The other side of the page is still in the frame

A notice scanned crisp and legible, with the sheet's far side bleeding through backwards behind it, never once landing in register with the front.

A cream tee pinned flat to a wall, printed across the back with the bold heading PLEASE KEEP THIS over dense grey paragraphs of illegible type, and a second, fainter pass of grey type spilling past the print's left edge at a different line height that never lines up with the paragraphs in front of it.

A sheet of dense grey type, headed in bold black: PLEASE KEEP THIS.

Behind it, fainter and never falling on the same lines, a second block of type is visible through the paper — paragraphs that belong to nothing printed on this side of the page.

The two blocks never meet. A page told to be kept is failing at the one job that requires it to stay legible.

Why does a second page show through the first one at all?

Because paper is not opaque, and a scanner's light goes looking for whatever it can find.

A page held up to a window has always done this — ink on the far side has always thrown some of its shape forward. A scanner does the same thing with more light and more attention: the bed illuminates the sheet from underneath as well as from above, and whatever sits on the reverse, or on the sheet lying beneath it in the stack, throws a faint image up through the fibres and into the result. Nothing was copied twice. One pass caught both surfaces at once, because both surfaces were there.

Why does the text behind it run in reverse?

Because you are reading the back of it from the front.

Type set on one side of a sheet faces one direction. Seen through the paper from the other side, it is mirrored the way a shop sign is mirrored from inside the shop — every letter correct in shape, wrong in orientation, and unreadable as a result. That reversal is the proof of what the faint block actually is. A double exposure or a second copy laid over the first would keep its own orientation; only light coming through the sheet from behind arrives backwards.

Why doesn't any of it line up with the text in front?

Because the two blocks were never set to agree, and nothing here tried to make them.

The paragraphs on the front run at their own line height, in their own margins, for their own reasons. The type bleeding through belongs to a different sheet with a different layout, so it crosses the front lines wherever it happens to fall — sometimes between them, sometimes straight through the middle of a line. That collision is what makes it unreadable rather than merely faint. Two texts in perfect register might still be legible, layered like tracing paper. A page that only sits a few degrees off square stays fully legible, because nothing about its own content moved. This is a different kind of misalignment — two separate texts that were never going to agree in the first place.

Who owns a page that was never meant to be read from both sides?

Nobody, if the page and the type on it are invented, which is what is printed here.

No real notice is reproduced and no institution's letterhead sits behind it. Where a genuine document is the whole point of a design — a form, a memo, a real letter — whoever wrote it or holds its copyright can license the reproduction, and that licensing is ordinary business for archives and publishers. Where the document is invented for the design, as it is here, there is nothing to clear. Style has no owner. A work does.

Does the same idea work on something other than a notice?

Yes, on anything with a front and a back worth printing on the same page.

A letter with a signature bleeding through from beneath it. A ledger page with the next entry already showing. A ticket with its own stub reversed behind the stub in front. What the subject needs is one side legible enough to anchor the print and a second side dense enough to survive as a ghost rather than vanish — a mostly blank reverse gives you nothing to show through. A sheet folded rather than stacked gives a smaller version of the same idea: a single ridge scanning as a line with its own shadow rather than a whole second page showing through it, but the same principle — a physical condition of the paper, arriving as a mark nobody drew.

Isn't bleed-through just a sign of a bad scan?

It usually gets treated that way, and the sheet gets rescanned or the setting adjusted. This is the copy that got kept instead — see what shows up when the scanner's own light finds nothing to bounce off for the same machine erring in the opposite direction, toward too little rather than too much. Bleed-through is not damage to the page; it is the scanner recording a second page that happened to be in the light's way, at the same settings as everything it was actually meant to record.

Does it print well?

Yes, provided the two layers stay unequal in weight.

The heading and the front paragraphs want to sit solid and black, holding their edges the way flat type always does on fabric. The bleeding-through block needs to print noticeably lighter — a grey fine enough that it reads as behind the front text rather than beside it, without going so faint it disappears into the fabric's own texture. Reversing that balance, or printing both at full strength, turns two legible layers into one illegible mess, which is the one failure this design cannot afford, since illegibility is meant to belong to the back layer only. Cream or a pale stone base holds both greys cleanly; a white base tends to bleach the fainter pass out altogether.

How do I get one made?

Describe the sheet at JustOG — what the heading says, how dense the bleed-through runs, and how far it drifts from the paragraphs in front. 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.

Whatever was on the other side of that page is still there, reversed, never once meeting the front.

Upload the image you already made, see it on the real garment, and get the physical piece. Made to order, no minimum.

Bring your image →

Related