Systems that kept the measure

A picture with two eyes in it

Several traditions draw the floor from one place and the people on it from another. This site has said no single camera produces both. What such a picture actually is has a measurement attached: give the rays their world points and ask for the one place they all pass through, and at a stride of separation the best answer misses them by six tenths of a metre.

Worth reading first: A carpet and the people on it · Assembled from several views · The point you have to stand at.

A Persian miniature draws the carpet as though from above and the figures standing on it as though from in front. This site has already measured what that asks for: two optical axes exactly ninety degrees apart, and no single camera has both.

The carpet and the people want optical axes 90° apartLeft, a camera on the carpet's normal: the carpet is a true square and a figure standing under the eye is drawn at exactly no height at all, whatever its height — off the axis its drawn length is proportional to how far off it stands, not to how tall it is. Right, a level camera: the people are right and the carpet is a 56 px band against its 271 px width. The convention takes the carpet from the left picture and the people from the right, and there is no camera that supplies both.a camera looking downthe carpet is true, the people are nota camera looking levelthe people are true, the carpet is notthe two views a miniature is assembled from90° apart, exactly
Fig. 1 The convention, and the contradiction in it. A camera on the carpet’s normal images it as a true square and images a standing figure as a mark whose length is its distance from the point under the eye.

That is a refusal. It says what such a picture is not. It does not say what it is, and the answer turns out to be an ordinary object with a number attached.

A picture plane that belongs to nobody

The first thing needed is a place to draw on that is not either eye’s coordinate system.

Everywhere else on this site a picture is a camera’s pixels. That will not do here, because two eyes cannot share a set of pixels that belongs to one of them. So the picture plane is put in the world: a real plane in the room, at a stated distance along the axis, and every centre draws onto it by the same rule — join the centre to the world point and see where the join crosses the glass.

Two eyes, one sheet of glassThe arrangement the picture is made in, seen from somewhere else. The rectangle is the picture plane — a real plane in the room, not either eye's pixels — and the two eyes both draw onto it. Every mark on it is where a line from one of the eyes to a point of the room crosses the glass. The two eyes are 1.41 m apart. Nothing in the drawing they make between them records which of them drew which mark.correct from 19 cm, at 160 mm widetwo eyes 1.41 m apart
Fig. 2 The arrangement. The rectangle is the picture plane, a plane in the room rather than anybody’s pixels, and both eyes draw onto it.

The check that makes this comparable with the rest of the site is that drawing from the camera’s own eye reproduces the camera’s own projection routine to the last bit, at two different depths of the plane — which also says the depth of the plane is a free choice that changes nothing, as the plane is a choice requires.

The picture plane tilted 14°Pointing the camera up tilts the picture plane with it, and three things happen at once: the verticals converge — 3.59° between the outer two — the horizon drops 213 px below the middle of the frame, and the vertical vanishing point arrives at 3425 px from the principal point. They are one fact: the product of those two offsets is f².correct from 20 cm, at 160 mm wideverticals converge 3.59° · horizon 213 px off centre
Fig. 3 The result that entitles the picture plane to be moved. One eye and two picture planes give a homography of the picture for any scene whatever, so where the glass is put is a decision about coordinates.

The picture

Now draw a room. The floor and the near box from the first eye; the two far boxes from a second eye a stride away.

The near half from one eye, the far half from anotherThe floor and the near box are drawn from the first eye; the two far boxes are drawn from a second eye 1.30 m away from it. The faint outlines are where the first eye would have put those same boxes, and the gap between the two answers about one world point runs to 902 px. Taking every drawn mark with the world point it is a mark of and asking for the one point all those lines pass through, the best answer misses them by 0.539 m — so this picture is a projection of this room from nowhere at all.two eyesno single viewpoint — the rays miss by 0.54 mtwo centres, 1.30 m apart
Fig. 4 The result. The faint outlines are where the first eye would have put the far boxes; the drawn ones are where the second eye put them, and the two answers about one world point differ by up to nine hundred pixels.
The near half from one eye, the far half from anotherThe floor and the near box are drawn from the first eye; the two far boxes are drawn from a second eye 0.40 m away from it. The faint outlines are where the first eye would have put those same boxes, and the gap between the two answers about one world point runs to 423 px. Taking every drawn mark with the world point it is a mark of and asking for the one point all those lines pass through, the best answer misses them by 0.249 m — so this picture is a projection of this room from nowhere at all.two eyesno single viewpoint — the rays miss by 0.25 mtwo centres, 0.40 m apart
Fig. 5 The two eyes closer together. The seam narrows and the picture looks more and more like a photograph.
One eye, and the ordinary photograph it makesThe floor and the near box are drawn from the first eye; the two far boxes are drawn from a second eye 0.00 m away from it. The faint outlines are where the first eye would have put those same boxes, and the gap between the two answers about one world point runs to 0 px. At zero separation the two eyes are one eye, the gap is nothing, and the picture is an ordinary photograph.one eyecorrect from 17 cm, at 160 mm wide50° across
Fig. 6 And at zero separation, which is the control. One eye, an ordinary photograph, and the figure prints a viewing distance rather than saying it has none.

The last of those three is the reason the other two mean anything. A generator that produced a two-centre picture at every setting could not distinguish its own defect from its own machinery; one whose slider passes through the ordinary case can.

The measurement

Here is the question worth asking: how far is this picture from having a centre?

It has an answer, and the answer uses machinery this site already had. Take every drawn mark together with the world point it is a mark of, and join the two — that is a line in space. A picture with a centre is a picture whose lines all pass through one point. So fit the best point and report how far it misses.

How far from having a centre, and it depends on knowing the roomTake every mark in the two-centre picture with the world point it is a mark of, join the two, and ask for the point all those lines pass through. There is none: at 2.40 m of separation the best point misses them by 1.118 m on average and 1.565 m at worst, and the miss falls to nothing as the eyes come together. That is the measurement — and it needs the room. Given the same picture and the room the picture is *consistent with*, the same fit returns a residual of 1.2e-15 m at every separation on this plot.00.500100.50011.502distance between the two eyes (m)how far the rays miss their own best point (m)worst raythe room it is consistent with1.12 m at 2.4 m apartzero for the absorbed reading
Fig. 7 The miss, against how far apart the two eyes are. At a stride and a half of separation the best point misses the rays by nine tenths of a metre on average, and the miss falls to nothing as the eyes come together.

That solver is closestPointToRays, and it is on its fourth object. It measures how far a refracted picture is from having a centre; how far a handscroll is; how far a curved mirror’s bundle is; and now this. Using the same solver is deliberate, because it makes the millimetres comparable across four different reasons for a picture to have no viewpoint.

The rays of a refracted picture, continued into the waterEvery ray leaves the pinhole, bends at the surface and carries on. Fitted to a common point they miss it by 9.9 mm — the circle is that miss drawn at the figure's own scale. With the water removed the same fit misses by 0e+0 m.the water surfacethe pinholethe rays miss by 9.9 mmno single viewpoint — the rays miss by 9.9 mmdry control: 0e+0 m
Fig. 8 Its first use: rays continued into water, missing their own least-squares centre by millimetres. A refracted picture is not a projection of anything from anywhere, measured the same way.
The rays of 27 m of scroll, and the point they miss by 8.57 mPlan of one section. Each ray leaves the eye at its own column, so the eyes lie along a track rather than at a point. The circle is the least-squares centre drawn at the radius of its own miss — 8.57 m, which the closed form puts at 8.57 m, the standard deviation of a track that long. A single column of the same scroll fits exactly.the best point, missed by 8.57 m27 m of the eye's trackno single viewpoint — the rays miss by 8.57 mthe eyes are a track, not a point
Fig. 9 And its third: a handscroll’s section, whose rays miss by metres because the eye translates. Same solver, different reason, comparable number.

What the miss depends on

Three things move it, and separating them is what makes it a measurement rather than a number.

The separation of the eyes, which is the plot above and is nearly linear over the range drawn. Doubling how far apart the two centres are roughly doubles the miss.

The depth spread of the scene. The miss is a parallax, and parallax needs depth: a scene entirely on one plane produces no miss at all whatever the two eyes do, because a plane seen from two points differs by a homography of the picture and a homography is a perfectly good picture from one point.

The picture plane tilted 14°Pointing the camera up tilts the picture plane with it, and three things happen at once: the verticals converge — 3.59° between the outer two — the horizon drops 213 px below the middle of the frame, and the vertical vanishing point arrives at 3425 px from the principal point. They are one fact: the product of those two offsets is f².correct from 20 cm, at 160 mm wideverticals converge 3.59° · horizon 213 px off centre
Fig. 10 The reason. Two eyes looking at one plane differ by a map of the picture, so a flat scene cannot record how many eyes drew it.

And which points are handed to which eye. The measurement is not a property of the two centres alone; it is a property of the composite, and a division that gives the second eye only far-away things produces a smaller miss than one that splits the scene near the camera.

The near half from one eye, the far half from anotherThe floor and the near box are drawn from the first eye; the two far boxes are drawn from a second eye 1.30 m away from it. The faint outlines are where the first eye would have put those same boxes, and the gap between the two answers about one world point runs to 1213 px. Taking every drawn mark with the world point it is a mark of and asking for the one point all those lines pass through, the best answer misses them by 0.838 m — so this picture is a projection of this room from nowhere at all.two eyesno single viewpoint — the rays miss by 0.84 mtwo centres, 1.30 m apart
Fig. 11 The second eye moved mostly upward rather than sideways. The miss is a different size because the displacement is along a different direction relative to the scene’s own depth.

The second of those is the one to carry, because it is a genuine limitation on the whole enterprise: a picture of a flat thing cannot be caught at having two eyes, however far apart they are. Every measurement here needs depth in the scene to work with.

Two frames stitched on the sky, with the pivot 50 mm behind the pupilThe far field registers to 1e-13 px, which is what makes the stitch look correct. The near field does not: the nearest post lands 3.2 px from where the other frame put it and the furthest 0.31 px, an error falling as 1/distance — the signature of a parallax rather than a calibration error.the far field — where the stitch was fitted2.2 m — 3.2 px out24 m — 0.3 px outthe sky registers to 1e-13 pxthe foreground does not — up to 3.2 px
Fig. 12 The same fact from the panorama side: rotating a camera about anything but its entrance pupil leaves parallax between frames, and the parallax depends on the depth of what was photographed. No depth, no parallax, nothing to detect.

Which of the four reasons this is

The other three are ways for a picture to lose its centre by accident. The medium bends the light; the eye moves during the exposure or along the roll; the reflector curves. In every one of them somebody wanted a projection and the physics did not supply one.

This one is different. The picture has no centre because it was drawn from two on purpose, and the traditions that do it are doing it for a reason — a floor that can be measured with one ruler, a figure whose parts are each seen from the direction that identifies them.

One cube in 5 parallel drawing systemsEvery one preserves midpoints exactly. What separates them is the axis scales, printed beneath each — isometric's are all 0.8165, which is equal and is not 1.elevationx 1.000y 1.000z 0.000cavalierx 1.000y 1.000z 1.000cabinetx 1.000y 1.000z 0.500isometricx 0.816y 0.816z 0.816dimetricx 0.943y 0.943z 0.471axis scales measured from the drawingall 5 preserve midpoints
Fig. 13 The five systems this site models, each a stated map from a world to a page. Two of them are two-centre pictures once they are looked at this way.
What each parallel system does to the three axesThe smallest axis scale is plotted; the note says how many of the three coincide and whether the system is an orthographic projection or an oblique construction. cavalier and isometric both make all three the same — but isometric is the only orthographic one that does, at 0.8165, and the oblique route gets there by drawing depth at true length.elevation0.0000two equal, orthographic ←cabinet0.5000two equal, obliquecavalier1.0000all three equal, obliquedimetric0.4714all three different, orthographicisometric0.8165all three equal, orthographictrimetric0.5479all three different, orthographicsmallest of the three axis scalesmeasured from each projection
Fig. 14 And what each buys with what it gives up, on one battery. A centre and a true measure are exclusive, and a system that wants both takes two centres.

The measurement is therefore not a criticism. It is a size: the miss is how much geometry the convention is spending, and a convention that spends a metre of it is making a bigger claim than one that spends a centimetre.

The miniature's two grounds are two centres, and the instrument is the same oneThis site already measured the miniature convention and found that its floor and its figures want optical axes 90° apart, so that no single camera produces both. Read as a two-centre picture the same arrangement has a number attached to it rather than a refusal: carrying the second eye up toward overhead, the rays of the composite miss their own best point by up to 1.12 m. The convention is not a failure to find one centre. It is a picture with two, and the two are far apart on purpose.00.5001020406080the second eye, carried toward overhead (degrees)how far the rays miss their own best point (m)two axes 90° apartmiss up to 1.12 m
Fig. 15 The miniature’s own arrangement run through the instrument. Carrying the second eye up toward overhead — which is what the convention’s ninety degrees asks for — the composite’s rays miss their best point by more than a metre.

Two centres, or a hundred

Nothing above needs the number two. The construction takes a list of parts, each with its own centre, and draws every part from its own — so a picture assembled from a dozen viewpoints is the same object with a longer list.

That matters because it puts two of the five systems this site models on one axis rather than in two categories.

A handscroll is drawn by an eye that translates continuously, imaging one column at a time. That is not two centres; it is a centre for every column, which is a continuum of them. Read through this instrument it is the limiting case of the composite picture, and the miss it produces has a closed form — the spread of the eye’s own track — which is exactly what a continuum of centres spread along a line should give.

A scroll in plan: the eye travels, and images one line at a timeThe eye runs along the track at the bottom. Each position images the single vertical plane it is level with, so a world point is drawn by exactly one position of the eye — the one at its own x. The paper advances 26 px for every metre of travel whatever the scene does, which is why the roll is a map along its length.the eye's trackthe eye at x = -10.5 mevery point is drawn by the one position of the eye that is level with itplan — the eye's track and the scans it makes28 m of travel
Fig. 16 The scroll’s arrangement in plan. Every column is drawn from a different point of a track, and the track’s own length is the miss.
A scroll keeps the midpoint along its length and loses it acrossLeft, a segment lying along the roll: the image of its midpoint and the midpoint of its image are the same point to 0e+0 px. Right, a segment running away from the eye: the two are 21.9% of the segment apart. One projection, two answers, because the eye is at infinity in one direction and seven metres away in the other.along the rollthe two midpoints coincide — 0e+0 pxone mark, drawn twiceacross itthey separate by 21.9%the image of the midpointthe midpoint of the imageno single viewpoint — the rays miss by 6.9 m21.9% of the receding segment
Fig. 17 And the consequence in the picture: the midpoint survives along the roll and not across it, because there is one eye per column and no eye for the whole.

An aspective figure is the other extreme: a small number of centres, each owning a named part, chosen so that each part is seen from the direction that identifies it.

The best single view keeps 3.00 of the 5 aspects; the composite keeps all of themEach bar is the share of that part which reaches the picture from the best single viewing direction, found by sweeping the sphere. The composite takes each part from its own direction, so every one of these would be 1. The sweep's total, 2.9999, agrees with the closed form √(2² + 2² + 1²) = 3.0000 — and the control is a figure whose parts all face one way, where the same sweep returns 5.000 of 5 and no convention is needed.head0.668profile — the outline that names a faceeye0.669frontal — an eye in profile is a wedgeshoulders0.669frontal — the width that says two armslegs0.668profile — a stride is a side viewpond0.326plan — a rectangle of water is a rectangle1.000 — what its own aspect keepsshare of each part that reaches the picture, from the best single direction|d · n| for each part3.000 of 5, swept and in closed form
Fig. 18 The aspective arrangement measured as an optimisation. Any single viewing direction keeps at most the square root of what several directions keep between them, which is the reason for taking more than one.

So the four-eyed miniature, the continuous scroll and the composite figure are three points on one scale — how many centres, and how far apart — and this instrument reads all three.

The seam, and where it is not

There is a natural expectation that a two-centre picture has a visible join, and it is half right.

The disagreement between the two eyes about one world point is real and can be large — nine hundred pixels here. But it is only visible where both eyes drew something. In the picture above each world point is drawn once, by whichever eye owns it, so there is no double image anywhere and nothing on the sheet is inconsistent with itself.

The near half from one eye, the far half from anotherThe floor and the near box are drawn from the first eye; the two far boxes are drawn from a second eye 2.20 m away from it. The faint outlines are where the first eye would have put those same boxes, and the gap between the two answers about one world point runs to 1452 px. Taking every drawn mark with the world point it is a mark of and asking for the one point all those lines pass through, the best answer misses them by 0.984 m — so this picture is a projection of this room from nowhere at all.two eyesno single viewpoint — the rays miss by 0.98 mtwo centres, 2.20 m apart
Fig. 19 A wide separation. The far boxes sit in the wrong place relative to the floor and neither of them is drawn twice, so nothing in the picture contradicts anything else in it.

What does show is a relational wrongness: the boxes stand somewhere the floor does not put them. That is the same complaint a viewer has about the miniature — the figures do not stand on the carpet in a way the carpet agrees with — and it is a complaint about the arrangement rather than about any mark.

A shadow across the creaseOne straight rod, one lamp, two receiving planes. Each piece of the shadow is dead straight — 7e-16 m and 1e-15 m from the line through its own ends — because each is a plane projectivity of the rod, and a projectivity takes a line to a line. They meet at 35.08°, and the corner is the image of the crease rather than anything about the rod.35.08°correct from 16 cm, at 160 mm widetwo maps, meeting at 35.08°
Fig. 20 A seam of the other kind, for contrast: one drawn on purpose where two receiving surfaces meet, which is a fact about the receiver rather than about the number of eyes.

The instrument’s own control

One more thing has to be checked before any of the numbers above are quotable, and it is the check every use of this solver on this site carries.

With one eye, the fit has to land exactly. The same points, the same picture plane, the same least-squares routine, with both halves drawn from the same centre: the rays meet to eight parts in ten thousand million million of a metre, which is the arithmetic floor.

One eye, and the ordinary photograph it makesThe floor and the near box are drawn from the first eye; the two far boxes are drawn from a second eye 0.00 m away from it. The faint outlines are where the first eye would have put those same boxes, and the gap between the two answers about one world point runs to 0 px. At zero separation the two eyes are one eye, the gap is nothing, and the picture is an ordinary photograph.one eyecorrect from 17 cm, at 160 mm wide50° across
Fig. 21 The control drawn rather than described. One eye, an ordinary photograph, and a fit that returns the eye.

Without that, the metres above would be indistinguishable from the solver’s own noise on a badly conditioned bundle, which is not a hypothetical worry — the rays of a narrow-angle picture are nearly parallel, and a least-squares point fitted to nearly parallel rays is exactly the kind of quantity that comes back large for no reason.

Two rays, 2.53 mm apart, in the plane that contains bothThe ray from the left eye through its mark and the ray from the right eye through its. With the marks placed exactly they meet, to 3.3e-14 m. With the same marks read to 1 px they miss by 2.53 mm at a range of 7.45 m. Triangulation is not an intersection; the reported point is a choice about what to minimise, and the gap is the part a residual alone will not tell you.midpoint — 2.53 mm gapfrom the left eyefrom the right eyegap 2.53 mm at 7.45 mexact marks: 3.3e-14 m
Fig. 22 And the shape of that worry in the two-view case, where two rays that should meet miss by millimetres for reasons that have nothing to do with the geometry being measured.

What this does not say

It says nothing about why a tradition would do this. That is a question about pictures and people and this site’s ruling on the conventions field is explicit: what is modelled here is a map from a world to a page, and every claim survives with no painting, no period and no object named.

It says nothing about how a viewer reads such a picture. The miss is a fact about rays; whether a person notices is a fact about seeing.

And it does not yet say what the picture is. A number for how far from having a centre it is, is not the same as an account of what it is a picture of — and the answer to that is the next rung, and it is more surprising than the measurement.

What has been established, and what has not

Worth separating before the next rung, because the two are easy to run together.

Established. A picture drawn from two centres has no centre of its own, and how far it is from having one is a number in metres of the world. The number needs the scene — the rays are joins of drawn marks to world points — and it needs depth in the scene, and it is zero when the two centres coincide.

Not established. That such a picture is wrong, or that it is not a projection of anything at all. Those are different claims and the second one is false, which is the finding of the next rung and is the reason this one stops here.

What each drawing system preservesEvery cell is read out of the system's own map rather than typed: the centre is fitted from its rays, measure is the site's own midpoint test, diminution is the drawn size of a fixed object at two depths, bounded depth asks whether infinite depth lands in finite page, and straightness is the sag of a receding line. Only perspective has a centre, and it is the one system with no true measure. No row has both.a centretrue measurediminutionbounded depthstraight linesperspectivehandscrollisometricdimetrictrimetriccavaliercabinetelevationmilitaryfilled means the system keeps itmeasured from each system's own projectionno row is filled in both of the first two
Fig. 23 The battery this site measures a drawing system on. Nothing in it asks whether a system is right; every entry asks what it keeps.

The distinction matters because “no centre” reads as a defect and is not one. A parallel projection has no centre either — its centre is at infinity, which is to say it has none in the room — and nobody regards an isometric drawing as a failed photograph.

Moving the object, in both familiesThe same box, in place and translated 2.4 m across the world. In the parallel drawing the second image is the first translated by 98.6 px and nothing else — every edge the same length to 3e-14 px. In the perspective drawing the edge lengths change by up to 77.0%, because the direction from the eye has changed and a projection through a centre depends on it.horizonisometric — the same drawing, movedperspective — a different drawingcorrect from 23 cm, at 160 mm wideparallel: 3e-14 px · perspective: 77.0%
Fig. 24 The system with no centre that nobody complains about. What it gives up is diminution and what it buys is a drawing a ruler can be used on.

Why the picture plane had to be in the world

One design decision underpins every measurement here and it is worth defending, because the obvious alternative fails in a way that would not have been visible.

The obvious alternative is to draw both halves in one camera’s pixel coordinates: project the near half with camera A, project the far half with camera B, and put both sets of pixels on the same canvas.

That produces a picture, and it is the wrong picture. Camera B’s pixels are its own — its principal point is at the centre of its frame, its focal length scales its image — so pasting them onto camera A’s canvas silently applies a similarity that has nothing to do with the geometry. The result depends on the two cameras’ frame sizes rather than on where the two eyes were.

The same lens behind five sensorsA 50 mm lens subtends 39.6° across full frame and 8.7° across a phone sensor. The focal length is the same in every one of these; what changes is the rectangle behind it. Two setups matched on angle rather than on focal length are correct from the same distance — 222 mm for a 160 mm print — whatever their formats.full frame · 39.6°APS-C · 26.6°Micro Four Thirds · 19.6°1 inch · 15.0°phone (1/1.7″) · 8.7°one 50 mm lens · the angle is a property of the rectangle behind itMicro Four Thirds: 2.00× diagonal, 2.08× wide, 1.85× tall50 mm across five formats39.6° down to 8.7°
Fig. 25 The quantity that would have leaked in. A focal length is not an angle until a frame of stated width is named, so two cameras’ pixels are not comparable without one.

Putting the plane in the world removes the question. There is one sheet of glass, both eyes draw on it, and the pixel coordinates are read off it afterwards by one map — camera A’s, chosen arbitrarily, and the choice affects nothing because it is applied to the whole picture at once.

The picture plane tilted 14°Pointing the camera up tilts the picture plane with it, and three things happen at once: the verticals converge — 3.59° between the outer two — the horizon drops 213 px below the middle of the frame, and the vertical vanishing point arrives at 3425 px from the principal point. They are one fact: the product of those two offsets is f².correct from 20 cm, at 160 mm wideverticals converge 3.59° · horizon 213 px off centre
Fig. 26 The result that says the choice is free: one eye and two picture planes give a homography of the picture, so which plane is used is a change of coordinates rather than a change of picture.

That is the same reason this collection insists on projecting from a stated camera rather than constructing by eye, applied one level up: the coordinates a measurement is made in have to be a choice that provably does not enter the answer.

The transferable form

A refusal becomes a measurement the moment the same instrument that fails is asked how badly. “No camera does this” and “the rays miss by 0.63 m” are different kinds of statement, and only the second can be compared with anything.

This site has made the first kind of statement about five drawing systems and about three optical arrangements. The second kind is what let the scroll’s miss be compared with the water’s, and it is what lets a convention’s departure from a projection be put on the same axis as a physical one.

The next question follows straight from having a number: a miss of a metre is a metre of something. The rays are not passing through one point, so they are passing through a family of points, and asking what that family is turns the whole arrangement into an object with a name.

What links here

Computed from the collection, not written here: the essays that point at this one.

Reads more easily once this is understood

Essays that name this one as worth reading first.

Shares its objects with

Essays that name at least two of the same things, and that neither author linked.

Named objects

A flat tag is an object no other essay names yet.

centre of projectionDegeneracyDemonstrationDrawing systemerror propagationFree parameterleast squaresParallaxPicture planeStation point