 with vibrato; 3200 Hz 2.3 straight, 0.7 with vibrato; 3520 Hz 2.0 straight, 0.7 with vibrato. Below 413 hertz the vibrato adds followable beats, because it sweeps a slow difference faster. Above it the vibrato removes them, because it carries the pair's difference past fifteen hertz most of the time; at 3,200 hertz, the singer's formant, a straight section gives 3.4 times as many. Instruments and their design

A vibrato hides a choir's tuning above four hundred hertz

Thirteen essays drew a vibrato that had always been running; a real one grows out of a straight tone over the first quarter-second of a note. Counting the beats a listener can follow between two voices of a section shows what that front edge is worth. A vibrato takes most of a section's countable beats away above about 400 hertz — seven in ten of them at the singer's formant — and adds some below. The straight start gives them back only while the vibrato is still growing, and in the band where a voice carries that is less than one beat.

 A dancer who hears the crowd as noise buys back most of what caution cost. The landmark stream, downbeats marked, 10 metres known, 25 per cent of the time lost to noise in stretches of the stated mean length: the share of metres recognised within the 16-step present by a listener who cannot tell noise from silence and so allows for losses everywhere, and by one who hears the noise and counts a silence only where the room was quiet. Not hearing it: one by one 27%, 2 27%, 4 28%, 8 30%, 16 32%, 32 34%, 64 35%. Hearing the noise: one by one 51%, 2 50%, 4 50%, 8 52%, 16 54%, 32 57%, 64 59%. Nothing lost at all: 81%. Nothing lost but the same caution: 48%. A listener who loses a quarter of the music and knows where does better than one who loses nothing and must allow for losing it. Rhythm and metre

A dancer who hears the crowd can trust the quiet

A late dancer in a noisy room who cannot tell noise from silence has to treat every silence as weak evidence, and that caution was the largest cost in every figure before this one. A dancer who hears the laugh that covers a bar knows which silences to discount and can trust the rest. Losing a quarter of the music, that dancer recognises more metres than one who loses nothing but must allow for losses — and the landmarks, which a noisy room had made the worse stream to follow, lead again once the noise comes in bursts. The quiet has to be trusted to one onset in nineteen.

 A bell is never minor alone. For bells whose hum is on C1, E1, G1, C2, E2, G2, C3, E3, G3, C4, struck so that the hum is at 20 to 110 decibels: whether the masked-threshold count delivers the tierce, the minor third above the prime, and the deciem, the major third above the nominal. C1: ···MMMM············; E1: ··MMMMMMMM·········; G1: ··MMMMMMMMMM·······; C2: ·MBBBMMMMMMMMMM····; E2: ·BBBBBBBBMMMMMMM···; G2: ·BBBBBBBBBBMMMMMM··; C3: BBBBBBBBBBBBBMMMMM·; E3: BBBBBBBBBBBBBBBMMM·; G3: BBBBBBBBBBBBBBBBMMM; C4: BBBBBBBBBBBBBBBBBMM (B both thirds, M the major only, m the minor only, a dot neither; levels in steps of five from 20). Of 190 cells, 82 deliver both thirds, 59 the major third only, 49 neither, and 0 the minor third alone. A dot in a cell marks a stroke at which fewer than two of the three modes that imply the strike note survive. Perception and the listener

A bell is never minor alone

A tuned bell carries two thirds: the tierce, a minor third above the prime, which is why a bell is called minor, and the deciem, a major third above the nominal. Run the masking count across every register and every stroke and the minor one is never delivered without the major one. Struck softly a bell delivers both; struck harder, only the major; and below C2 the tierce is never delivered at all. The deciem survives because it is masked by one mode barely louder than itself, the tierce by two that outweigh it.

 A function label gives back a first hearing's place only if it does not recur. A through-composed piece of eight one-minute sections, heard for the first time by a listener told its plan, with a tenth of lapses taking every count and lasting 40 seconds on average: proportions told apart across the whole piece, against the chance that on returning into music not yet heard the listener recognises the section's function — opening (O), continuing (C), intensifying (I) or closing (Z) — for four arrangements of the functions. OIOIOIOI, two functions alternating: 0% 29.8, 25% 31.5, 50% 33.6, 75% 36.2, 100% 39.5. OCCCIIIZ, an arc in long runs: 0% 29.8, 25% 31.4, 50% 33.3, 75% 35.7, 100% 38.6. OOCCIIZZ, four in blocks of two: 0% 29.8, 25% 32.1, 50% 35.1, 75% 39.1, 100% 45.0. OCIZOCIZ, four, cycled: 0% 29.8, 25% 33.1, 50% 37.8, 75% 45.4, 100% 60.8. With no label the listener reads 29.8; knowing the piece, 66.1. Form and structure

A section that says what it is for places a first hearing

A first-time listener who loses count in a through-composed piece comes back into music they have never heard, and new music says only that it is new. Suppose instead each section announces what it is doing — opening, continuing, intensifying, closing — and the listener recognises it. How much of their place that gives back depends hardly at all on how many functions there are and almost entirely on how soon the same function comes back. Four functions cycled restore 85 per cent of what a first hearing costs at forty-second lapses; the same four in the familiar arc, two sections each, restore 42; two functions alternating restore 85 per cent at twenty seconds and 14 at a minute.

 An honest drum that comes late is too late, however long it plays. Two bars of ground and twenty-four of a tune that marks the bar, over a walking bass whose root arrives late in every bar, with an honest drum on every downbeat from a stated bar to the end: the share of passages a listener choosing among three sets of habits ends on the true barline, for four starting credits for the drum. Credited 0.75: from bar 1 100%, from bar 2 0%, from bar 3 0%, from bar 4 0%, from bar 5 0%, from bar 6 0%, from bar 7 0%, from bar 9 0%. Credited 0.9: from bar 1 100%, from bar 2 99%, from bar 3 29%, from bar 4 0%, from bar 5 0%, from bar 6 0%, from bar 7 0%, from bar 9 0%. Credited 0.95: from bar 1 100%, from bar 2 100%, from bar 3 88%, from bar 4 0%, from bar 5 0%, from bar 6 0%, from bar 7 0%, from bar 9 0%. Credited 0.99: from bar 1 100%, from bar 2 100%, from bar 3 100%, from bar 4 88%, from bar 5 0%, from bar 6 0%, from bar 7 0%, from bar 9 0%. At the credit the drum has always had, three in four, a drum that misses the first bar never recovers the reading, though it then plays twenty-five bars running. Each bar the late root is believed before the drum arrives raises the credit it would need; at 0.99, ninety-nine to one, the drum must still start by the fourth bar. Harmony and voice leading

An honest drum has to come first

Two strokes of an honest drum at the very start of a passage protect a listener from a bass whose root arrives late in every bar. The same two strokes a bar later do nothing at all, and neither does a drum that then plays every downbeat to the end. A listener that keeps a record of its witnesses scores the late drum against the reading the root has already made, and convicts it within two bars. Every bar the lie is believed first multiplies the trust the drum would need, and by the seventh bar no trust short of certainty is enough.

 A section's vibratos must lock almost completely before its tuning is exposed again. Two voices of a section spread by 15 cents, each with a 71-cent vibrato near six hertz, their vibratos pulled together so that their differences of rate and phase shrink to a stated share: followable beats a second at three partial frequencies, against how far they are locked, with the same voices sung straight as the dashed line beside each. 660 Hz (straight 5.14): 0 2.77, 0.25 2.90, 0.5 3.23, 0.7 3.85, 0.8 4.45, 0.85 4.98, 0.9 5.02, 0.95 5.31, 0.98 5.17, 1 5.15; 1217 Hz (straight 4.66): 0 1.60, 0.25 1.80, 0.5 2.04, 0.7 2.58, 0.8 2.88, 0.85 3.29, 0.9 3.88, 0.95 4.41, 0.98 4.67, 1 4.66; 3200 Hz (straight 2.33): 0 0.68, 0.25 0.70, 0.5 0.80, 0.7 1.13, 0.8 1.10, 0.85 1.45, 0.9 1.70, 0.95 1.98, 0.98 2.12, 1 2.23. At the singer's formant, half-locked vibratos expose 34 per cent of what a straight section does, locked to nine-tenths 73 per cent, locked to ninety-five 85 per cent; only at about ninety-eight per cent does it come within a tenth of a straight section. Instruments and their design

A choir's vibratos must lock to two milliseconds

A vibrato hides a section's mistuning from a listener counting beats, and the obvious way it could stop hiding it is for the singers' vibratos to move together. They would have to move very nearly together. Half-entrained vibratos leave a section as hidden at the singer's formant as independent ones; locked to nine-tenths they expose three-quarters of what a straight section does; only past nineteen parts in twenty does the section's tuning show as if they sang straight. At six cycles a second that is two vibratos peaking within two milliseconds of each other — and the better a section is tuned, the tighter the lock has to be.

 In a loud room the accent is worth more, if it is loud enough to get through. Ten metres known, entered at a random step, recognised within the 16-step present by a listener who hears the room's noise: how much marking the downbeat adds to the share recognised from the landmarks, when the noise covers a tenth, a quarter or half of the time in stretches of sixteen steps, every marked downbeat is heard through it, and a stated share of the unmarked strokes are. 10% noise: 0% 23.3, 25% 24.0, 50% 24.4, 75% 24.7, 100% 23.4 points; noise masking everything 21.4. 25% noise: 0% 22.9, 25% 24.6, 50% 25.9, 75% 26.7, 100% 23.4 points; noise masking everything 18.1. 50% noise: 0% 19.8, 25% 23.8, 50% 27.0, 75% 28.9, 100% 23.4 points; noise masking everything 11.3. In a quiet room the accent is worth 23.4 points. Rhythm and metre

In a loud room the accent has to be loud

A late dancer who hears the crowd as noise ignores everything played under it. Real noise is not that thorough. It hides the soft strokes and lets the loud ones through, and an accented downbeat is a loud one. Let the noise pass loud strokes and some soft ones, and marking the downbeat is worth more to the dancer than in a quiet room: with half the time under noise, 29 points of recognition against 23. Noise that swallows the accent halves its worth. In a quiet room no amount of loudness bought an accent anything extra; in a loud one, loudness is the whole of what the accent is for.

 A tuned chord struck in step drifts out of step as soon as it is mistuned. The tuned close seventh over C2, every note struck at the same instant and held, each note detuned by a draw with the stated spread in cents: the mean correlation between filters' envelopes, in half-second windows every quarter-second through eight seconds, averaged over eight draws. 0 cents: 0.61, 0.59, 0.59, 0.59, 0.59, 0.59, 0.59, 0.59; 0.1 cents: 0.61, 0.56, 0.50, 0.41, 0.36, 0.32, 0.28, 0.25 (halfway down by 4 s); 0.25 cents: 0.60, 0.43, 0.30, 0.23, 0.20, 0.18, 0.18, 0.17 (halfway down by 1.75 s); 0.5 cents: 0.57, 0.29, 0.19, 0.17, 0.17, 0.16, 0.18, 0.19 (halfway down by 1 s); 1 cent: 0.50, 0.18, 0.17, 0.18, 0.14, 0.18, 0.14, 0.17 (halfway down by 0.5 s), at whole seconds. The tempered chord's level is 0.16. Exactly in tune the chord stays in step; a tenth of a cent takes four seconds to lose half of it, a cent half a second. Timbre and acoustics

A tuned chord stays in step for half a cent-second

A close seventh chord tuned 4 : 5 : 6 : 7 low on the keyboard fluctuates in every auditory filter at once, but only if its notes start within a millisecond of each other — and a tuned chord, it was said, keeps its phases for as long as it sounds. It keeps them only as long as it stays exactly tuned. Mistune its notes by a fraction of a cent and they slide against each other, and the chord drifts out of step in a time inversely proportional to the mistuning: about half a cent-second. A tenth of a cent holds it for four seconds, a cent for half a second. A chord that starts out of step never drifts into it.

 Four functions cycled lose their advantage the moment neighbours are confused. A through-composed piece of eight one-minute sections, a first hearing, a listener told the plan of functions, lapses of forty seconds on average: proportions told apart across the whole piece against the chance that a recognised function is heard as its neighbour in the arc opening, continuing, intensifying, closing, for five arrangements. OCIZOCIZ, four, cycled: 0% 60.8, 5% 47.7, 10% 44.1, 15% 41.3, 20% 39.0, 30% 36.2, 40% 35.1, 50% 34.9. OOCCIIZZ, four in blocks of two: 0% 45.0, 5% 40.0, 10% 35.3, 15% 34.4, 20% 33.0, 30% 32.3, 40% 31.3, 50% 31.5. OIZOIZOI, three, cycled: 0% 51.4, 5% 43.2, 10% 41.9, 15% 40.8, 20% 39.8, 30% 37.9, 40% 37.1, 50% 36.7. OZOZOZOZ, opening and closing alternated: 0% 39.5, 5% 39.5, 10% 39.5, 15% 39.5, 20% 39.5, 30% 39.5, 40% 39.5, 50% 39.5. OCOCOCOC, opening and continuing alternated: 0% 39.5, 5% 37.3, 10% 35.3, 15% 33.7, 20% 31.8, 30% 31.0, 40% 30.4, 50% 30.6. Unlabelled, 29.8; knowing the piece, 66.1. Form and structure

Two functions far apart outlast four that blur

A first-time listener who loses their place in a through-composed piece gets most of it back if each section announces what it is doing and the same function does not come back too soon: four functions cycled restore 85 per cent of what a first hearing costs. That rests on hearing each function exactly. Let a listener mistake a function for its neighbour in the arc — a continuation for an intensification — one time in ten, and the cycle keeps less than half of its gain. Two functions at the opposite ends of the arc, opening and closing alternated, lose nothing to such confusion, and from a little over one confusion in five no arrangement does better.

 A break lets a late drum win only if the listener comes out of it holding no reading and almost no record. Two bars of ground and twenty-four of a tune that marks the bar, over a walking bass whose root is late in every bar, the choosing listener, and an honest drum on every downbeat from bar 5, 9 or 13 to the end, entering after a break at which the listener keeps a stated share of its witnesses' records. Solid: the break also wipes its reading. From bar 5: 96% keeping 0, 96% keeping 0.02, 96% keeping 0.05, 91% keeping 0.1, 46% keeping 0.2, 42% keeping 0.3, 30% keeping 0.5, 8% keeping 0.75, 7% keeping 1. From bar 9: 95% keeping 0, 92% keeping 0.02, 47% keeping 0.05, 43% keeping 0.1, 11% keeping 0.2, 2% keeping 0.3, 0% keeping 0.5, 0% keeping 0.75, 0% keeping 1. From bar 13: 93% keeping 0, 57% keeping 0.02, 44% keeping 0.05, 10% keeping 0.1, 1% keeping 0.2, 0% keeping 0.3, 0% keeping 0.5, 0% keeping 0.75, 0% keeping 1. Dashed, on the floor: the same breaks leaving half the reading's margin, where the drum wins no passage at any share of the record kept. The horizontal axis is drawn as the square root of the share. Harmony and voice leading

A break rescues a late drum only by ending the reading

A drum that enters after a late root has made the reading needs odds in the hundreds to undo it. Bring it in after a break in the texture, and the obvious guess is that the break helps by making the listener forget whom it trusted. It barely does. What rescues the drum is the listener forgetting what it heard: with the old reading gone, a drum on bar 9 needs odds of nine to one, and with the record gone too it needs less than the three to one it has always had. A break that leaves even a hundredth of the old reading behind rescues nothing at all.

 A vibrato short of perfect lock takes an exactly tuned chord out of step. The tuned close seventh over C2, exactly in tune and every note struck at the same instant, held for eight seconds: the mean correlation between filters' envelopes, in half-second windows, averaged over eight draws. Straight: 0.61, 0.59, 0.59, 0.59, 0.59, 0.59, 0.59, 0.59. With a 71-cent vibrato on every note, fully locked: 0.59, 0.57, 0.57, 0.57, 0.57, 0.57, 0.57, 0.57. Locked 98 per cent: 0.59, 0.55, 0.52, 0.47, 0.43, 0.39, 0.37, 0.33, halfway to the tempered chord's 0.16 by 6 s, at whole seconds. A straight tuned chord stays in step for as long as it is held; a vibrato can only keep it there by being locked completely. Timbre and acoustics

A shared vibrato cannot hold a tuned chord together

A tuned close seventh chord binds its auditory filters together only while its notes stay within a third of a millisecond of each other, and mistuned by a fraction of a cent it drifts out of step in half a cent-second. A vibrato shared by every voice looked like a way to hold it: if all four notes swing together, they slide together. They do not stop sliding. A perfectly locked vibrato leaves the drift exactly as fast as before, because it changes every note's frequency by the same factor and leaves their differences alone; a vibrato locked ninety-eight per cent takes an exactly tuned chord out of step in six seconds, and one locked ninety per cent in one.

 Heard a little late, two vibratos lock in step; heard a third of a cycle late, they lock in opposition. Two singers' vibratos at 6.0 and 6.2 a second, each pulled towards the phase it hears from the other with a strength of 10 radians a second, the hearing delayed by a stated time: how far apart the two vibratos' peaks settle, in milliseconds, after twenty seconds. 0 ms: 1.6 apart (4°); 20 ms: 2.4 apart (5°); 40 ms: 61.9 apart (-169°); 60 ms: 70.7 apart (-176°); 80 ms: 79.5 apart (-176°); 100 ms: 88.1 apart (-176°); 120 ms: 95.7 apart (-175°); 140 ms: 1.6 apart (4°); 160 ms: 1.6 apart (4°); 180 ms: 1.8 apart (4°); 200 ms: 69.6 apart (-176°). Under about forty milliseconds the lock is near in phase; from about forty to a hundred and twenty it is near antiphase, the peaks most of half a cycle apart; from about a hundred and forty, nearly a whole cycle late, it is near in phase again. The shaded band is the two milliseconds the formant asked for. Instruments and their design

Singers who hear each other late lock their vibratos in opposition

A choir's vibratos give back the section's tuning at the singers' formant only if they are locked to within about two milliseconds. Singers who lock do it by hearing each other, and they hear each other late. Two vibratos pulled towards what each hears of the other settle near in step while the delay is under about forty milliseconds, and even then come within two milliseconds only if each singer corrects a phase error in about thirty. From forty to a hundred and twenty milliseconds late — a third to two-thirds of a vibrato cycle — they lock in opposition, peaks seventy to ninety milliseconds apart, and hide the section's tuning more completely than vibratos that never locked at all.

 Talk buries a melody's accent ten decibels before it buries a low drum. A band at 70 decibels marking its downbeats four ways, each mark at the band's level, under a crowd whose talk has the long-term spectrum of speech, at a stated level above the band: the chance the marked downbeat is heard, band by band at its clearest, with the crowd's level fluctuating three decibels either way. a louder melody note: -10 dB 100%, -5 dB 100%, 0 dB 100%, +5 dB 94%, +10 dB 47%, +15 dB 4%, +20 dB 0%; a bass note an octave below: -10 dB 100%, -5 dB 100%, 0 dB 100%, +5 dB 100%, +10 dB 100%, +15 dB 90%, +20 dB 34%; a low drum: -10 dB 100%, -5 dB 100%, 0 dB 100%, +5 dB 100%, +10 dB 100%, +15 dB 100%, +20 dB 82%; a high bell: -10 dB 100%, -5 dB 100%, 0 dB 100%, +5 dB 100%, +10 dB 100%, +15 dB 93%, +20 dB 43%; the melody's unmarked strokes, eight decibels under the accent: 100%, 99%, 72%, 14%, 0%, 0%, 0%. Talk is loudest from 250 to 400 hertz, where a melody's fundamentals are; it is fifteen decibels quieter at 63 hertz and twelve quieter at 3 kilohertz. Rhythm and metre

In a crowd the downbeat belongs to the drum

In a loud room a marked downbeat is worth more to a late dancer than in a quiet one, so long as the mark gets through. Whether it does depends on where its sound lies, because talk is not loud everywhere: it is loudest from 250 to 400 hertz, exactly where a melody's notes are, and fifteen decibels quieter at the bottom of a drum. A louder melody note is heard half the time through talk ten decibels over the band; a low drum survives a crowd twelve decibels louder than that. With the talk fifteen decibels over the band, a downbeat marked in the melody is worth two-fifths less to the dancer than one marked by the drum.

 A listener who has to learn the plan loses most where the plan said most. Eight one-minute sections, a first hearing, lapses of forty seconds: proportions told apart across the whole piece, against the chance a recognised function is heard as its neighbour in the arc, for a listener told the plan of functions (dashed) and one who must infer it from the functions heard before each lapse, among five plans it thinks equally likely (solid). four, cycled: told 60.8, 44.1, 39.0, 34.9; learning 44.2, 38.5, 35.4, 33.3; four in blocks of two: told 45.0, 35.3, 33.0, 31.5; learning 39.1, 33.2, 32.4, 30.7; opening and closing alternated: told 39.5, 39.5, 39.5, 39.5; learning 37.1, 37.1, 37.1, 37.1 at confusions of 0, 0.1, 0.2 and 0.5. A listener who knows the piece reaches 66.1; one with no functions, 29.8. Form and structure

A listener who learns the plan keeps only the far pair's immunity

A first-time listener who recognises what each section of a through-composed piece is doing gets most of their place back after a lapse — if they know the plan of functions. A listener who has to infer the plan from the sections heard so far loses most where the plan said most: four functions cycled give up more than half of what their labels were worth, nearly all of it to lapses in the first two minutes, before the second function has settled which plan the piece has. Opening and closing alternated lose a little and stay immune to confusing neighbouring functions. The far pair's immunity was never knowledge of the plan; it was the arc.

 A silent break ends the listener's reading only after as many bars as the reading has votes in hand. The choosing listener, every root late, and a break of silence before bar 3, 5, 9 or 13 through which it goes on expecting a downbeat once a bar at the reading it holds, each failure charging that reading 1 vote: the share of passages whose reading the silence ends, against the break's length in bars. Before bar 3, mean margin 2.9 votes: 1 bar 0%, 2 bars 0%, 4 bars 100%, 8 bars 100%, 16 bars 100%; half ended by 3 bars. Before bar 5, mean margin 6.8 votes: 1 bar 0%, 2 bars 0%, 4 bars 0%, 8 bars 88%, 16 bars 100%; half ended by 8 bars. Before bar 9, mean margin 20.7 votes: 1 bar 0%, 2 bars 0%, 4 bars 0%, 8 bars 0%, 16 bars 0%; half ended by 24 bars. Before bar 13, mean margin 38.5 votes: 1 bar 0%, 2 bars 0%, 4 bars 0%, 8 bars 0%, 16 bars 0%; half ended by 48 bars. The horizontal axis is logarithmic in one plus the length. Harmony and voice leading

A silent break has to outlast the reading's lead

A listener waiting through a break goes on expecting downbeats at the reading it holds, and each one that fails to come can be charged against that reading. Charged at a witness's vote apiece, the silence ends the reading only after as many bars as the reading has votes in hand: three bars after two bars of tune, eight after four, twenty-one after eight. A one-bar break would need a missing downbeat to cost twenty votes. And a silence that does end the reading leaves the record standing, which convicts the drum anyway.

 With the tierce at 1.25 times the prime, the lower modes point only at the strike note's own name. For bells whose hum is on C1, E1, G1, C2, E2, G2, C3, E3, G3, C4, struck with the hum at 20 to 110 decibels, the tierce at 1.25 times the prime, 386 cents: the series the delivered lower modes — hum, prime, tierce and quint — fit best, within thirty cents. C1: ···················; E1: ·······SSS·········; G1: ·····SSSSSSS·······; C2: ··SSSSSSSSSSSS·····; E2: ·SSSSSSSSSSSSSSS···; G2: ·SSSSSSSSSSSSSSS···; C3: SSSSSSSSSSSSSSSSS··; E3: SSSSSSSSSSSSSSSSSS·; G3: SSSSSSSSSSSSSSSSSS·; C4: SSSSSSSSSSSSSSSSSSS (S the strike note's name, 3 the note a major third below it, a dot fewer than two lower modes delivered; levels in steps of five from 20). 124 cells point at the strike's name, 0 a major third below, 66 at nothing. Perception and the listener

The raised tierce points the bell at its own note

A bell's four lowest modes — hum, prime, tierce and quint — can imply a pitch of their own. With a major tierce they are the second, fourth, fifth and sixth harmonics of the prime two octaves down: the strike note's own name, with one empty slot. It was the minor tierce that pointed elsewhere all along. Its lower modes are the fifth, tenth, twelfth and fifteenth harmonics of the note a major third below the strike, and in 82 of 190 struck bells that is the series the delivered modes fit. Raising the tierce removes a competitor rather than adding one.

 An announcement of the plan pays the cycle only when it is nearly always right. Recognised exactly: proportions told apart across the whole piece, averaged over announcements that name the true plan with the chance on the axis and otherwise one of the four wrong plans, for a listener who puts exactly that chance of its prior on the plan named; dashed, the listener told the plan. four, cycled: 0.2 44.2, 0.3 44.2, 0.4 44.2, 0.5 44.4, 0.6 46.6, 0.7 46.2, 0.8 46.6, 0.9 56.1, 0.95 56.8, 0.99 60.6; told 60.8, learning 44.2; four in blocks of two: 0.2 39.1, 0.3 39.1, 0.4 39.1, 0.5 37.3, 0.6 37.8, 0.7 40.9, 0.8 41.6, 0.9 42.0, 0.95 42.5, 0.99 44.8; told 45.0, learning 39.1; opening and closing alternated: 0.2 37.1, 0.3 37.1, 0.4 37.1, 0.5 38.6, 0.6 38.2, 0.7 38.6, 0.8 38.9, 0.9 38.8, 0.95 39.2, 0.99 39.4; told 39.5, learning 37.1. The cycle recovers half of what being told gives it only at 90 per cent reliability. Form and structure

An announced plan pays the cycle and punishes the blocks

A slow introduction, a title or a programme note can tell a first-time listener the plan of a piece before the first lapse — and be wrong. Trusted as far as it deserves, an announcement gives back half of what being told the plan is worth to four cycled functions only once it is right nine times in ten. But how far to trust it depends on the piece: a cycle rewards believing even a coin-toss announcement completely, while functions in blocks of two are made worse by an announcement right half the time, and a confident wrong one can cost a listener forty points of the first minute.

 A drum that is always felt is worth no more than one that is always heard, until the crowd drowns the hearing. Ten metres known, entered at a random step, recognised within the 16-step present, the crowd covering half the time at a stated level above the band: the points marking the downbeat adds, for each way of marking it heard through the crowd, and for a low drum that is felt as well and so always arrives. a louder melody note: +0 28.9, +5 20.4, +10 14.7, +15 11.6, +20 11.3, +25 11.3, +30 11.3; a bass note an octave below: +0 28.9, +5 22.1, +10 19.8, +15 18.0, +20 13.7, +25 11.4, +30 11.3; a low drum: +0 28.9, +5 22.1, +10 19.8, +15 19.8, +20 17.3, +25 12.9, +30 11.4; a high bell: +0 28.9, +5 22.1, +10 19.8, +15 18.2, +20 14.3, +25 11.5, +30 11.3; a low drum, felt as well: +0 28.9, +5 22.1, +10 19.8, +15 19.8, +20 19.8, +25 19.8, +30 19.8. A quiet room's mark is worth 23.4. Rhythm and metre

A felt drum reaches only where the heard one already does

A low drum near a stage is felt in the body as well as heard, and talk cannot mask a vibration. Given that second channel, a drum-marked downbeat always arrives — and is then worth 19.8 points to a late dancer, exactly what an always-heard drum was worth, because the crowd has already taken the soft strokes that made a mark worth more. The felt channel pays only under crowds twenty decibels over the band. And it is driven by the drum's own level, which in a reverberant room is high enough to be felt only where the drum is already loud enough to be heard.

 Locked 95 per cent, the vibratos carry the chord back into step 13 times in a second. The tuned close seventh over C2, exactly in tune, with the typical vibrato on every note — 71 cents at about six a second — the four vibratos 95 per cent locked to the bass's: how far ahead of or behind the bass each upper note's waveform runs, in milliseconds, through the fourth second, for one draw of rates and phases. The band is ±0.3 ms, within which the chord counts as in step. In this second the chord is in step for 180 ms in 13 stretches; over eight seconds and eight draws, 29 per cent of the time, in stretches averaging 22 ms after the first two seconds. One cycle of the chord's own flutter takes 61 ms. Timbre and acoustics

A chord comes back into step too briefly to bind

A vibrato that is shared imperfectly does not carry a tuned chord's notes ever further apart: each note's time is warped back and forth, and the chord passes back through step twice in every swing. Locked ninety-seven or ninety-eight per cent, it does so about ten times a second. But each return lasts on average less than one cycle of the chord's own flutter, which is the rhythm the ear's filters would have to share to show the binding — fewer than one return a second lasts that long. The chord is in step often and never for long enough.

 Where a pair would lock in opposition, a section of sixteen does not lock at all. Sixteen singers in two rows of eight, sixty centimetres apart, each hearing every other after the sound's travel and a common latency, each vibrato pulled towards all it hears with a total strength of 10 radians a second; four draws of rates and phases. Against the latency: R1, the length of the mean of the vibratos' phases, one when every vibrato is in step; and R2, the same for twice the phases, one also when the section has split into two opposed halves. 0 ms: 0.96, 0.86; 10 ms: 0.94, 0.83; 20 ms: 0.89, 0.71; 30 ms: 0.72, 0.52; 40 ms: 0.47, 0.31; 50 ms: 0.09, 0.24; 60 ms: 0.07, 0.22; 70 ms: 0.08, 0.23; 80 ms: 0.08, 0.22; 90 ms: 0.10, 0.24; 100 ms: 0.19, 0.22; 110 ms: 0.68, 0.46; 120 ms: 0.90, 0.73; 135 ms: 0.96, 0.85; 150 ms: 0.97, 0.87. A section split in two would show R2 high with R1 low; it never does. Instruments and their design

Sixteen singers cannot all oppose each other

Two singers who hear each other a third of a vibrato cycle late lock their vibratos in opposition and hide their tuning better than if they had not listened at all. Sixteen cannot do it, because no arrangement puts every one of them opposite every other. At the latencies where a pair opposes, a section of sixteen does not lock at all: each singer keeps a vibrato rate of its own, and a listener hears the section's tuning exactly as through independent vibratos. Outside that band the section locks in step and gives most of its tuning back. A section that listens to itself can expose its tuning or leave it as it was, but never hide it.

 The tuning that saves roughness rises only downward, and keeps the clock only if the low octave leaves the goals alone. Five ways of setting the beat rates of a gamelan's paired instruments over three octaves, the rate rising as the pitch to the power 2 where it rises: the roughness the detuned pairs add to the passage, against the bits of position a listener who fuses a goal's swells has left before each goal — filled with the goals struck in all three octaves, open with them struck only in the middle and high octaves. middle octave's rates repeated: roughness 3.06, bits 0.15 and 0.15; rise carried on: roughness 3.37, bits 0.40 and 0.40; repeated at the goals, rising elsewhere: roughness 3.42, bits 0.15 and 0.15; rising below the middle, repeated above: roughness 2.45, bits 0.40 and 0.15; the same, goals repeated in every octave: roughness 2.96, bits 0.15 and 0.15. Scales and modes

A gamelan's shimmer should rise only downward

A gamelan tuner who lets each pair's beat rate follow its pitch saves roughness in the low octave, where the pairs slow, and pays it in the high octave, where they race. Repeating the middle octave's rates only at the goal tones and letting the rest follow the pitch, as the lead proposed, is rougher than repeating everywhere. The tuning that saves is the one that lets the rate follow the pitch below the middle octave and repeats it above — a fifth less roughness at the square — and it keeps the clock only if the low octave does not double the goals.

 After a silence long enough to end the reading, a bar of eight fill strokes hands the next reading to the drum. The choosing listener, every root late, a silent break before bar 3, 5 or 9 at one vote a failed downbeat, then a bar of 8 fill strokes for the true barline heard as the drum, then the drum on every downbeat: the share of passages ending on the barline against the silence's length (solid), with the silence alone dashed. Before bar 3: 0 bars 86%, 2 bars 100%, 4 bars 100%, 8 bars 100%, 12 bars 100%, 24 bars 100%; silence alone 4 bars 44%, 8 bars 44%, 12 bars 44%, 24 bars 44%. Before bar 5: 0 bars 0%, 2 bars 0%, 4 bars 53%, 8 bars 99%, 12 bars 100%, 24 bars 100%; silence alone 4 bars 0%, 8 bars 5%, 12 bars 7%, 24 bars 7%. Before bar 9: 0 bars 0%, 2 bars 0%, 4 bars 0%, 8 bars 0%, 12 bars 0%, 24 bars 99%; silence alone 4 bars 0%, 8 bars 0%, 12 bars 0%, 24 bars 0%. The horizontal axis is logarithmic in one plus the length. Harmony and voice leading

A fill cannot end a reading, but it can win the next one

A drum fill in the break before a late drum is a witness that speaks while the old reading still stands, so every stroke it plays is scored as disagreeing, and its credit falls to a coin toss within two or three strokes: at three to one it can cast 1.37 votes in all, against a lead of twenty by the middle of a tune. A fill cannot end a reading. But once a silence has ended it, the same fill sets the next one, and the record that convicted the drum through every length of silence starts convicting the bass instead: after twelve bars of silence before bar 5 the drum goes from 7 per cent of passages to all of them.

 Two parts entering together mask each other unequally: the carrier loses a partial or two, the passenger up to eleven. Every pairing of two parts entering together on a chord of the passage in which the quieter part sits under its mask on average, 71 of them, sorted. Up: how many of the carrier's partials that are clear of the mask with the carrier alone over the chord are pushed under it by the passenger. Down: how many of the passenger's are pushed under by the carrier. In all, carriers lose 19 partials and passengers 135. 54 carriers lose none, 15 lose one and 2 lose two; passengers lose 0 in 25, 1 in 25, 2 in 6, and three or more in 15, the most being 11. Form and structure

A carrier gives up room, not partials

A passenger masks its carrier as the carrier masks it, so the arrival might be a new spectrum belonging to neither part. Across the seventy-one pairings of the passage it almost never is. Carriers lose nineteen partials in all to their passengers and passengers lose a hundred and thirty-five to their carriers; a voice beside a violin takes thirteen decibels of the violin's room and not one of its partials. The nine pairings that leave neither part a majority are mostly on one chord, and there the chord had already done it.

 Locked 99 per cent, the higher the chord the more of its returns are long enough to bind. The tuned close seventh over C2, C3 and C4, exactly in tune, with the typical vibrato on every note, the four vibratos 99 per cent locked, each register judged against its own onset tolerance and its own flutter: how many times a second, after the first two seconds, the chord comes back into step, and how many of those returns last a whole cycle of its flutter, over eight draws. C2 (tolerance 0.306 ms, flutter 61.2 ms): 4.6 returns a second, 0.54 a flutter cycle long, in step 88 per cent of the time; C3 (tolerance 0.153 ms, flutter 30.6 ms): 9.7 returns a second, 3.83 a flutter cycle long, in step 65 per cent of the time; C4 (tolerance 0.070 ms, flutter 15.3 ms): 12.0 returns a second, 5.13 a flutter cycle long, in step 38 per cent of the time. Timbre and acoustics

Two octaves up, a sung chord can bind at every swing

A vibrato's swing between two notes is the same number of milliseconds in every register; the tolerance for being in step halves each octave, and so does the period of the chord's own flutter. The first two make a scaling law — an octave higher is the same chord with its vibratos disagreeing half as much — and the third breaks the tie. Matched for time in step, the higher chord's returns carry twice as many flutter cycles: over C2 fewer than one return a second is long enough to bind, over C4 with the vibratos locked ninety-nine and a half per cent nearly every return is.

 A displaced stroke helps a listener keep its place only if it makes a figure the bell never makes. The standard bell pattern, 2 2 1 2 2 2 1, with both pulses heard half the time, a listener who learns how often strokes are lost: the surprise a step once placed, averaged over the second to sixth cycles, when every third cycle one stroke of the bell is moved one step into an empty step. Each bar is one of the 10 possible displacements, labelled by the varied cycle's gaps, on a logarithmic axis; the lines are the plain bell (0.0078) and the other order of the gaps, 2 2 2 1 2 1 2 (0.0016). 1 3 1 2 2 2 1 (a figure the bell never makes): 0.0055; 3 1 1 2 2 2 1 (a figure the bell never makes): 0.0056; 2 2 1 2 1 3 1 (a figure the bell never makes): 0.0062; 2 2 1 3 1 2 1 (a figure the bell never makes): 0.0066; 2 2 1 1 3 2 1 (a figure the bell never makes): 0.0077; 2 2 1 2 3 1 1 (a figure the bell never makes): 0.0080; 2 1 2 2 2 2 1 (the other order): 0.0107; 1 2 1 2 2 2 2 (the other order): 0.0107; 2 2 2 1 2 2 1 (the bell, shifted): 0.0252; 2 2 1 2 2 1 2 (the bell, shifted): 0.0253. Rhythm and metre

A varied bell can be a landmark or a false start

In a thin texture the standard bell pattern keeps a listener placed five times worse than another order of its gaps, because its two short gaps look alike. A lead drummer's variation — one stroke moved, once every few cycles — was predicted to give the bell back what it loses. The best single displacement every other cycle recovers seventy per cent of the gap; every third cycle, a third; every sixth, a seventh. Every variation makes the bell slower to arrive by. And the displacements that turn the varied cycle into the bell begun somewhere else make keeping one's place three times worse, because they sound exactly like having lost it.

 Held to the shimmer it promised, the downward tuning is rougher than repeating the middle octave's rates. A gamelan's paired instruments over three octaves, the rate rising as the pitch to the power 2 where it rises: the roughness the detuned pairs add to the passage for each tuning at the average shimmer it actually has (open) and with every rate rescaled so the average is the one the families promised (filled). middle octave's rates repeated: average 100% of the promised, roughness 3.06, held 3.06; rise carried on: average 100% of the promised, roughness 3.37, held 3.37; repeated at the goals, rising elsewhere: average 100% of the promised, roughness 3.42, held 3.42; rising below the middle, repeated above: average 63% of the promised, roughness 2.45, held 3.26; the same, goals repeated in every octave: average 83% of the promised, roughness 2.96, held 3.32; repeated, slowed to the same average: average 63% of the promised, roughness 2.27, held 3.06. Scales and modes

A gamelan's shimmer has no cheaper shape, only a slower one

The tuning that let each pair's beat rate follow its pitch below the middle octave and repeated it above saved a fifth of the roughness — and slowed the ensemble's average shimmer by more than a third while doing it. Held to the shimmer it promised, it is rougher than repeating the middle octave's rates everywhere, and the middle octave's rates simply slowed to the same average are smoother than either. With the average fixed, the cheapest shape runs the other way, and a knee is never cheaper than a step.

 A third part takes from the carrier what it would have taken alone; the room the passenger spent adds almost nothing. Every pairing of carrier and buried passenger, with each other part of the roster entering beside them as a third: 284 arrivals. For each third part, the carrier's partials — clear of the mask over the chord alone — that are under it with all three sounding, split into those the passenger alone already took, those the third part would take on its own, and those lost only to the two together. clarinet: 46 arrivals, 19 partials (14 to the passenger, 3 to the third alone, 2 only together), the third alone costing 1.5 dB of room; oboe: 50 arrivals, 20 partials (9 to the passenger, 9 to the third alone, 2 only together), the third alone costing 1.3 dB of room; brass: 46 arrivals, 23 partials (8 to the passenger, 14 to the third alone, 1 only together), the third alone costing 1.0 dB of room; voice: 47 arrivals, 40 partials (14 to the passenger, 20 to the third alone, 6 only together), the third alone costing 7.3 dB of room; violin: 47 arrivals, 127 partials (17 to the passenger, 108 to the third alone, 2 only together), the third alone costing 8.8 dB of room; flue: 48 arrivals, 209 partials (14 to the passenger, 187 to the third alone, 8 only together), the third alone costing 14.1 dB of room. In all 438 partials: 76 to the passengers, 341 to the thirds on their own, 21 only to the combination. Form and structure

A third part takes what it would have taken alone

A passenger costs its carrier room rather than partials, and the obvious worry is that the room compounds: that a third part, entering beside a carrier already ten decibels nearer its limit, finds it fragile. Across 284 three-part arrivals it does not. Of the 438 carrier partials lost with three parts sounding, 21 are lost only to the combination; the rest are taken by one of the two parts alone. Masking adds in power, so a threshold the passenger has raised rises less when a third part adds to it. What decides the loss is which part enters third, and two parts — the flue and the violin — take five sixths of it.

 A cadence hands the next section to a two-stroke fill only by emptying the reading, not by weakening it. The choosing listener, every root late, an honest drum from the bar after a cadence that leaves the share on the axis of the reading's lead (logarithmic, with "none" an emptied reading), with a bar of 2 fill strokes heard as the drum between them: the share of passages ending on the barline. Before bar 3 (mean lead 2.9 votes): 100% 7%, 50% 92%, 30% 100%, 20% 100%, 15% 100%, 12% 100%, 10% 100%, 7% 100%, 5% 100%, 3% 100%, 2% 100%, 1% 100%, 0% 100%. Before bar 5 (mean lead 6.8 votes): 100% 0%, 50% 0%, 30% 0%, 20% 26%, 15% 64%, 12% 87%, 10% 89%, 7% 90%, 5% 93%, 3% 95%, 2% 95%, 1% 96%, 0% 100%. Before bar 9 (mean lead 20.7 votes): 100% 0%, 50% 0%, 30% 0%, 20% 0%, 15% 0%, 12% 0%, 10% 0%, 7% 0%, 5% 1%, 3% 1%, 2% 1%, 1% 2%, 0% 100%. Before bar 13 (mean lead 38.5 votes): 100% 0%, 50% 0%, 30% 0%, 20% 0%, 15% 0%, 12% 0%, 10% 0%, 7% 0%, 5% 0%, 3% 0%, 2% 0%, 1% 0%, 0% 84%. Harmony and voice leading

A cadence has to empty the reading, not weaken it

A drum fill after a long enough silence wins the next section for an honest drum, because it speaks into a reading that is empty. A cadence might do the silence's work in a bar. It does, but only on one condition: before bar 9, a cadence that removes ninety-eight per cent of the listener's conviction leaves a two-stroke fill winning one passage in a hundred, and one that removes all of it lets the fill win every one. The fill turns the reading either way. What the last sliver decides is whether the fill's first stroke is recorded as a disagreement — and that record is what the late root then uses to win the reading back.

 A row that hears its neighbours late hides the tuning between them, exposes it between every second singer, and leaves the whole as it was. Sixteen singers in one row sixty centimetres apart, a draw of fifteen cents apart with 71-cent vibratos, each pulled at ten radians a second towards what it hears, each voice's pull weighted by the inverse of its distance to the power 8; four draws. The beats a second a listener can follow at 3,200 hertz, against the latency, for neighbouring pairs, for pairs two places apart, and over every pair. neighbours: 0 ms 1.62, 20 ms 1.36, 40 ms 0.69, 60 ms 0.30, 80 ms 0.28, 100 ms 0.29, 120 ms 0.78, 150 ms 1.63; every second singer: 0 ms 1.23, 20 ms 0.90, 40 ms 0.82, 60 ms 0.94, 80 ms 1.21, 100 ms 1.15, 120 ms 0.90, 150 ms 1.21; every pair: 0 ms 0.96, 20 ms 0.77, 40 ms 0.72, 60 ms 0.67, 80 ms 0.67, 100 ms 0.69, 120 ms 0.75, 150 ms 0.92. Sung straight 2.33, independent vibratos 0.68, a pair in opposition 0.19. Instruments and their design

A row that hears its neighbours moves its tuning, it does not hide it

Sixteen singers who all hear each other alike cannot lock in opposition, because opposition between every pair is impossible. A row that hears mainly its neighbours has no such obstacle — opposition along a chain is consistent — and at the latencies that make a pair oppose, it half-alternates: neighbours opposite each other, every second singer nearly in step. That hides the tuning between neighbours as well as a duet hides it, and exposes it between every second singer more than independent voices would. Over the whole row the two cancel to the hundredth. With no latency, the same local hearing stops the row locking at all, and hides nearly half of what a whole-row lock would show.

 In a ring of minor bells, five of the eight hidden pitches fall off the scale the ring rings. A ring of eight bells with minor tierces tuned to a major scale, the tenor's strike note C4: for each bell and each stroke with the hum at 40 to 90 decibels, what the delivered lower modes — hum, prime, tierce and quint — imply. C4 (hidden A♭, off it): 333333333SS; D4 (hidden B♭, off it): 3333333333S; E4 (hidden C, on the scale): 33333333333; F4 (hidden D♭, off it): 33333333333; G4 (hidden E♭, off it): 33333333333; A4 (hidden F, on the scale): 33333333333; B4 (hidden G, on the scale): 33333333333; C5 (hidden A♭, off it): 33333333333 (3 the note a major third below the strike, S the strike's own name, a dot too few lower modes delivered). 33 strokes imply a note on the ring's scale and 52 a note off it. Perception and the listener

A ring of minor bells rings a second scale a major third below

A minor-tierce bell's four lowest modes, read as a residue, imply a note a major third below the bell's strike note. Put eight such bells in a ring tuned to a major scale and the implied notes are that scale moved down a major third — a ring in C carries the scale of A♭ major underneath it. Only three of the eight land on notes the ring rings, from the mediant, the submediant and the leading note; the tonic, the octave, the supertonic, the subdominant and the dominant all imply notes the ring never sounds, each a semitone from two that it does. Above a tenor on C3 the lower modes deliver the implied note at nearly every stroke. A reading that weights the whole spectrum still hears none of it.

 Felt on every landmark, the stream recovers most of a quiet room's recognition even when no accent is felt. Ten dances known, entered at a random step, the crowd covering half the time at the level on the axis above the band: the share of entries in which the dance is recognised within the present, for four ways of carrying the landmarks — a drum on the downbeat and the melody's softer strokes between, heard; the same with the downbeat felt as well; a kick drum on every landmark eight decibels softer than the downbeat, heard; and the same felt, with a felt accent recognised 0 per cent of the time. A quiet room gives 80.9. drum downbeat, melody between: 0 dB 58.4, 5 dB 41.0, 10 dB 37.4, 15 dB 37.3, 20 dB 34.9, 25 dB 30.4, 30 dB 28.9; the downbeat felt: 0 dB 58.4, 5 dB 41.0, 10 dB 37.4, 15 dB 37.3, 20 dB 37.3, 25 dB 37.3, 30 dB 37.3; a kick on every landmark, heard: 0 dB 80.9, 5 dB 80.9, 10 dB 68.1, 15 dB 50.2, 20 dB 35.7, 25 dB 30.4, 30 dB 28.9; a kick on every landmark, felt: 0 dB 80.9, 5 dB 80.9, 10 dB 80.9, 15 dB 80.9, 20 dB 75.8, 25 dB 71.6, 30 dB 70.1. Rhythm and metre

The felt channel carries the inner beats, not the downbeat

A low drum felt in the body gets the downbeat through any crowd, and is then worth less than it was in a quiet room, because the crowd has taken the soft strokes between. Put a kick drum on every landmark of the dance and let all of it be felt, and in a crowd thirty decibels over the band a late dancer recognises the dance in seventy per cent of entries against thirty-seven for the felt downbeat — with no felt accent at all. The accent, felt or not, is worth about ten points on top. The worry that a stream felt alike has no downbeat turns out to be the small half of the arithmetic.

 Over C4, width and disagreement trade one for one only while the lock is tight; loosely locked, a narrow vibrato stays in step far longer. The tuned close seventh over C4, exactly in tune, a vibrato on every note, its width chosen so that the width times how far short of complete the lock falls is the value on the axis: the share of eight seconds the chord spends within this register's onset tolerance, over eight draws, for four locks. If width and disagreement entered only as their product the four curves would coincide. 99.5%: 0.071 cents 100% (width 14.2), 0.142 cents 92% (width 28.4), 0.355 cents 62% (width 71), 0.71 cents 38% (width 142), 1.42 cents 20% (width 284), 2.84 cents 8% (width 568); 98%: 0.071 cents 100% (width 3.55), 0.142 cents 93% (width 7.1), 0.355 cents 61% (width 17.75), 0.71 cents 37% (width 35.5), 1.42 cents 19% (width 71), 2.84 cents 7% (width 142); 90%: 0.071 cents 100% (width 0.71), 0.142 cents 100% (width 1.42), 0.355 cents 72% (width 3.55), 0.71 cents 32% (width 7.1), 1.42 cents 15% (width 14.2), 2.84 cents 6% (width 28.4); 50%: 0.071 cents 100% (width 0.14), 0.142 cents 100% (width 0.28), 0.355 cents 100% (width 0.71), 0.71 cents 100% (width 1.42), 1.42 cents 65% (width 2.84), 2.84 cents 15% (width 5.68). Timbre and acoustics

A narrow vibrato needs no lock

A tuned close seventh binds and releases at the vibrato's swings only where its notes' timing swings by about the width of its onset tolerance, and the swing is the vibrato's width times how far the singers' vibratos disagree. Down to a ninety-five per cent lock the two trade one for one: a 71-cent vibrato locked ninety-nine and a half per cent is the same chord as a 7-cent one locked ninety-five. Below that the trade fails, in the singers' favour. With no lock at all the swing stops growing with the disagreement, and a vibrato of three cents over middle C — twelve over the bass — binds the chord at most of its swings with each singer's vibrato its own.

 Slowing every pair by one factor smooths the shimmer steadily and leaves the goal-tone clock alone until a fifth of the promised speed. A gamelan's paired instruments with the middle octave's rates repeated in every octave, the rate rising as the square of the pitch across the degrees, and every rate multiplied by the factor on the axis; a cycle of 160 seconds, sixteen skeleton steps of 10.0 seconds, with a swell of beating in the step before each of four goal strokes. The roughness the pairs add, as a share of the promised shimmer's 3.06 (the falling line), and the share of the swells' value to the clock that survives (the other line): the skeleton alone leaves 0.646 bits of position unknown and the swells at the promised shimmer 0.146. 0.1: clock 49%, roughness 15%; 0.12: clock 84%, roughness 18%; 0.14: clock 84%, roughness 21%; 0.16: clock 84%, roughness 24%; 0.18: clock 84%, roughness 27%; 0.2: clock 84%, roughness 29%; 0.21: clock 100%, roughness 31%; 0.22: clock 100%, roughness 32%; 0.25: clock 100%, roughness 36%; 0.3: clock 100%, roughness 42%; 0.35: clock 100%, roughness 48%; 0.4: clock 100%, roughness 53%; 0.5: clock 100%, roughness 63%; 0.6: clock 100%, roughness 72%; 0.7: clock 100%, roughness 80%; 0.8: clock 100%, roughness 87%; 0.9: clock 100%, roughness 94%; 1: clock 100%, roughness 100%; 1.1: clock 100%, roughness 105%; 1.2: clock 100%, roughness 110%. The clock first loses at 0.20; the floor is 0.21. Scales and modes

A gamelan's shimmer can slow to a fifth before its clock notices

Roughness falls steadily as every pair's beating is slowed by one factor, and the listener who times a goal's swell does not object until the shimmer is a fifth of the promised speed. The limit is not how many beats a swell holds — a ten-second step holds eight even there — but the once-a-second boundary below which a beat is waited out rather than counted, and it bites when the second-slowest goal degree crosses it. At the quick tempi where a swell could hold too few beats, the skeleton already carries the clock alone.

 A goal degree that returns is better marked once than twice, and best told apart. A cycle of 160 seconds, sixteen skeleton steps, a swell of beating in the step before each of four goal strokes; two of the four goals share a slendro degree and so a beat rate, at every pair of positions. Bits of position left unknown with both of that degree's swells, with the first or the second silenced, with neither, and with the two told apart by some other mark. steps 3 and 7: both 0.034, first silenced 0.049, second 0.062, neither 0.143, told apart 0.003; steps 3 and 11: both 0.146, first silenced 0.049, second 0.082, neither 0.128, told apart 0.003; steps 3 and 15: both 0.046, first silenced 0.049, second 0.067, neither 0.126, told apart 0.003; steps 7 and 11: both 0.066, first silenced 0.062, second 0.082, neither 0.202, told apart 0.003; steps 7 and 15: both 0.101, first silenced 0.062, second 0.067, neither 0.127, told apart 0.003; steps 11 and 15: both 0.077, first silenced 0.082, second 0.067, neither 0.188, told apart 0.003. Scales and modes

A goal tone that returns should swell only once

When two of a gamelan cycle's four goal strokes land on the same degree, their swells of beating sound alike, and a listener who hears one cannot tell which goal is coming. Across all 625 ways five degrees can fill four goals, 365 are served better by an elaboration that anticipates fewer goals than four. The repeated pair costs most at slow cycles, can cost more than every swell is worth at moderate ones, and is spared only when its first swell comes straight after the gong.

 A cadence that votes as a witness of its own moves the line from an emptied reading to a fiftieth of one. The choosing listener, every root late, a cadence before bar 9 leaving the share on the axis of the reading's lead (logarithmic, "none" an emptied reading); at the cadence its arrival casts one vote for the barline, heard as the tune's long note, as the drum, or as a witness of its own, or is silent; then 1 fill stroke heard as the drum, then the honest drum: the share of passages ending on the barline. a silent cadence: 100% 0%, 50% 0%, 30% 0%, 20% 0%, 15% 0%, 12% 0%, 10% 0%, 7% 0%, 5% 0%, 3% 0%, 2% 0%, 1% 0%, 0% 6%. heard as the tune: 100% 0%, 50% 0%, 30% 0%, 20% 0%, 15% 0%, 12% 0%, 10% 0%, 7% 0%, 5% 1%, 3% 1%, 2% 1%, 1% 10%, 0% 11%. heard as the drum: 100% 0%, 50% 0%, 30% 0%, 20% 0%, 15% 0%, 12% 0%, 10% 0%, 7% 0%, 5% 1%, 3% 1%, 2% 1%, 1% 2%, 0% 100%. heard as its own witness: 100% 0%, 50% 0%, 30% 0%, 20% 0%, 15% 0%, 12% 0%, 10% 0%, 7% 0%, 5% 1%, 3% 51%, 2% 90%, 1% 93%, 0% 93%. Harmony and voice leading

A cadence that votes must not vote as the drum

A cadence lands on a downbeat, so it could cast a vote for the barline at the moment it loosens the listener's grip. Whether that vote helps depends on whose it is taken to be. Heard as the tune's long note, it speaks with credit the tune spent long ago and is worth almost nothing. Heard as the drum, it is exactly one more stroke of fill, and is convicted as the fill's first stroke was. Heard as a witness of its own, it carries a third of a vote and no record to lose — and that moves the line from an emptied reading to a fiftieth of one.

 A straight tone's slow wander drifts a chord's notes out of step and leaves them there. The tuned close seventh over C4 held without vibrato, each note's pitch wandering on its own by 0.5 cent rms with its spectrum falling above two hertz: how far ahead of the bass each upper note's timing has run, in milliseconds, over eight seconds of one draw, with this register's onset tolerance of ±0.070 ms shaded. Over eight draws the chord is in step 4.3 per cent of the time and first leaves the tolerance after 0.16 s on average; a three-cent vibrato on every note, each its own, keeps it in step 48.2 per cent. Timbre and acoustics

A straight tone drifts out of step and stays there

A tuned close seventh binds — its filters flutter together — only while its notes keep their timing within a fraction of a millisecond, and a narrow vibrato keeps them there because a sinusoid's timing offset swings inside a fixed range. A straight-tone singer's pitch is never quite straight: it wanders slowly, a cent or less, at random. That wander's timing offset is a random walk. Over C4 a wander of half a cent takes the chord out of step within a fifth of a second and leaves it in step four per cent of the time, where a three-cent vibrato holds it nearly half. What matters is the wander's slow part, and adding a vibrato on top never helps.

 Struck, a suspension's roughness falls whether or not it resolves, and the step itself sends it back up. One 4–3: C5 prepared over a C chord, tied over G3, D4 and G4, and stepping to B4, each moment 1 s. Roughness, Plomp–Levelt over eight partials of every pair weighted by the product of their amplitudes. Sustained: suspension 0.327, resolution 0.325. Struck, with a T60 of 12 s at middle C: suspension 0.149, resolution 0.054, and with the note left hanging 0.024 at the same moment. Harmony and voice leading

A tied note is the oldest thing in its own dissonance

A suspension is taught as a dissonance that resolves, and the sum is done on a chord symbol with no time in it. Put the time back. On voices or an organ, a 4–3 is no smoother after its step than before: of eight spacings, two fall and four rise. On a piano the tied note was struck a chord ago, so it supplies under a third of the roughness it is blamed for, the fall to the resolution is the clock's, and the step itself makes the sound about three times rougher than leaving the note where it was.

 A loud, slow tune spends a singer's whole breath before each phrase ends, and the extra breaths have to go inside the phrases. Ode to Joy from D4 at crotchet 40 and 90 dB, sung by a trained singer with 2.8 litres of usable air and a flow set by each note's pitch and level. The line is the air left. The plan is the cheapest that reaches the end, with breaths at notated phrase ends costing nothing: 6 breaths, 4 of them inside a phrase. The tune lasts 48 s. Form and structure

A singer needs a quarter of the breaths the page offers

A breath in a sung phrase looks like evidence of a boundary, and also like a physical necessity, and it cannot be both at once. Solved exactly over three tunes, a trained singer is forced to breathe inside a phrase on 8 settings of 147, all of them slow and loud; at a comfortable level the air asks for about a quarter of the phrase ends a tune offers. So nearly every breath is a choice, which is what makes it a cue. Where the air does force one, it lands on a listener's boundary no more often than chance.

 Letting songs move a tone adds between five and twenty points at any count of fifths, and never makes the trade disappear. Twelve minor songs, three each in E, A, D and B minor, with three capo changes and no key twice running, searched exactly. For each number of neighbouring songs a fifth apart, the largest share of a moving capo's saving, with every song in its own key and with up to 12 moved up or down a tone. 0: 84% and 92%; 1: 84% and 92%; 2: 84% and 92%; 3: 84% and 92%; 4: 78% and 83%; 5: 70% and 75%; 6: 63% and 75%; 7: 61% and 75%; 8: 54% and 66%; 9: 46% and 58%; 10: 40% and 57%; 11: 37% and 57%. Pitch and tuning

A set free to change key becomes a set in two

A set list orders its songs for the music and a capo wants them ordered by fret, and the trade between the two looked as if it might disappear if a singer would move a song a tone. It does not. Searched exactly, moving songs adds five to twenty points of the capo's saving at any count of fifths and stops improving at six moved; at its limit the best set rocks between B minor and E minor, because a count of fifths rewards a pendulum. A set that really walks the chain of fifths has three to spend, not eleven.

 Up every wind instrument's compass the partials draw together, so whether a note parts depends mostly on how high it is. Five modelled bores, every semitone of a playing compass: how far apart the first and last of a note's partials reach 50 per cent of their steady amplitude, against a threshold of twenty milliseconds for hearing a partial out. trumpet: 165 Hz 26 ms, 262 Hz 21 ms, 416 Hz 16 ms, 660 Hz 10 ms; horn: 87 Hz 36 ms, 138 Hz 29 ms, 219 Hz 24 ms, 348 Hz 15 ms, 552 Hz 9 ms; trombone: 82 Hz 33 ms, 130 Hz 36 ms, 207 Hz 30 ms, 328 Hz 21 ms, 521 Hz 14 ms; clarinet: 147 Hz 33 ms, 233 Hz 24 ms, 370 Hz 16 ms, 588 Hz 11 ms, 933 Hz 0 ms, 1482 Hz 0 ms; alto saxophone: 139 Hz 43 ms, 221 Hz 45 ms, 350 Hz 26 ms, 556 Hz 13 ms. Below twenty milliseconds from: trumpet 311 Hz, horn 293 Hz, trombone 368 Hz, clarinet 311 Hz, alto saxophone 441 Hz. Perception and the listener

A blown note parts by its register, not its family

A wind note's partials reach their steady levels at different rates, and one ranking of five instruments had the saxophone's parting soonest and the horn's latest. Run at every note rather than one, that ranking dissolves: it was read at pitches from 126 to 428 hertz, and pitch was doing most of the sorting. Moving up one instrument shifts the point of parting three times as far as changing instrument at one pitch, and read at half of steady level, all five come together within a fifth of each other, between about 290 and 440 hertz.

 The same kind of walk covers more of the keyboard on a scale with larger steps. A walk of thirty notes with the tunes' distribution of steps in scale degrees, pulled back towards its start, on the major scale. This seeded walk uses 8 different notes and spans 12 semitones. Over three thousand such walks the mean is 11.4 semitones, 6.6 degrees and 7.4 different notes. Scales and modes

A tune's range is a number of notes, not an interval

An earlier essay derived the width of a tune from a walk pulled back towards its middle, and read it in semitones. That reading has a scale inside it. Put the same walk on five notes, seven and twelve, and at thirty notes it covers 16, 11 and 7 semitones, while using the same seven or so different pitches on all three. The width a tune inherits from its walk is a count of notes, which an interval then measures. And on a pentatonic scale, the taught rule about leaps cannot even be stated in semitones.

 A section whose upper partials beam keeps its pizzicato deep into a hall on its axis, and loses it faster than an unpointing one a little way off. The margin at the strike for a pizzicato section of twelve over a held clarinet, 15 m into a hall, from C4 to F♯5, with each partial's direct sound scaled as from a piston of 10 cm radius. Above zero the pluck owns the note. no beam: A♭4 to C♯5 -0.3, -1.7, -3.4, -3.5, -3.3, -1.7; on the axis: A♭4 to C♯5 5.2, 4.1, 2.8, 3.0, 3.5, 5.3; 15° off: A♭4 to C♯5 4.4, 3.2, 1.8, 1.8, 2.2, 3.9; 30° off: A♭4 to C♯5 1.9, 0.3, -1.6, -2.2, -5.1, -2.1. Timbre and acoustics

A section keeps its pizzicato only inside a cone

In a hall a section's pizzicato loses its corpus notes to a held clarinet within two metres, on the assumption that a violin radiates equally in every direction. It does not, and twelve players sharing a beam cannot average it away. Directly on the players' axis, the same section keeps every note for twenty-five metres. Thirty degrees off, with the beam's upper partials thrown elsewhere, it is worse than if it had never pointed. A question about how far back a pizzicato carries turns out to be a question about which way the listener's seat faces.

 A mark does little until it is nearly always heard, and then it does everything, whichever stroke was moved. The standard bell pattern with a varied cycle every 3, both pulses heard half the time, a listener who learns how often strokes are lost: the surprise a step once placed, averaged over the second to sixth cycles, against how reliably each stroke's mark is heard. The plain bell reads 0.0078 and the other order of the gaps 0.0016. the bell shifted: 50% 0.0252, 60% 0.0238, 70% 0.0204, 80% 0.0154, 90% 0.0092, 95% 0.0062, 100% 0.0027; a new figure: 50% 0.0055, 60% 0.0055, 70% 0.0052, 80% 0.0048, 90% 0.0041, 95% 0.0036, 100% 0.0025; nothing moved: 50% 0.0076, 60% 0.0076, 70% 0.0075, 80% 0.0070, 90% 0.0057, 95% 0.0047, 100% 0.0026. Rhythm and metre

A bell can mark a cycle without moving a stroke

A lead drummer's varied cycle helps a listener or misleads one, according to which stroke it moves, because a moved stroke can make the bell look as if it had started somewhere else. Give the varied cycle a different sound as well and the distinction goes. Heard nine times in ten and a misleading variation is no longer one; heard always, it no longer matters which stroke moved, or whether any did. A mark heard seven times in ten does almost nothing, and the arithmetic of that threshold is the same as a coin's.

 A trill between a suspension and its resolution stays rougher than the chord struck once, all the way through. One 4–3 on a struck instrument, T60 12 s at middle C: C5 over G3, D4 and G4 for the one second of the suspended chord. Mean roughness over the chord: tied 0.077, struck once 0.119, trill to the upper neighbour D5 at 8 notes a second 0.108, trill to the resolution B4 0.141. Trills are read per whole cycle of two notes. Harmony and voice leading

A trill is half a consonance unless it borrows its own resolution

A struck suspension loses roughness because its tied note is old, and the obvious repairs are all things a player already has. Three of them fail in ways nobody would predict. A re-strike under about nine decibels below the chord is smoother than the tie it replaces. No rate of trill stops the fading, because the three notes under the trill are what is fading. And a trill on the upper neighbour spends half its time on a chord note: only the trill that alternates a suspension with its own resolution gives back more than a single strike.

 A vibraphone's motor makes a note quieter and far longer, with a tremolo that is shallower than the valve's two states at middle C. The power a struck bar and its tube radiate at middle C, in decibels, with the coupling switched at 6 openings a second. A tube's filling time here is 98 ms. With the bar holding the same energy, the open valve radiates 13.8 dB more than the shut one; with the motor running the tremolo's depth is 10.0 dB. Instruments and their design

A vibraphone's valve is a door on the sound and a drain on the bar

A vibraphone's motor is usually described as a duty cycle: loud while the valve is open and quiet while it is shut, with a tremolo as deep as the gap between them. Put the tube's own time back and the description comes apart. The door and the drain obey different laws. A low note's tremolo is a third shallower than a high note's at the same speed, the note rings almost three times longer than with the valve left open and well beyond any mean of the two states, and the same stroke can land eight decibels apart according to where the motor happens to be.

 On an unequal scale the two meanings of a melody's pull come apart at particular notes, and they do not do it evenly. The hirajōshi scale, with the walk's centre on its first note and both pulls fitted to a post-skip reversal of 60 per cent: the chance of moving inwards at each degree. Pull in semitones κ 0.715, pull in degrees κ 0.849. It takes about 337 notes of melody to tell the two at a likelihood ratio of 99. Scales and modes

A melody's pull can be counted only where a step is a skip

A melody drifts back towards its centre, and nobody has said in what unit: a distance on the keyboard or a distance in the scale. On even scales the two cannot be told apart. On unequal ones they can, and an exact reckoning puts a price on it: about five hundred thirty-note tunes on the major scale, a dozen on hirajōshi. The price falls with the square of how uneven the scale is, and then drops seventeenfold at the point where a step of the scale gets wide enough to be counted as a skip.

 Antiphonal second violins own their worst pizzicato note almost nowhere, unless a hard wall stands close behind them. A plan of a 24 by 36 m hall with the second violins at the right of the stage, plates turned away from the audience, and a perfectly reflecting wall 2 m behind them. A seat is shaded if the section's worst note from A♭4 to C♯5 stays a pizzicato rather than a clarinet with a click. The seconds keep 30 per cent of seats, the firsts on the other side 30. Timbre and acoustics

A wall puts antiphonal seconds back beside the firsts

Second violins seated opposite the firsts have their plates turned away from the audience, and in a hall a section's pizzicato depends on its beam. With nothing behind them they keep their worst note in about one seat in a hundred. A perfectly hard wall two metres behind gives them back a cone as large as the firsts' own, but a wall that loses three decibels takes most of it away again. And the seats that cone serves turn out to be, within a seat or two, the seats the same section would serve sitting beside the firsts.

All themes