Nobody drew the shape the dust makes

A flat black field. Scattered across it, bright white specks — some round, some short and thread-like — and a handful of thin straight lines crossing between them.

Connect the brightest points on any dark field and you get something that looks drawn on purpose. Nothing here was.

Every mark in the panel is a mote of dust or a scratch on a sheet of glass, caught by a light that was never meant to photograph either one.

Why are the specks bright white against black rather than dark?

Because of where the dust is sitting relative to the light, not because of anything about the dust itself.

Dust on a camera lens sits between the light and the sensor, so it blocks a little of what comes through and shows up as a dark, soft-edged smudge. Dust on a scanner's glass sits directly against the source — the bed's own light is aimed straight up at it from underneath, close enough that a speck scatters that light back at full strength rather than shadowing anything behind it. The result is the opposite of a dirty lens: not something blocking the picture, but something the light found and returned in full.

Why do some of the marks look like lines rather than dots?

Because they are a different kind of defect, caught by the same light.

A round mote of dust scatters light from a single point and reads as a dot. A hair, a fibre, or a fine scratch across the glass scatters it along its whole length instead, and reads as a thin bright line. Both are surface conditions on the same sheet of glass, lit the same way, at the same pass. Nothing distinguishes them except their shape, and their shape is simply the shape of whatever was actually there — the same rule that turns a ridge in a folded sheet into a straight line with its own shadow: the light reports the physical shape of whatever is in its way, and does not care whether that shape was meant to be there.

Then why does it look like a chart, with lines connecting the brightest points?

Because a scatter of bright points on black is exactly the pattern a person has always looked at and found shapes in, and a scanner bed covered in motes and scratches is close enough to that pattern to trigger the same reflex.

The lines that appear to link the specks into a net are the scratches themselves, sitting where they happen to sit — not drawn to connect anything, and not aligned to any of the dots on purpose. A handful of them cross near enough to a speck that the eye completes the rest, the same way it always has with an actual night sky. The chart is not on the glass. It is the oldest trick the eye plays on a field of points, applied here to grime instead of stars.

Who owns a scattering of dust and scratches?

Nobody. Dirt on glass belongs to no one, and there is no photograph or chart being copied here — the field is built entirely from invented marks standing in for what a dirty scanner bed produces.

If a real constellation chart, with its named stars and its centuries of established figures, were the actual point of a design, that specific chart's illustrator or publisher could hold rights in its particular drawing — the positions of real stars are not owned by anyone, but one artist's way of connecting them can be. Nothing here reproduces any such chart. Style has no owner. A work does.

Does this work for anything other than a scan?

Yes, for any surface where accumulated small marks are the entire subject rather than a flaw in one.

A windscreen after a long drive. A whiteboard that was never fully wiped. The underside of a glass tabletop, seen from below with something lit under it. What the idea needs is a field dark enough to carry contrast and a light source close enough to the marks to throw them forward rather than let them sit as texture — a dim, evenly lit field just reads as dirty rather than as a record of anything. It is a different family of unintended pattern from the sweeping bands a photographed screen produces, which come from two grids meeting rather than from anything settling on a surface, but both are marks nobody drew, made legible only because a light happened to catch them.

Isn't this just an unwashed scanner, not a design?

It is exactly an unwashed scanner, which is the whole design. Every flatbed accumulates this over years of use — the same machine that, left with its lid up, records a leaf against total absolute black is also, on every ordinary scan it makes, quietly logging every mote that has ever settled on its glass and never been cleaned off. Nobody scans the glass on purpose. This is what happens when the thing being copied is thin or pale enough that the glass's own condition gets recorded along with it.

Does it print well?

Yes, and cleanly — flat black holding one small colour of mark is close to the simplest thing fabric can carry.

The black field needs to stay unbroken with no banding, since the whole image depends on true black reading as empty rather than as a dark grey with texture in it. The specks and lines want to stay crisp and hard-edged rather than softened into a glow — a soft dot reads as a star with atmosphere around it, which this is not. One ink for the marks against the garment's own black is enough; a second colour has nothing to do here.

How do I get one made?

Describe the field at JustOG — how dense the specks run, and whether any of them are allowed to look like lines. Pick a direction, drag the crop frame, see it composited on the real garment, and it's made to order and shipped.

Designs other people have published are in the shop.

Somewhere a scanner bed is still carrying today's dust into tomorrow's scan, uncleaned and unnoticed.