Colour Science & Perception
what the eye settles, and what it cannot
Theme

The thread: Computed, not quoted — page 26

Every swatch begins as a spectral power distribution and is carried through the colour-matching functions as it is drawn. None is a hex code recalled from a table.
Scaled by the height, the two rooms close their windows together. The deepest change in the painted floor's chroma over each length's solved shorter sides, against the floor's length over the room's height, for boxes 2.5 and 4 units tall. The 2.5-tall box's dip crosses nought at 2.59 units, 1.04 times its height; the 4-tall box's at 4.04, 1.01 times. What a scene does

A taller room keeps its floor losing in proportion

A satin finish on a room's grey faces nearly always adds a little colour to a painted floor, and in a box 2.5 units tall it takes some away only in a pocket of proportions: floors shorter than about 2.6 units, with a shorter side near 0.4. The prediction was that the pocket scales with the room's height — that a box four units tall keeps losing until its floor is about 4.2 units long, 1.04 times its height as in the lower box. It loses until 4.04 units, 1.01 times its height: the crossing scales, a little short of the prediction. The pocket moves too. At a shorter side of 0.4, where the lower box lost, the taller box's floor gains at every length; its losses sit at 0.6 — at three and four units long — and at 0.8 when three units long, and its deepest loss lies near a sixth of its height, as the lower box's did. Every box is still all or nothing across seventy-two paints, and every box's change carries the sign of its finish's tilt across reflectance.

 Bounded, one session stops finding a macula that never moved. For the three bands with the macula unmoved, how many of twenty one-session experiments report it moved beyond twice its standard error: from the census's two starts, from the best of five starts, and from the best of five with the shift held within ±6 nm. The bound brings the false shifts at 415 and 430 nm from 8 and 6 to 1 and 2. What the eye does

A bound on the macula trades false moves for hidden ones

Fitted thoroughly, one session of an older observer's colour matches finds a second valley in a quarter of its experiments — the macula's peak moved ten or more nanometres from the model's, where no real macula sits — and reports a moved macula where there is none in up to eight experiments of twenty. The prediction was that holding the macula's shift within six nanometres either way, the range real pigment mixtures span, restores every observer's one-session counts to within two of the census's and leaves about one fit in four pressed against the bound. The false moves vanish: an unmoved macula is reported moved in one, two and one experiments of twenty for the three bands, against eight, six and two unbounded. Seven of nine observers keep their count of moved maculas within two of the census's. But only one fit in six ends on the bound, and beside the band at 445 nm the bound hides real moves: a macula truly three nanometres short is found in three experiments of twenty, where the census found seven. Two sessions barely notice the bound.

 The prior trades the macula's moves for the lens band. For each observer, how many more experiments of twenty find the lens band under the prior than under the census's fits, against how many more find the macula moved. The observers moved short or unmoved gain up to seven bands and lose one to five moves — false ones where the macula never moved, real ones where it did; those moved long gain no band and lose one or two moves, the 445 nm observer losing three bands as well. What the eye does

A soft prior on the macula frees the lens band

One session of an older observer's colour matches, fitted thoroughly, can drift into a valley where the macula's peak sits ten or more nanometres from any real eye's, and a hard bound at ±6 nm removes those false moves at the cost of hiding some real ones — beside a lens band at 445 nm, a macula truly three nanometres short is found in three experiments of twenty where the census found seven. The prediction was that a Gaussian prior four nanometres wide, which lets a fit go further when the matches insist, keeps the false moves away and restores those real ones. It keeps the false moves away — none, one and one for the three bands. It does not restore the real moves: it finds them as seldom as the bound or more seldom, two of twenty beside 445 nm, because it pulls every fitted shift towards nought, to a mean of 1.2 to 1.9 nanometres for moves of three. What it buys instead is the lens band. With the macula held near the standard, the band beside it is found in ten to twelve experiments of twenty, where the census found three to seven.

 A sixth gained above L* 20, most of the cure lost below it. For the eight cases binding above L 26, the dulling below L 20 and between the binding point and L* 20, with the cap held, eased straight and eased on the bent ramp. The bent ramp cuts the upper stretch from 6.18 to 5.20 thousandths and raises the deep shadow from 3.85 to 7.16. What it takes to deliver it

A bent ramp takes more ink and dulls more

Easing a four-ink table's total-ink cap by a fifth on a straight ramp from L* 20 to the floor straightens the deep shadow's bend and halves the dulling of dark colours, and a ramp started higher trades the deep shadow's cure for the upper stretch's. The prediction for a ramp bent at L* 20 — a quarter of the ease spent between L* 32 and 20, the rest below — was that it keeps nearly all of the deep shadow's cure while taking a third off the high-binding colours' upper stretch, their total falling by a fifth. The bent ramp takes at least as much ink as the straight one at every lightness. It dulls every group more. The high binders' upper stretch improves by a sixth; their deep shadow loses half its cure, and their total rises from 0.0052 to 0.0060; the brown's low binders, left with almost nothing by the straight ramp, keep 0.0039. The deep shadow's cure follows how steeply the cap falls below L* 20 — a fifth of the cap over six units, not three twentieths — and not how much ink is taken there.

 Two rooms' floors as long as the room is tall, read as one row. The mean change in the painted floor's chroma over seventy-two paints under a satin finish, against the floor's shorter side as a share of the room's height, for the floor four units long in the box four units tall and two and a half long in the box 2.5 tall. Where the two rooms were solved at the same proportions their points coincide. Read together the floor loses from 0.148 to 0.176 of the height, the dashed lines; the deepest loss is 0.0056% at 0.16. What a scene does

A room scaled whole loses its floor to the last digit

A satin finish on a room's grey faces costs a painted floor a little of its chroma only in a pocket of proportions, and a box four units tall was found losing in proportion to a box 2.5 tall — its crossing at 1.01 times its height against the lower box's 1.04, its pocket near a sixth of the height in both. The prediction was that the answer is the room's proportions alone: that the four-tall box built as the 2.5-tall box multiplied by 1.6 gives the lower box's loss to the solver's precision, and that the floor four units long loses from about 0.55 to 0.65 wide. The scaled box gives the lower box's numbers to eleven significant figures, and a pair the census had already solved agrees to twelve. The window sits higher than predicted at both edges, from 0.59 to 0.70. And the gap between the two rooms' crossings was the census's sampling: read at the pocket, the taller room crosses at 1.029 times its height against 1.036.

 The curve holds inside the bin and fails outside it. Four error lines learned on eighty-nine white LEDs at a Duv of nought and ±0.003, each read at those three values with every LED left out and at ±0.0015 and ±0.006 on 119 LEDs none of them was fitted on. The slope and curve per temperature reads ±0.045 and ±0.043 inside the bin and ±0.153 and ±0.161 at twice its edge, where a slope per temperature alone reads ±0.093 and ±0.110. What a camera does

The bin's curve holds between its points and breaks beyond them

A camera's colour error under a white LED was read from an ambient sensor across a manufacturing bin by a line with a slope and a curve in Duv for each colour temperature, fitted at three Duv values — on the locus and 0.003 either side — and reading the LEDs beside the locus within ±0.056 and ±0.048 ΔE₀₀. A curve through three points passes through them whatever happens between them. The prediction was that 119 new LEDs built at ±0.0015 are read within ±0.06, and those at ±0.006 worse than ±0.10. Both halves hold: ±0.045 and ±0.043 inside, ±0.153 and ±0.161 outside. But the error's own mean course stays on its parabola within 0.02 out to ±0.006. The line's bend is not the error's: it is the gap the line leaves between the locus LEDs and the bin's edges, and a gap spent as a curvature quadruples at twice the distance.

 With the level placed, the line bends as the error does. At each temperature, the curvature in Duv of the camera error's mean course through nought and ±0.003, beside the curvature each slope-and-curve line implies, averaged over the three pumps. At 2,700 K the course bends by +7,916, the line without a level by +2,803 and the line with one by +8,198; at 4,000 K the course bends by +2,196, the line without a level by +14,591 and the line with one by +2,195; at 6,500 K the course bends by +363, the line without a level by −13,349 and the line with one by +445. What a camera does

A level for each temperature lets the curve be the error's

An error line with a slope and a curve in Duv for each colour temperature read white LEDs inside a manufacturing bin within ±0.045 ΔE₀₀ and LEDs at twice the bin's edge only within ±0.16, though the camera's own error stays on its parabola there. Its curve was borrowed: the line's shared terms could not place each temperature's level, and the square of the Duv absorbed the gap. The prediction was that one more number a pump, the square of the mireds, brings the locus LEDs within ±0.05 and those at ±0.006 within ±0.08. It brings the locus to ±0.025 — three times better than the locus-only line that started this run of essays — and the outer LEDs to ±0.049 and ±0.058. The curvature the line now implies at each temperature is the error's own within a fifth; what it leaves is the warm whites' scatter at the bin's edges.

 Each widening of the prior buys moves with bands. Over the six observers whose macula truly moved three nanometres, the real moves one session finds beyond twice their error, against the lens bands found for the four observers whose band the four-nanometre prior freed — under priors four, six and eight nanometres wide, the census's own fits and the hard bound. 4 nm: 14 moves, 44 bands; 6 nm: 22 moves, 34 bands; 8 nm: 25 moves, 25 bands; census 27 and 22; bound 19 and 20. What the eye does

A wider prior gives the moves back by spending the band

One session of an older observer's colour matches, fitted under a Gaussian prior four nanometres wide on the macula's peak shift, finds the lens band beside the macula in ten to twelve experiments of twenty and real three-nanometre moves of the macula as seldom as a hard bound does. The prediction was that a prior six nanometres wide keeps the false moves at two or fewer, finds the real moves within two of the bound's counts and keeps two thirds of the band's gain, and that at eight the far valley reopens at 415 nm. Three parts hold. The fourth fails: six nanometres keeps 55 per cent of the band's gain. Read across the widths the prior traces one trade — 14 real moves and 44 bands at four nanometres, 22 and 34 at six, 25 and 25 at eight, against the unconstrained fit's 27 and 22 — and the hard bound, at 19 and 20, is worse than six nanometres on both.

 A narrower flank keeps most of the Lorentzian's losses. The forty-nanometre band pigment with its steep side fourteen, ten, seven and five nanometres wide at half depth: how many of the clear mercury lamp's 25 splitting directions it loses by more than one per cent, as a Gaussian and as a pseudo-Voigt with a twentieth of Lorentzian, beside the counts predicted. The Gaussian loses 4, 0, 0, 0; the Voigt 18, 13, 12, 7, where four were predicted at seven nanometres and none at five. Where the model breaks

A narrower flank moves its wing in with it

Under a clear mercury lamp, a band pigment whose steep side is a Gaussian fourteen nanometres wide loses four of the twenty-five directions that split the green line from the yellow, and with a twentieth of Lorentzian in that side it loses eighteen, because the wing three half-widths out out-absorbs the whole Gaussian there. A Lorentzian wing falls as the square of the distance over its half-width, and the prediction was that narrowing the side to seven nanometres brings the twentieth-Lorentzian pigment back to four losses and to five nanometres to none. It loses thirteen at ten nanometres, twelve at seven and seven at five, while the Gaussian loses nothing at all three. The wing still decides: the narrower band does not stay where the wider one sat but moves fourteen nanometres in towards the line it must pass, so its wing there falls fourfold rather than eightfold.

 The census line is a chord across a curve. For four pigments, the cancelling strength less the census line's at each sheet's glow — the light, office and premium sheets and the three bleached and yellowed premium sheets — in order of glow. The light and premium sheets sit above their own line and the office sheet under it, the orange by −0.0161, deeper than any bleached sheet: the line through the three census sheets is a chord across a sagging curve, and the bleached sheets sag beneath it as the office sheet does. What the brain does

A yellowing base leaves the ink where the glow put it

A light ink cancels a pale tint's residual hue turn on brightened paper at one strength for each pigment, and as a premium sheet's brightener bleaches that strength rises back along the straight line in glow through three census sheets — a shade under it, by up to 0.013. Real photo-ageing yellows the base paper as well, and the prediction was that a base yellowed in step lifts the strengths onto the line within 0.01 and that the office sheet's ink still holds every residue within one and a half degrees. The ink holds, at 1.38°, 1.13° and 1.38°. The strengths do not lift: no pigment's moves by more than 0.0023, and its gap under the line by more than 0.0006. The yellowing raises the sheet's b* by 0.31 at most while the bleaching raises it by five. And the gap was never the base's: the office sheet itself sits 0.016 under the line, deeper than any bleached sheet. The line is a chord across a sagging curve.

 Even two sessions fall into the far valley, and the prior lifts them out. For the observer with its band at 415 nm and its macula truly three nanometres short, every two-session experiment's fitted shift at the best of five starts and under a prior six nanometres wide. Three best fits lie beyond 7.5 nm short — −11.84 drawn to −1.90, −7.73 drawn to −3.05, −11.00 drawn to −0.94 — and the rest move by a tenth or two; the mean goes from −3.89 to −2.57. What the eye does

A second session leaves the prior one job

Under one session of an older observer's colour matches, a prior on the macular pigment's peak shift trades the lens band against real moves of the macula, six nanometres wide at the knee of the trade. Two sessions already find the band in twelve to twenty experiments of twenty and real three-nanometre moves in twelve to eighteen, and still report an unmoved macula moved in two to three. The prediction was that the six-nanometre prior, applied to two sessions, removes those false moves and costs nothing else. It removes none of them — two, two and none remain, about the rate at which a test at twice the standard error calls a move by chance. It costs nothing at five observers of six. Beside 415 nm, with the macula moved short, it finds four fewer moves: three of the best fits there were in the far valley, eight to twelve nanometres short, and counted as moves because their sign was right. With two sessions the far valley is the prior's only remaining job.

 Per channel, a tenth of green lands where a half was graded. Each mixture's CIECAM16 hue quadrature on the master and through the per-channel curve, between unique green at 200 and unique blue at 300. On the master a tenth of green sits at 312.9 and a half at 277.5; per channel on the 100 cd/m² display the tenth lands at 260.9, past the master's half. Matching and measuring

An appearance model sees the per-channel sky turn twice as far

BT.2390's tone curve, applied to each channel of a 1,000 cd/m² master shown on a 100 cd/m² display, raises a sky-blue's green share from a tenth to 69 per cent of its blue and turns its ICtCp hue 23°, where the same curve applied to the largest channel with all three scaled keeps the hue within 3°. ICtCp's hue is not an appearance, and the prediction was that CIECAM16, read under the display's own white, still shows the per-channel turn as more than twenty hue-quadrature units and more than ten ΔEITP, the scaled curve under three of each. The appearance model sees it as fifty-two units, more than half the way from unique blue to unique green, and its lightness more than doubles. The scaled curve keeps hue and lightness to a quarter of a unit. And ΔEITP between master and display ranks the two curves backwards: in absolute light the per-channel sky is the nearer one.

 A blue's intensity sits below where the curve compresses. BT.2390's curve from the 1,000 cd/m² master to displays of 100 and 300 cd/m², with the level a pure blue, a tenth-green and a half-green sky present to it when the curve is applied to ICtCp intensity, and the level their blue channel presents when it is applied per channel or to the largest channel. The skies' intensity corresponds to 79 to 387 cd/m², against knees of 27.9 and 161.4; their blue channel sits at 1,000, the top of the curve. Matching and measuring

A curve on intensity cannot see a blue's brightness

ITU-R BT.2390's tone curve applied to each channel of a bright master turns a tone-mapped sky's appearance fifty-two hue-quadrature units towards green on a 100 cd/m² display; applied to the largest channel with all three scaled, it keeps hue and lightness to a quarter of a unit. BT.2390 also describes applying the curve to ICtCp's intensity and scaling its chroma, which keeps ICtCp's hue by construction, and the prediction was that it keeps the appearance's hue within three units and its lightness within two. It keeps the hue, within 2.3 units. It cannot keep anything on the display: a saturated blue's intensity is the PQ level of 79 cd/m², under the 300 cd/m² display's knee, so the curve barely touches it and every sky asks for two to seven times the blue the display has. Clipped channel by channel, the sky turns as far as the per-channel curve turned it; scaled back into the gamut, it lands where the scaled curve put it. The step back into the gamut decides the appearance, not the curve.

 Seven channels keep every light within a unit through the noise. Every light of the enlarged census read a hundred times by sensors of three, five, seven and twelve channels whose signals each carry independent noise of a quarter, a half and one per cent, each light by the matrix fitted without it: how many lights are read beyond a ΔEITP unit at least once. 3 channels: 40, 42, 46; 5 channels: 7, 9, 22; 7 channels: 0, 0, 0; 12 channels: 2, 2, 2. Difference and uniformity

Seven channels carry their noise as lightly as three

An ambient-light sensor of seven narrow channels, fitted to a room's forty-seven lights and four triphosphor siblings, reads every one of the fifty-one within a ΔEITP unit of a dim moon picture's tint, each light held out of its own fit. A fine sensor was expected to read chromaticity from small differences between neighbouring channels, with large weights of both signs that multiply every channel's noise. The prediction was that seven channels stay within a unit at a quarter of a per cent of channel noise and are no better than five at one per cent. The first half holds. The second fails completely: at one per cent seven channels still read every light's every one of a hundred readings within a unit, while five lose twenty-two lights. The premise was wrong. Seven channels' matrix barely cancels — its contributions sum with one sign — and per unit of noise it moves the reported chromaticity no more than three broad channels do.

 Every stock's ink lies on the premium stock's curve. Five pigments' cancelling strengths at a ten-per-cent target against the sheet's glow: the premium stock at eleven brightener loadings from a tenth of its own to all of it, the light, office and premium census sheets, and the premium stock bleached until its glow equals the office's and the light's. Every census sheet lies on the premium stock's curve, and the bleached premium sheets on the census sheets — the orange at 0.5581 and 0.6461 against 0.5578 and 0.6456. What the brain does

The glow alone sets the light ink

A light ink cancels a pale tint's residual hue turn on brightened paper at one strength for each pigment, and that strength falls with the share of the sheet's luminance that is glow — along a curve, not the straight line through three census stocks, which runs as much as 0.016 above it. The prediction was that the strength is one curve in the glow whatever the stock: that a premium sheet bleached until its glow equals the office sheet's, or the light sheet's, needs that sheet's ink within 0.003 for every pigment. It needs it within 0.0007, though the two papers differ in base, in whiteness and in how much of the brightener their ink shields. Traced across eleven loadings of the premium stock, the curve runs through every census sheet within 0.003, sags under every chord, and falls five to nine times as steeply on a dim sheet as on a bright one for the orange, red and violet.

 Wings raise every core-built ceiling's margin over the peak. The reversed tube with every line a 2.4 nm Gaussian core convolved with Lorentzian wings, read through slits of four, five and six nanometres, each window's ceiling built for an unnotched line of the core's width: the least ceiling over the true peak at any column depth to five, against the wings' width. 4 nm: −5.7%, +2.2%, +13.0%; 5 nm: −11.6%, −3.3%, +8.0%; 6 nm: +7.3%, +15.4%, +27.8%. What light is

Wings make a broad line easier to bound from its core

A spectral table read through a wide slit can still bound a reversed mercury tube's true peaks with a ceiling built for an unnotched line, so long as the slit is more than about two and a half times the lines' Gaussian core; at a ratio of 2.08 a five-nanometre table fails from a column depth of 3.5. A real discharge line has Lorentzian wings, which widen the width a table declares, and the prediction was that wings move that threshold out in proportion to the declared width, while a ceiling built on the declared width keeps it where it was. Both halves fail, the first in direction. With wings a tenth of a 2.4 nm core's width the five-nanometre table holds to 4.5, and with a quarter it holds at every depth, by eight per cent: wings take power from a line's peak and give it to the table's entry. A ceiling built on the wider declared width fails three of six winged tables that the core-built ceiling serves.

 Read by size, the steepest ramp triples the brown's error. The same cases read by the mean size of the chroma error, overshoot and shortfall alike. On the ramp from L 17 the brown's low binders err by 15.72 thousandths against 5.04 from L 20; the blue's error is 5.19 against 5.21. What it takes to deliver it

A steeper ramp cures the shadow by overshooting it

A four-ink press's total-ink cap dulls the deepest colours it binds, and easing the cap by a fifth on a straight ramp from L* 20 to the floor halves the deep blue's dulling. A ramp bent to take more ink above L* 20 dulled every group more, so the cure seemed to follow how steeply the cap falls below L* 20. The prediction was that a ramp steeper still — the whole fifth spent from L* 17 — cures more. Read as the census has read every ramp, by the mean chroma lost, it does: every group dulls less at each step from L* 20 to 18.5 to 17. But the brown's low binders cross nought: their prints come out more chromatic than the colour asked, by as much as 0.07, and read by the error's size the brown ray's mean error more than doubles. The measure that ranked every earlier ramp was signed, and a signed measure cannot see an overshoot. Eased by a tenth, the steep ramp is no worse anywhere and better on three cases of four.

 The overshoot follows how fast the cap gives ink back. For each ramp and ease, the brown ray's deepest overshoot — the most its print exceeds the chroma asked — against the rate at which the eased cap returns ink below the ramp's start. A tenth from L 17 and a fifth from L 20 return ink at the same 7.3 points a unit and overshoot by 0.013 and 0.019; a fifth from 17, at twice the rate, by 0.070. What it takes to deliver it

Each dark ray wants its own ease

Easing a four-ink press's total-ink cap below L* 20 cures the dulling of the deepest colours it binds, and a steeper ramp — the whole ease spent from L* 17 — cures more on the signed mean, but eased by a fifth it prints the brown ray's low binders up to 0.07 more chromatic than asked and more than doubles the brown's error read by size. The prediction was that three twentieths from L* 17 is the ease the steep ramp wants: below every earlier setting on both rays. The brown still overshoots, by 0.039, and errs nearly twice as much as on the gentle ramp. The overshoot follows one number — how fast the eased cap gives ink back per unit of lightness — and the two rays want that number set differently. Hard, the blue errs least on three twentieths from L* 17 and the brown on three twentieths from L* 20; the brown's best costs the blue a fifteenth more error, the blue's best costs the brown four fifths more.

 A tilt moves every light's reading much the same way. Each of the fifty-one lights' shift in reported u′v′ when the seven-channel sensor's gains are tilted by +1 and −1 per cent. At +1 per cent every light moves by about 6.0 in u′ and 3.4 in v′, ten-thousandths, towards the long wavelengths, and no light lies more than 3.3 from that mean; −1 per cent mirrors it. Difference and uniformity

A tilt moves every light's reading alike

An ambient-light sensor of seven narrow channels reads all fifty-one of a room's lights within a ΔEITP unit of a dim moon picture's tint and keeps them there through one per cent of independent noise on every channel, because its matrix barely cancels. A drift shared by the channels does not average out: as a photodiode warms its response tilts, rising at long wavelengths and falling at short. The prediction was that a tilt reaching ±1 per cent at the ends of the range leaves the seven-channel sensor within a unit for every light and that the twelve-channel sensor, whose Z row cancels, loses more lights than seven at every tilt. Seven channels hold to ±2 per cent, the worst light at 0.56. Twelve loses two lights at every tilt — the two it misreads with no tilt at all — and none to the tilt. Every sensor's reported chromaticity moves by about seven ten-thousandths in u′v′ per per cent of tilt, every light nearly the same way.

 The wedge of loss comes to a point at 1.04 times the height. Every box of the 2.5-tall room near the wedge, by its floor's sides as shares of the height, marked by whether every one of the seventy-two paints gains, every one loses, or they split, with the crossing along the row at 0.16 of the height. The floors 1.08 and 1.12 times the height long gain at 0.15, 0.16 and 0.17; the floor 1.04 times as long gains at 0.15 and 0.17 and splits at 0.16, 55 paints gaining; the row crosses nought at 1.0399. What a scene does

The wedge of loss comes to a point

A satin finish on a room's grey walls and ceiling adds colour to a painted floor in almost every proportion of room, and costs it a few thousandths of a per cent of its chroma in one narrow wedge of proportions: floors about as long as the room is tall and a sixth of its height wide. The wedge was known to close between floors 1.0 and 1.2 times the height long. The prediction was that a floor 1.04 times the height long still loses at a sixth of the height and one 1.08 times as long gains at every width near it, placing the tip between them a little above the coarse crossing of 1.036. The longer floor gains everywhere, as predicted. The 1.04 floor sits on the tip itself: its change in chroma is two parts in ten million, the row crosses nought at 1.0399, and there — and in no other box of the census — the seventy-two paints part company, fifty-five gaining and seventeen losing.

 Every paint has its own tip, and the pale ones cross last. Each of the seventy-two paints' own crossing from loss to gain along a sixth of the height, against how far its change swings between 1.0 and 1.08 times the height relative to the mean's, by band width. The crossings run from 1.0387 to 1.0413 of the height; the swings from ×0.37 to ×3.12, and the paints that swing most cross last. What a scene does

The pale paints cross the tip last

A satin finish on a room's grey faces costs a painted floor its chroma only in a narrow wedge of proportions, and at the wedge's tip — a floor 1.04 times the room's height long and a sixth of it wide — the mean change is nought and, alone in the census, the seventy-two paints part company: seventeen lose, fifty-five gain. The prediction was that the losers are the paints with the narrowest reflectance bands, whose chroma depends most on how much grey light mixes in, and that they lose most below the tip and gain least above it. The opposite holds. No narrow band loses; half the broadest do, and fourteen of the seventeen have the highest peak — the pale paints. They lose most below the tip and gain most above it: they are the paints the finish moves most, up to three times the mean, and each paint has its own tip, from 1.0387 to 1.0413 times the height, the palest last.

 A five-nanometre flank loses nothing up to a fiftieth of Lorentzian. The forty-nanometre band pigment's lost splitting directions against the Lorentzian share of its steep side, for a steep side five nanometres wide — the first search and the better of it and a second — and fourteen. At five nanometres: 0, 0, 0, 5 lost at no Lorentzian, a hundredth, a fiftieth and a twentieth, where the first search had 0, 0, 1, 7; at fourteen, 4, 11, 14, 18. Where the model breaks

A five-nanometre flank carries a fiftieth of Lorentzian

A band pigment splitting mercury's green and yellow lines with a steep side five nanometres wide loses none of the twenty-five splitting directions as a Gaussian and seven with a twentieth of Lorentzian in its flank, because the band slides in towards the passed line and carries its wing there. The prediction was that at five nanometres the count follows the single band's error, which grows in proportion to the share: about two lost at a fiftieth and none at a hundredth. A hundredth loses none, as predicted. The first search lost one direction at a fiftieth — and searched again from the best pigments found at the neighbouring shares, that direction reaches within two thousandths of the ideal: a fiftieth loses none either. The same second search recovers two of the twentieth's seven, leaving five. A sharp enough flank tolerates a fiftieth of Lorentzian where a fourteen-nanometre one lost fourteen directions to it.

 The loss deepens to 0.16 and jumps to a gain a hundredth wider. The same row between 0.15 and 0.175 of the height. The loss deepens at every step from −0.0016% at 0.150 to −0.0056% at 0.160, turns to +0.0010% at 0.164, sits at +0.00004% at 0.168 and loses again, −0.00076%, at 0.175. What a scene does

The pocket has a step where its bottom should be

Along a floor as long as the room is tall, a satin finish on the grey faces costs the floor a few thousandths of a per cent of its chroma in a pocket around a sixth of the height wide, deepest at 0.16 of the height by three samples a hundredth of the height apart. The prediction was that boxes between those samples trace one smooth dip with its bottom between 0.39 and 0.41 units, every box from 0.375 to 0.425 losing. The loss does deepen smoothly, at every step from 0.375 to 0.40. Then, a hundredth of a unit wider, at 0.41, every one of the seventy-two paints gains; at 0.42 they split; at 0.4375 every paint loses again. The row changes sign three times in a tenth of a unit of width, and the one step from 0.40 to 0.41 rises three times as fast as the pocket falls. The finish's brightening, which the chroma change rides on, falls smoothly straight through. What the census has called the pocket's bottom is a step.

 Even half a core of wings leaves the three-nanometre table short at the deepest column. The three-nanometre table's worst ceiling over the true peak at every column depth, built for the core's width, with the lines bare and with wings a tenth, a quarter and a half of the core's width. It fails from depths of 2, 2, 2.5, 5 as the wings widen; with wings a half it holds to 4.5 and falls 1.1 per cent short at five. What light is

Wings rescue a three-nanometre table only to its deepest column

Lorentzian wings on a reversed tube's lines make every four-, five- and six-nanometre table's ceiling, built for the lines' 2.4 nm core, safer by eight to twelve points of margin a step, so that wings a quarter of the core's width carry all three to every column depth. The three-nanometre table, at a slit-to-core ratio of 1.25, is the one a bare broad line failed earliest, from a depth of two. The prediction was that it still fails with a quarter's wings, from about three, and holds at every depth only with a half's, because a narrow slit averages less of a line's skirt. The first half holds, half a step early: a quarter's wings fail it from 2.5. The second fails by a hair: with wings half the core's width the three-nanometre table holds to a depth of 4.5 and falls 1.1 per cent short at five. Its margin climbs by 6.6, 9.0 and 13.2 points a step — the smallest lifts of any table, from the lowest start.

638 essays on this thread, page 26 of 27 · all threads · all essays