Edge leak

Lifted from commissioning the pinhole camera, whose two-sheet leak test identifies it, and from Break/Fix — unexpectedly soft images and wrong reciprocity corrections.

What you see

A broad soft band along one edge of the sheet, densest at the edge and fading inwards over perhaps a fifth of the sheet. The picture, where there is one, is otherwise normal.

The decisive observation is that the band continues under the strips that were covered by the paper stops. Image-forming light and flare both stop dead at the edge of the image circle; light that came in any other way does not.

Likely causes

1. A body seam or corner that is not light-tight.
2. The mouth of the dark-slide slot, admitting light around the slide.
3. A joint that runs along an edge — the commonest place for a home-built camera to fail, because a joint along an edge presents a long path and a small angle.

What is happening: the chemistry and physics

Light entering at an edge does not form an image, because it did not pass through the hole. It illuminates whatever part of the sheet it can reach, and the part it can reach best is the part nearest the gap, so the density falls off with distance from the entry point. That gradient is the signature.

The reason a leak too faint to see matters is the exposure time. A pinhole exposure is minutes, and the leak is multiplied by all of them. A leak far too faint to notice by eye can put half a stop of density along one edge of a sheet over thirty minutes.

Paper punishes leaks harder than film does, because it is a short-scale, high-contrast material — the same non-image light that a long-scale film would record as a density difference you would need an instrument to find shows on a paper negative as an obvious band.

Diagnostic questions

• Does the band cross the strips that were under the paper stops? Yes means a leak or fog. No means it came through the hole, and you are looking at flare or a hot spot instead.
• Is the control sheet clean? A sheet from the same box, handled identically, that never went in the camera carries the box, the safelight and the developer and nothing else. Clean control means the fault is in the camera.
• Run the two commissioning sheets. Shutter closed, thirty minutes in full sun, tests every path except the shutter. Shutter open with the pinhole capped tests the shutter and nothing else. A band on the closed-shutter sheet is a body leak.
• Is the mark a soft band or a hard-edged wedge? A soft band fading inwards is this entry. A sharp streak or wedge is a seam or hole leak, and the design names the two separately because the shapes point at different paths.

Corrective action

Find the edge, tape it with black photographic or gaffer tape, and re-run the closed-shutter sheet alone. Recording which pass found the leak matters, because a camera with two leaks will pass the second test only after both are sealed and you will not know that unless you tested one at a time.

Where the dark-slide slot is the path, tape is a diagnosis rather than a repair — the slot has to open. The permanent answer is a light trap of felt or flocked material in the slot mouth.

Prevention

Re-run the two-sheet leak test after any work on the camera and after adding a frame to the stack. A stack that passes at one focal distance can leak at another, because each added frame adds two joints.

Test at thirty minutes as standard, and at sixty and a hundred and twenty when you want to find the faintest leak the camera has. A leak that appears at 120 minutes and not at 30 is a measurable quantity rather than a mystery.

📓Name the edge, not just an edge

Record the leak against a named edge — “the hinge edge”, “the slot edge” — and mark the camera itself so that the same edge is meant next time. A leak table whose entries say “along an edge” cannot be compared between sessions.