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

Gamut — the series

52 essays on one idea, from the one that introduces it to the one that assumes the rest.
  1. The CIE 1931 chromaticity diagram with its unreachable region marked. The spectral locus encloses every chromaticity a human eye can see. Cells inside the sRGB triangle are drawn in their own colour; the 85 per cent outside it are hatched, because no value this display accepts is the colour belonging there.

    Most of this diagram cannot be shown

    The chromaticity horseshoe is the canonical illustration of colour science, and nearly every printed copy is filled edge to edge with colours the page cannot produce. The honest version marks them, and the marking covers most of the picture.

    part 1 · matching
  2. Two identical grey patches on different surrounds. Both inner squares are #868686. The one on the dark field looks lighter. The values are checked to be equal before the figure is drawn, so the claim is a fact about the drawing rather than a promise.

    Matching is not appearance

    CIE XYZ predicts when two lights will look the same under identical viewing conditions. It was never a model of how anything looks, and most of the confusion in applied colour comes from using it as one.

    part 2 · matching
  3. sRGB, Display P3 and Rec. 2020 compared on the chromaticity diagram. Three nested triangles inside the horseshoe. sRGB covers 74 per cent of the area P3 covers. Rec. 2020's red and green primaries sit on the spectral locus itself, within 0.000 and 0.002 of it, meaning they are monochromatic.

    What a gamut costs

    Three primaries reach a triangle and the visible region is not a triangle, so something has to give. Widening the primaries helps, has a price in precision and compatibility, and runs into a limit that is geometric rather than technological.

    part 2 · matching
  4. The sRGB transfer function, and the gamma 2.2 curve it is not. Code value against relative luminance. The sRGB function is piecewise — a short linear segment near black, then a 2.4 power law with an offset — and it is close to but not the same as a plain 2.2 power law. Half-way along the axis of stored values sits at 21 per cent luminance, and half the luminance of white is at code 188.

    A hex code is not a colour

    Six hexadecimal digits identify three numbers. Turning three numbers into a colour needs a colour space, a transfer function, a white point and a display, and leaving any of them unstated is the everyday version of every confusion in colour management.

    part 2 · matching
  5. The 1976 uniform chromaticity diagram, with the unreachable region marked. The u′v′ diagram: the same spectral locus and the same sRGB primaries as the 1931 picture, projectively transformed. Straight lines stay straight, so mixtures and the gamut triangle survive; what changes is the distribution of area, and the green region that dominates the 1931 diagram is much reduced. 8% of the cells sampled inside the locus are reachable at Y = 0.55; the rest are hatched.

    The diagram was replaced in 1976

    The CIE knew the 1931 diagram was badly distorted and published a better one. Half a century later almost every chromaticity plot in print is still the old one, and both are still printed filled edge to edge with colours no display can show.

    part 3 · matching
  6. Three lightness scales, seventy years apart. Munsell value, CIELAB's L* and CIECAM16's J against luminance, all rescaled to run 0 to 100. Each is somebody's answer to how evenly spaced lightness steps map onto light. They put the midpoint of the scale at 19.8%, 18.4% and 28.0% of the white's luminance respectively — close enough to be three measurements of one thing, far enough apart to be three measurements rather than one restated twice.

    Colour by catalogue

    A colour order system arranges physical samples on a regular lattice so a person can find one and name it. The space it samples is not regular, and everything interesting about such systems is what happens at that mismatch.

    part 3 · matching
  7. A 20 nm bandpass, and what it does to the sample. The true reflectance, the same reflectance as a 20 nm instrument reporting every 20 nm returns it, and the slit function responsible, drawn at its own scale around 550 nm. The reconstruction is what every calculation downstream will use, and it differs from the truth by ΔE 2.97 under D65. Nothing in the file the instrument writes marks which values were measured and which were interpolated.

    What the instrument reports

    A spectrophotometer sees a sample through a slit of finite width, at a finite number of wavelengths. What it writes down is the truth convolved with its own bandpass, and nothing in the file says which values were measured and which were invented.

    part 3 · matching
  8. One pair of colours, blended in four spaces. Every row starts and ends at the same two colours and visits different ones in between. The chroma of the middle falls furthest in linear light, by 83 units below the straight line between the endpoints' chromas — the grey dead zone that every blue-to-yellow gradient has and that no endpoint mentions. Any colour a route visits that this display cannot show is hatched rather than clipped.

    A gradient is a path

    Two colours fix the ends of a blend and nothing else. The middle is decided by the space the interpolation happens in, and four spaces in daily use put the halfway point of one ordinary gradient as much as thirty-seven units of colour difference apart.

    part 3 · matching
  9. Dominant wavelength as a construction on the diagram, against equal-energy E. A ray is drawn from the white point through each sample and continued until it leaves the diagram. The sample at (0.28, 0.52) leaves through the spectral locus at 537 nm, at excitation purity 0.43. The sample at (0.36, 0.19) leaves through the line of purples, so it has no dominant wavelength at all and is written 544c nm — the crossing on the opposite side, marked with a c. The hatched region is colour this display cannot show and is not drawn as though it could.

    Not every colour has a wavelength

    The question every reader arrives with is which wavelength a colour is. For close to a third of the directions on the chromaticity diagram the honest answer is that there is none, and the construction colorimetry offers instead is a statement about a diagram rather than about light.

    part 4 · matching
  10. What a coarse wavelength grid costs, by source. Colour error against grid size, for four sources through one reflectance. Daylight survives every grid tested: 0.67 ΔE00 even at 40 nm. A source with lines in it does not — the narrowband source reaches 16.2. The grid is not a property of the arithmetic; it is a claim about what the light has in it.

    Five nanometres is a choice

    Every integral here is taken in five-nanometre steps, and the interval has never had to be defended. Coarsening it to twenty costs daylight two hundredths of a colour difference and a fluorescent tube six and a half — and which way of coarsening is used decides a further factor of six.

    part 4 · matching
  11. One match, and what it says about the observer making it. The anomaloscope: a monochromatic 589 nm yellow set against a mixture of 545 and 670 nm. Only two cone classes respond at those wavelengths, so the match is two equations in two unknowns and has one solution for any observer whose two pigments differ. The bar is the fraction of the accepted band; the mark is the solution. A normal observer accepts 0.7 per cent of the scale; an observer whose two pigments are the same accepts all of it, because their two equations are one equation twice. Nothing here is fitted to clinical data: the pigments are the same template used for every observer here, at stated peaks, and the match is the solution of the linear system.

    One match names the observer

    A yellow at 589 nanometres set against a mixture of 545 and 670 is two equations in two unknowns. It has one solution for a normal observer, a solution somewhere else for an anomalous one, and no unique solution at all for a dichromat — whose two equations are one equation twice.

    part 4 · matching
  12. A colour-order system's chips, sorted by what they would be called. A regular lattice in lightness, chroma and hue — the idealisation of a swatch book, and regular by construction because a person has to be able to find a page. Named, it comes apart: 1320 chips inside the gamut divide into 198 for grey and 66 for blue. And 22 per cent of one-step moves in the lattice change the name, so a page of a swatch book is not a page of a vocabulary.

    A catalogue is not a vocabulary

    A colour order system is a regular lattice, because a person has to be able to find a page. Sorted by what its chips would be called, that lattice comes apart into piles differing by a factor of three — and a fifth of all one-step moves in it change the name.

    part 4 · matching
  13. sRGB, Display P3 and Rec. 2020 compared on the chromaticity diagram. Three nested triangles inside the horseshoe. sRGB covers 74 per cent of the area P3 covers. Rec. 2020's red and green primaries sit on the spectral locus itself, within 0.000 and 0.006 of it, meaning they are monochromatic.

    Six numbers make a space

    An RGB colour space is eight numbers — three primary chromaticities, a white point, and a transfer function — and everything else about it is derived. Deriving it rather than copying the matrix is the difference between having a colour space and having a table somebody else computed.

    part 5 · matching
  14. Grassmann's four laws, exact — and the two things that break them. For a linear observer every one of the four is exact and the residual is floating point, which is the control that makes the two failures below measurements rather than artefacts. Rods break a cone-metameric match by 23 per cent of a rod excitation at dusk; bleaching breaks it by 0.59 per cent of a cone excitation in the sun. Both are stated as fractions of a receptor's own response, so they can be put on one scale.

    The laws that make colour add up

    Colorimetry is an integral, and an integral assumes matching is linear. Grassmann's four laws are exact for a linear observer, to floating point — and they fail at both ends of the light range, by two different mechanisms, leaving colorimetry an operating band of three and a bit decades that no standard states.

    part 5 · matching
  15. A fourth primary, swept — every setting an exact match, none of them the same. Four primaries matching three numbers leave one degree of freedom. Along the horizontal axis it is the fourth primary's share of the white's luminance; at each value the other three powers are solved exactly, so every point on this plot is a floating-point-exact match for the reference member — worst residual 1.3e-15 — and no colorimeter can tell them apart. What the population sees runs from 13.7 ΔE00 at the ninety-fifth percentile to 17.3, a factor of 1.26. The best setting is the largest share the arithmetic admits, so what stops it is not colour but the requirement that four powers stay positive.

    Four primaries have a choice

    Three primaries matching three numbers have one answer. Four have a family of them, every member exact to floating point for the observer they were solved for — and the members are not equally good for anybody else, so a display with a fourth primary has a setting that is robust to who is looking at it and a setting that is not.

    part 5 · matching
  16. What a fourth primary actually buys. All three displays are floating-point-exact matches for the reference observer, so no colorimeter can tell them apart. The bars are the 95th percentile of what two hundred other eyes report. Held to the same gamut floor of 1.4× sRGB, the four-primary design leaves the population 2.6 times closer together than the three-primary one. That is what the extra emitter is worth, and it is not more colour — the gamut is held fixed while it is measured.

    A fourth primary is a design

    A display's fourth emitter is sold as more colour. Optimised instead against how far apart two hundred eyes are about its white — with the gamut held fixed so it cannot cheat — it buys agreement, and two and a half times closer together than three primaries reaching the same area, and the wavelengths it chooses are not the ones anybody would pick.

    part 6 · matching
  17. Which of these paints the display can show, and to how many people. Each row is a real surface under D65, and the bar is the share of 120 observers for whom a non-negative mixture of this display's three primaries reproduces it. The question has no observer-free answer: the paint is a reflectance, the primaries are emission spectra, and whether one matches the other is a fact about somebody's cones. A dot marks the rows the 1931 observer calls displayable. 2 of them are rows some real people cannot see, and 4 more go the other way.

    A gamut has a population

    Whether a display can reproduce a paint is a fact about somebody's cones, so the boundary of a gamut is not a curve but a band. On a laser projector, ten of twenty-eight boundary surfaces are ones the standard observer calls displayable and some real people cannot see — and the wider the gamut, the wider the band.

    part 7 · matching
  18. The same twenty-four samples, measured two standard ways. How far apart a 45°/0° instrument and a sphere with its gloss port closed are, on samples running from three per cent reflectance to seventy. The whole of the difference is the interface reflection — four per cent of the light, returned without ever meeting a pigment, thrown away by one geometry and collected by the other. It is the same four points in every row, which is why the disagreement is a property of how dark the sample is rather than of what colour it is: ΔE00 8.7 on the darkest samples against 1.93 on the lightest.

    An instrument has a geometry

    Every reflectance here arrives through a model with a bandpass, a sampling interval and no position at all. Real instruments say where they were standing, and the two standard answers disagree by ΔE00 8.35 on a dark gloss sample — a difference that adds rather than multiplies, and that no adaptation removes.

    part 7 · matching
  19. Outside the set of colours a reflecting surface can be. How far each stock sits from the boundary of the object-colour solid, as a fraction of the bound. The line at zero is the boundary: a perfect diffuser sits exactly on it, and every reflectance ever made sits to its left. The pale marker is the sheet measured with the ultraviolet excluded and the dark one with it included. Two of the six cross the line — they are brighter, in a direction that can be written down, than any reflecting surface of their colour could be. This is a proof rather than a hull: for each sample a direction is found in which the largest value any reflectance can reach is computed exactly, and the sample exceeds it.

    A white that is not a reflectance

    The object-colour solid is the hardest boundary in colorimetry — the set of tristimulus values any reflecting surface can produce, with no assumption about pigments in it at all. A coated press stock under a measurement standard's own lamp sits 1.5 per cent outside it, and a heavily brightened one 4.0, and with the ultraviolet removed both come back inside.

    part 8 · limits
  20. The locus and the triangle, drawn on one of the diagrams. The spectral locus and the sRGB triangle in CIE xy (1931) — the default of the discipline, and the default used here. The triangle covers 33.6% of the enclosed area here. Nothing about the observer or the display has changed between this picture and any other in the family; the coordinates have, and the coordinates are what an area is measured in.

    The diagram has no area

    A chromaticity diagram is a projective picture of a three-dimensional space, and the freedom colour matching leaves in the observer acts on it as a projective map. Straight lines and mixture ratios survive that; area, distance and angle do not — so half of what the diagram is used to say is a statement about the paper.

    part 8 · matching
  21. What share of the diagram the sRGB triangle covers, in twelve published coordinate systems. Each row is a chromaticity diagram somebody has printed, and each bar is the fraction of the enclosed visible area that the sRGB triangle covers in it. Every row describes exactly the same observer and exactly the same gamut. The answer runs from 8.5% to 38.4%, a factor of 4.52, because area is not preserved by the projective maps that carry one of these diagrams to another. The familiar "about a third" is a fact about CIE xy.

    Two thirds is not a property of the eye

    It has been said from the beginning here that about two thirds of the chromaticity diagram cannot be shown on a screen. The figure is right, on the diagram it was measured on, and across twelve published diagrams the same triangle covers anything from 8.5 to 38.4 per cent of the same locus. Counting stimuli instead gives an answer that does not move.

    part 9 · matching
  22. A display's primaries, scored as the adaptation basis they are. Four primary sets ranked by the mean ΔE00 an adapted observer is left with when the white point moves — which for a display is a gain on R, G and B, and so a von Kries adaptation in the inverse of its own primary matrix. sRGB leaves 2.36, as much as scaling XYZ directly and therefore as much as having no cone basis at all. Rec. 2020 leaves 1.09, better than every published adaptation transform fitted to corresponding-colour data. Nobody chose that: it is what wanting a wider gamut does to a primary's spectral selectivity.

    The gamut race chose the basis

    Twenty years of arguing about how much of the diagram a display should cover has produced primaries whose inverse is a better adaptation basis than any transform ever fitted to corresponding-colour data. On the invariant count of what those displays can actually show, the same twenty years produced nothing at all.

    part 10 · matching
  23. What a display's red primary is allowed to be, under four requirements at once. A close view of the chromaticity plane around one designed primary, 0.101 units across. Four outlines: the set of positions the primary can take before each of four requirements gets one per cent worse — how well a gain in the display's own basis undoes a change of light, how much of the diagram the three primaries enclose, how many real surfaces fall inside them, and whether a light of that colour exists at all. The shaded region is where all four hold. It is 5% of the smallest outline's area, because the outlines are long and thin and cross at an angle rather than nesting. adaptation holds 42% of its boundary, gamut holds 10% of its boundary, realisable holds 48% of its boundary.

    A primary is chosen for four things

    A display's primaries have to adapt well, cover the diagram, hold the surfaces anybody photographs, and be colours a light can actually have. Drawing all four tolerance regions around one primary shows that no single requirement decides where it can go, and that one of the four never decides anything.

    part 10 · matching
  24. The bowl the eigenvalues describe and the bowl a sample found. Six points on a logarithmic vertical axis — the distance from the optimum of the adaptation residual to a 5 per cent rise along each of the six directions the objective can see — with a shaded band behind them showing the whole range 24 random directions reported. The eigen-radii run from 1.2e-2 to 3.6e-1, a factor of 29.8. The band runs from 2.2e-2 to 1.8e-1, a factor of 8.0, and sits entirely inside the ends of the true range: a random direction in nine dimensions carries a share of every eigenvector and so reports the middle of the bowl, never an end of it.

    How long is the bowl

    The optimum of an adaptation objective is twenty-nine times longer one way than another, and the two ends have names — the stiffest direction is almost entirely the short-wavelength row of the basis, the flattest almost entirely the long-wavelength one. Twenty-four random directions reported a factor of eight, and no number of them would have done better.

    part 11 · matching
  25. How far a fitted transform is from anybody's eyes. Five groups of three bars: for each published adaptation transform, the distance from the population's own cloud to the confusion point that transform is committed to, measured in the population's standard deviations on that point. On the protanope's point every one of them is between 2.2 and 14.4 out, and on the deuteranope's between 3.7 and 11.8. On the tritanope's, 5 of the five are within three standard deviations — indistinguishable from a member of the population. The claim that these matrices are not cone responses is safe, and the evidence for it is two points out of three.

    Two points out of three

    Every published adaptation transform implies three dichromat confusion points, whether or not it was fitted to any. Measured in the population's own standard deviations they are two to fourteen out on the protanope's point and four to twelve on the deuteranope's — and between half a standard deviation and two and a half on the tritanope's, which is inside the population. The claim that these matrices are not cone responses is safe. The evidence for it is two points.

    part 11 · matching
  26. The same tolerance, in the two numbers somebody actually sets. The plane a maker of a single-peak emitter works in: peak wavelength across, full width at half maximum up. Each marker is a candidate emitter whose chromaticity falls inside the colorimetric tolerance drawn for this display's green primary. They occupy a narrow band — peaks from 528 to 535 nanometres, a span of 7, against widths from 25 to 45 — so a tolerance stated as a region in chromaticity becomes ±3.5 nanometres of peak and a great deal of latitude in width. 2.0% of the 2501 candidates land inside at all: most of a region drawn in chromaticity is a colour no single-peak emitter makes.

    A tolerance in the wrong coordinates

    A display primary's tolerance is written as a region in chromaticity, because that is where the colorimetry lives. Nobody has a knob for chromaticity. What a maker of an emitter sets is a peak wavelength and a bandwidth, and the map between the two is so anisotropic that on the red primary its condition number is over eleven thousand.

    part 11 · matching
  27. Exactly flat everywhere, and eigenvectors at one point only. Two columns over the same nine places. On the left, how far from zero the objective's second derivative is along a row-scaling direction, on a logarithmic axis — it is between 10⁻⁹ and 10⁻⁶ of the largest eigenvalue at every one of them, which is a numerical zero. Scaling a row of the basis is a straight line along which the cost does not change, and that is true at every point, not only at the optimum. On the right, the angle between those three directions and the Hessian's own three smallest eigenvectors: 0.025 degrees at the optimum and up to 88 away from it. An invariance is a property of the function; being an eigenvector is a property of the function at a minimum, and the two coincide only where everybody computes.

    Only the flat directions keep their names

    Three of the nine numbers a colour match leaves free do nothing, and they do nothing everywhere — exactly, at every basis in this collection's table. They are the objective's own principal directions at one point only, and everywhere else the directions carrying the curvature have turned by tens of degrees.

    part 11 · matching
  28. A published residual is a mean, and the worst object in the room costs twice it. Three bars for each of the 14 changes of light in the adaptation census, ordered by how uneven the change is across surfaces. The first bar is the published mean residual. The second is the worst single surface in the audit's published test set. The third is the worst surface anywhere in the region that set is drawn from, found by search rather than by reading a maximum off a lattice. The mean-to-worst ratio runs from 1.90 to 4.02 and averages 2.43, so every published adaptation number has a worst case about twice it that no essay had ever quoted. The gap between the second and third bars is the other finding: a maximum over 125 sampled points understates the region's own maximum by up to 34 per cent.

    An extremum is still not a sample

    Two rounds ago three measurements turned up that took a maximum over a sample of a set and were short by up to a factor of two. The same error was live in a fourth place the whole time, on the set of surfaces every adaptation number is averaged over, and it is short by up to a third.

    part 12 · matching
  29. What six of this collection's published numbers do when the unit changes. Six quantities, from six calculations that share nothing: a change of light after an observer has adapted, a camera profile's error, the gap between the two standard observers, a metameric pair under the lamp that breaks it, the same image on two papers, and an observer two seconds into a new room. Each is recomputed under all six units and every unit is calibrated onto ΔE2000's scale first, so the bar is not a change of units in the ordinary sense. The bar is the ratio of the largest reading to the smallest, and it runs from 1.71 to 2.30. Five of the six are printed in ΔE2000 by the essays that report them; the sixth is printed in CAM16-UCS, because the model it comes out of defines that unit.

    The observers differ by a unit's worth

    The gap between the 1931 and 1964 standard observers is the one quantity in this collection's audit with no published number under it — nothing here reports it as a single figure over a stated set. It also has the second-largest dependence on which colour-difference formula is used, running from 1.60 to 3.55 across the menu.

    part 12 · matching
  30. Two slabs with one reflectance, and two colours through an aperture. Two constructed media whose bulk reflectance agrees at every wavelength to fifteen figures, and whose diffusion lengths differ by a factor of four. The upper curve is that shared reflectance — both slabs lie on it exactly. The two patches on the right are what a 4 millimetre radius returns from each, and they are 6.3 ΔE₀₀ apart. This is a metamerism with no observer in it: the two samples are the same colour to anybody under any light, and the instrument separates them because it is measuring a kernel through a hole rather than measuring a reflectance.

    A pair the aperture separates

    Two constructed slabs with the same reflectance at every wavelength, to fifteen figures — the same colour to any observer under any light — and 6.25 ΔE₀₀ apart when measured through a four-millimetre aperture. It is a metamerism with no observer in it, no illuminant in it, and no spectral difference to construct it from.

    part 12 · matching
  31. MacAdam's twenty-five ellipses, measured in each unit. The uniformity instrument used here, applied to units rather than to spaces. The upper bar is anisotropy — the mean over the twenty-five of the largest radius divided by the smallest, where 1 would be a circle. The lower is spread — the largest mean radius divided by the smallest across all twenty-five, which asks whether a step of the same size means the same thing in different parts of the diagram. Reading down the three CIELAB-based formulae in the order they were published, the anisotropy falls 3.42 → 2.89 → 2.74 and the spread rises 3.24 → 3.59 → 4.12: the weighting divides a difference by the chroma it was measured at, which equalises directions at a point and unequalises magnitudes between points. Neither number is scaled, so no calibration is applied here. CAM16-UCS is ahead on both.

    A unit rests on a space that was ranked

    This collection ranks three colour spaces by how nearly they make MacAdam's ellipses circles, and CIELAB comes last. It then publishes every difference it computes in a formula built on CIELAB. Turning the same instrument on the formulae rather than the spaces shows the repair works — and that it buys roundness by paying in evenness.

    part 13 · matching
  32. What this collection's grid does to its own observer audit. Two bars per light: the mean departure of the observer computed on this collection's five-nanometre grid, and the same computation on a quarter-nanometre one. For five of the six lights the two agree to two decimal places, which is what a well-sampled spectrum looks like. For the laser projector the coarse answer is exactly zero and the fine one is 1.80 — the largest in the table. On a five-nanometre grid a three-line spectrum with lines at 465, 532 and 638 nanometres is a one-line spectrum, and a single wavelength is a stimulus every observer agrees about to the last bit. The two departures do not compound here; the first conceals the second.

    One wavelength is everyone's colour

    A stimulus with a single wavelength in it produces the same relative cone excitations for every observer, exactly, whatever their age or field size. A display made of three such stimuli is where observers disagree most. Both statements are consequences of the same algebra, and the second is why laser projection has an observer problem.

    part 13 · matching
  33. The same white, matched at six primary widths. At every width the three primaries are solved to match D65 exactly for the reference member; the bands are what the population sees. A broad primary integrates the observer differences over a band and averages them away; a narrow one samples them at a point and passes them straight through. From 40 nm to 2 the ninety-fifth percentile rises from 11.3 to 17.9 ΔE00, monotonically, and the technology has been moving from left to right for thirty years.

    A narrow primary buys a disagreement

    The observer audit decomposes what a display costs a population. Narrowing the primaries raises the pigment-peak departure monotonically, moving one raises or lowers the macular departure, and the two respond to different design variables — so a wide gamut and an observer-robust display are bought with the same money.

    part 14 · matching
  34. One surface dimmed sixteen times, and the two things that happen to it. A single surface, dimmed by successive halvings, with the lens age departure measured on it at every level. The tristimulus deviation falls by exactly the dimming factor — 16 times over the sweep, to the last bit, because the colour integral is linear in the stimulus. What a unit of that deviation is worth rises by 8.2 times over the same sweep. The colour difference the audit reports is the product of the two, and it falls by only 1.95.

    A deviation is not a difference

    The previous round priced twenty departures in colour differences and treated each number as a property of the thing that departed. Every one of them is a product of two factors — how far the reading moved, which is linear and belongs to the departure, and what a unit of that movement is worth where it landed, which is not linear and belongs to the colour. Dimming one surface sixteen times scales the first by exactly sixteen and the second by eight.

    part 15 · matching
  35. The angle between two departures, which nothing in the audit records. Every pair of the six departures on every surface — 2520 pairs — binned by the angle between their two deviations in the local metric. The distribution reaches both ends: 440 pairs sit under thirty degrees and point almost the same way, and 615 sit above a hundred and fifty and point almost opposite. On 1095 of the 2520 the two together cost less than the larger of them alone. A table of magnitudes cannot say which case it is in.

    A size is not a direction

    An audit that reports magnitudes cannot say what two of them cost together. Over two and a half thousand pairs of observer departures the angle between them in the local metric runs from one degree to a hundred and seventy-nine, and on forty-three per cent of them the two together cost less than the larger of the two alone. Adding the angle predicts the composition to one and a quarter per cent; Pythagoras is out by twenty-eight.

    part 16 · matching
  36. The straight piece under the cube root, and where it stops. CIELAB's lightness against relative luminance, over the bottom 5.0 per cent of the range. Below Y = 0.008856 it is a straight line of slope 7.787; above it, a cube root. The two meet at L* 8 in value and in slope, exactly — the CIE's two constants are chosen to make that true. The dashed curve is the pure cube root, which reaches negative lightness before it reaches zero luminance and has an infinite slope there. The break is marked, and the axis runs to Y = 0.050.

    The straight piece under the cube root

    CIELAB's lightness is described everywhere as a cube root and below a stated luminance it is a straight line, spliced on with two constants chosen so the join is exact in value and in slope. Every black a delivery chain reaches is inside that straight piece — a press black at L* 2.4, a projected black at 1.1 — where the compression does not compress, the price of a deviation is flat to two parts in a thousand, and the second derivative the composition of two departures needs does not exist.

    part 16 · matching
  37. What one tolerance accepts, around four colours. The surface in tristimulus values that ΔE₀₀ 1.0 draws around four colours, each outline scaled to its own size so the shapes can be compared. The volumes they enclose differ by a factor of 1.0e+5 across the sRGB cube, and the longest axis of one shell is between 3.3 and 29.9 times its shortest. A tolerance is written as one number and is a different set of samples at every colour it is applied to.

    What one number accepts

    A delivery tolerance is written as a single colour difference and it acts on three tristimulus values, so what it actually accepts is a closed surface. Measured over sixty-four colours in the sRGB cube, the volume inside that surface varies by a factor of a hundred thousand and its longest axis is between three and thirty times its shortest. The same contract, applied to a dark colour and a light one, is two different requirements.

    part 17 · matching
  38. What a code lattice costs, and where. Twelve thousand colours quantised to 8 bits per channel through the sRGB transfer function and read back, with lightness across the bottom and the colour difference the rounding cost up the side. The mean is 0.191 and the worst case is 1.15, a factor of 6.0. The bars are band means, and they rise: the encoding spends its codes in the shadows, so the top of the ramp is where the lattice is coarsest against a metric that does not compress as hard.

    A lattice has no derivative

    Every departure priced here was priced by perturbing something and reading the answer, which requires the thing being perturbed to have a derivative. A file written on a code lattice does not have one — its output is flat almost everywhere and jumps on a set of measure zero — so quantisation can be bounded and never propagated. The bound is 1.15 colour differences at eight bits per channel against a mean of 0.19, and it is worst where the encoding spends fewest codes.

    part 17 · matching
  39. What a small neutral deviation costs as the grey darkens, in two lightness scales. The price of a fixed fraction of deviation on a neutral — how much lightness it is worth per unit of luminance — from L 40 down to L 0.25, both axes logarithmic. CIELAB's L is a straight line below L 8 and its price there is exactly flat. The appearance model's J′ has no straight piece: its price keeps rising, by a factor of 5.0 across the same range, as the luminance to the power -0.442 — the derived exponent is -0.441 in an average surround.

    The appearance model has no straight piece

    CIELAB's lightness is a straight line below L* 8, so a deviation near black has a fixed price and a black has a floor under it. CIECAM16 has no such piece. Its lightness near zero goes as the luminance to a power set by the room, the price of a deviation rises without limit as the black deepens — as Y to the power −0.44 in a lit room and −0.58 in a dark one — and on a projected black a lightness unit in the model allows under half the luminance change a unit of L* allows.

    part 17 · matching
  40. Two primaries mixed, and the line a reader assumes they take. The additive mixture of two sRGB primaries, walked in twenty steps, plotted in the a and b of CIELAB. The filled points are where the light actually goes, which is exactly straight in tristimulus values because that is Grassmann's second law. The open points are the straight line between the two readings. They part company by 26.7 ΔE₀₀ at their furthest, at 60 per cent of the way along, and the half-and-half mixture misses the midpoint by 21.1.

    The mixture line bows

    Grassmann's second law says an additive mixture is exactly linear in tristimulus values, and tested here it is exact to floating point. Nothing downstream of the three numbers preserves it. The physical half-and-half mixture of two colours sits a median of 5.5 colour differences from the midpoint of their two readings and up to thirty; on a green and a blue display primary it is twenty-one, which is a quarter of the distance between them.

    part 18 · matching
  41. Which way each tolerance is tightest, colour by colour. Sixty-four colours on a lattice through the sRGB cube, with lightness across the bottom. For each, the tolerance ΔE₀₀ 1 is pulled back into tristimulus values and its shortest and longest axes are found. The filled points are the angle between each colour's shortest axis and the direction common to all of them: a median of 15 degrees, ninety per cent within 32. The open points are the same for the longest axis, at 23. The common short axis is 3.1 degrees from X up and Y down, which is the direction of a*.

    A tolerance has a grain

    A colour tolerance pulled back into tristimulus values is a long thin shape, and across the whole sRGB cube its shortest axis points the same way, fifteen degrees off at the median — the direction in which X rises as Y falls, which is the direction of a*. So what one colour difference allows depends on which way a delivery drifts. An instrument's X filter may age 1.4 per cent before the tolerance is used up, its Z filter 5.1, and the light on the sample 9.3.

    part 18 · matching
  42. Four mixtures of display primaries, bowing by different amounts in two units. The half-and-half mixture of each pair of sRGB primaries, measured from the midpoint of the two readings as a share of the pair's own separation. The upper bar is ΔE₀₀ and the lower is the appearance model's J′a′b′. In ΔE₀₀ green and blue bows most, at 25 per cent; in the model it is red and blue, at 20, and blue and yellow falls from 22 to 12. The right-hand column gives the model's distance with its power correction, which is smaller than the Euclidean one for every pair here.

    Which mixture bows most depends on the ruler

    The physical half-and-half mixture of two lights misses the midpoint of their two readings in any unit a person is shown, and the share it misses by is about the same in ΔE₀₀ and in the appearance model's own space — 12.8 and 14.6 per cent at the median. Which mixtures miss most is not. The rank correlation between the two units is 0.57, a display's blue and yellow is the worst pair in one and the mildest in the other, and the smaller number usually quoted for the appearance model is its power correction rather than a straighter line.

    part 19 · matching
  43. Where a press's variation lies, and where the tolerance charges for it. Each printed patch's sheet-to-sheet variation under four stated mixes of press variation, split along the three axes of a one-unit ΔE₀₀ tolerance at that patch, tightest first. The upper bar of each pair is the share of the variation along each axis and the lower the share of the price, averaged over 29 patches. For an even mix the tightest axis holds 1.5% of the variation and pays 29% of the price, and the loosest holds 83% and pays 36%. For inking alone the tightest axis holds 2.0% of the variation and pays 34% of the price, and the loosest holds 79% and pays 28%. For a gain-led press the tightest axis holds 1.8% of the variation and pays 29% of the price, and the loosest holds 83% and pays 39%. For a trap-led press the tightest axis holds 1.6% of the variation and pays 29% of the price, and the loosest holds 82% and pays 35%.

    A press is charged for the direction it barely moves

    A press run varies almost entirely along the direction a colour tolerance forgives. Split along the tolerance's own axes under four stated mixes of press variation, its tightest axis holds about one per cent of the sheet-to-sheet variation and pays a quarter to a third of the price — and on a blue overprint four fifths, for the balance between two inking units rather than the level of either.

    part 19 · matching
  44. The straight line between two colours, and the formula's own shortest path. Five gradients seen from above, in the a and b plane of CIELAB: the straight line between the two colours in grey, and the shortest path under ΔE₀₀'s own local metric in colour. The lightness coordinate bows too and is not drawn. red to green saves 3.6 per cent and leaves the straight line by 13; blue to yellow saves 7.5 per cent and leaves the straight line by 21; cyan to magenta saves 3.9 per cent and leaves the straight line by 10; black to white saves 0.0 per cent and leaves the straight line by 0; red to blue saves 0.3 per cent and leaves the straight line by 4 CIELAB units at the widest. Black to white is the flat case: its shortest path is the straight one.

    The straight line is not the shortest gradient

    A colour difference formula says what a small step costs at every colour, and that is enough to ask which path between two colours is shortest. It is not the straight line: between red and green the shortest path bows thirteen CIELAB units away, saves four per cent — and stays inside sRGB on every step where the straight line leaves it. The formula's own answer for the two endpoints, meanwhile, is neither length.

    part 20 · matching
  45. The metamerism index, computed three ways, as the reference match loosens. The special metamerism index of 6 metameric pairs under an incandescent test light, against how well each pair matches under the reference light. Uncorrected, the index absorbs the reference mismatch and rises from 2.86 to 3.76. Corrected multiplicatively it rises to 3.31 and additively to 3.87. All three are the same number when the pair matches exactly, which is the only case the definition covers.

    The metamerism index has two corrections

    The index for a metameric pair is defined for a pair that matches exactly under the reference light, and no real pair does. The standard's remedy is to correct the sample first, and it names two corrections — scale the tristimulus values, or add the difference. On six pairs matched to one colour difference, the two answers differ by a tenth to four tenths of an index unit; at two, by a whole one.

    part 20 · matching
  46. Two uniform spaces, two answers: cyan to magenta. The gradient from above, in the a and b plane. The grey line is the straight path in CIELAB; the two coloured curves are the shortest paths under ΔE₀₀'s own local metric and under CAM16-UCS's. They are 15.7 CIELAB units apart at their furthest, against bows from the straight line of 9.9 and 12.2. Both spaces are published as uniform and both are used to decide what lies between two colours; they do not agree.

    Two uniform spaces disagree about between

    ΔE₀₀'s own local behaviour says which colours lie between two colours, and so does CAM16-UCS's. They do not agree. On cyan to magenta the two shortest paths run fifteen CIELAB units apart — more than either runs from the straight line — and on that gradient ΔE₀₀ would rather have the straight line than CAM16-UCS's answer. Getting the comparison at all takes noticing that one of the two has no local metric: its published difference is a Euclidean distance raised to the power 0.63, and a power below one has an infinite derivative at zero.

    part 21 · matching
  47. What a gamut charges a gradient, against what the relaxation's noise is. For each gradient, how much longer the best path that stays inside sRGB is than the free shortest path, as a percentage of the free one. The shaded band is the relaxation's own noise, measured by relaxing the same free path from two different starting points: 0.42 per cent at its worst. Every excess is inside it. So holding a gradient inside a display's gamut costs nothing measurable in the metric's own units, on any of these pairs — including the ones whose free path is outside at most of its points.

    A gamut charges a gradient nothing

    On red to green the shortest path happened to stay inside sRGB where the straight line did not, and that was called a coincidence of mechanism. Made into a measurement it is not a coincidence: constraining a path to stay inside the gamut costs less of the metric's own length than the relaxation's own noise, on every gradient tested — including the ones whose free shortest path is outside at fifteen of its twenty-one steps. The gamut's boundary and the metric's cheap region point the same way, and the price of the constraint is nothing.

    part 21 · matching
  48. How often the choice of correction changes a pair's grade. Eighteen metameric pairs walked to each of eleven reference mismatches, with the share whose index falls in a different band under the two corrections the standard allows. At an exact match the share is zero and must be: there is nothing for either correction to correct. It rises to 44 per cent at a reference mismatch of 2, which is the quality a dyehouse reaches rather than the quality a laboratory constructs. The banding is a five-step convention at 0.5, 1, 2 and 3, stated here rather than quoted, and how much the count depends on it is drawn separately.

    The ambiguity is largest where the index is used

    The metamerism index is defined for a pair that matches exactly under the reference light, no real pair does, and the two corrections the standard allows for the residual give different answers. Over eighteen pairs at eleven match qualities the gap is nearly a function of the mismatch alone — its middle half spans a factor of 1.4 at the mismatches a dyehouse reaches — so it could be tabulated. And it is largest exactly there: zero at a laboratory's match, and re-grading eight pairs in eighteen at a trade's.

    part 21 · matching
  49. Three uniform spaces, three answers: red to blue. The gradient from above, in the a and b plane: CIELAB's straight line, and the shortest paths under ΔE₀₀'s local metric, under CAM16-UCS's and under Oklab's, whose shortest path is its own straight line carried back into CIELAB. They bow from CIELAB's line by 14.4, 42.5 and 42.6 units. ΔE₀₀'s and CAM16-UCS's paths run 28.3 apart at their furthest, Oklab's runs 29.0 from ΔE₀₀'s and 5.6 from CAM16-UCS's, and the space furthest from the other two here is ΔE₀₀.

    A third space breaks the tie only once

    ΔE₀₀ and CAM16-UCS disagree about which colours lie between two colours. Oklab was the obvious tie-breaker, and it breaks the tie on one gradient of five: on red to blue its path runs within six CIELAB units of CAM16-UCS's and twenty-nine from ΔE₀₀'s. On red to green and cyan to magenta it keeps to CIELAB's straight line while both others bow, and on blue to yellow it bows further than either. Each of the three spaces is the odd one out somewhere. Asking also exposed a CAM16-UCS path that had never converged.

    part 22 · matching
  50. Six starts for every held gradient, and where each one lands. Each row is one gradient held inside a gamut, relaxed from six starts: the straight line, the free shortest path, and the straight line bent towards and away from the neutral axis and up and down in lightness. Each dot is how much longer than the free shortest path that start's result is, on a logarithmic scale from a hundredth of a per cent to a thousand; the ring marks the best. The first 10 rows are gradients between colours on a coated CMYK press's boundary whose straight line leaves the press: their best routes cost a median of 0.12% and at most 1.1%, and on 5 of them some start lands at more than twice the free length. The next 3 are the display gradients held inside sRGB, whose best cost at most 1.3% and whose starts spread by at most 4.2%. The last 6 are press gradients that never leave, where no start is trapped.

    A press makes the cheap route a search

    Holding a gradient inside a display's gamut cost nothing measurable, and the essay that found it credited the gamut's convexity. A press's gamut is not convex: 216 of 630 straight lines between colours on its own hue ring leave it. Holding gradients inside the press still costs little at best, a median of a tenth of a per cent. But the best route depends on where the relaxation starts, the free path is the best start on one gradient of ten, and on half of them some start is trapped at more than twice the free length. On the display no start is trapped.

    part 22 · matching
  51. The same gradients, priced by a penalty and by a projection. Each row is one gradient held inside a coated press from six starts. The pale dots are the penalised relaxation — a free step, with a price for leaving the press — and the dark ones are the projected relaxation, which takes the free step and then moves each point to the nearest printable colour. Across is how much longer than the free path each result is, logarithmic. On the 10 gradients whose straight line leaves the press, the penalty leaves 9 starts at more than twice the free length and the projection leaves none. The projection's best route costs a median 0.02% against the penalty's 0.13%, and its spread across starts is 0.95% against 132.5%.

    A projection has no reason to detour

    Holding a gradient inside a press by penalising the excursion turned the gamut's price from a number into a search: on five of ten crossing gradients some starting point leaves the relaxation trapped at more than twice the free length, and the spread across six starts runs to 271 per cent of the free path. Replacing the penalty with a projection — take the free step, then move each point to the nearest printable colour — leaves no trap on any gradient and a spread of 0.06 to 2.4 per cent.

    part 23 · matching
  52. How much of a coated press is left at each margin inside its boundary. The share of a coated press's printable volume in CIELAB that lies at least a given distance inside its boundary, for margins from half a unit to 32. A margin of one unit keeps 88%, two keep 80%, four 66% and eight 45%. At every margin up to 30 what is left is a single connected piece; at 31 units, with 0.14% of the volume left, it first splits, into a core of 623 cells and 2 fragments of one or two cells — the deepest point is 32.7 units inside, so what splits is the last crumb of the core, not a waist.

    A margin costs a press its corners

    A gradient held inside a coated press by projection was predicted to tear if the press were first shrunk by a safety margin, at any pinch narrow enough for the shrinking to cut. The press has no such pinch: shrunk by any margin up to thirty CIELAB units it stays in one piece, and projected routes on it are neither trapped nor torn. What a margin costs is concentrated at the corners — the solid yellow moves nearly four times the margin to get inside, like the tip of a 31-degree spike.

    part 24 · matching

All series