Uneven enlargement from mixing chamber, alignment, lens falloff or mechanical vignette

Owned by Break/Fix — the room that fogs and the light that is not even, whose unevenness differential sorts eight causes with the same two questions this entry asks — does it change with aperture, and is it symmetric about the centre — and adds a third cut this entry did not have: rotate the paper through 180° and print again, which separates a pattern fixed to the equipment from one fixed to the sheet, and which that page is careful to label as its own reasoning rather than published practice. The optics belong to enlarger optics and design. This entry was written earlier from Part VI’s optical work and Part XIV’s insistence that an exposure field is mapped.

What you see

A print that is not evenly exposed across its area when the negative says it should be. The pattern matters more than the amount, because each cause draws a different one:

Pattern	Cause it points at
Bright centre, falling smoothly and symmetrically to all four corners	Source or mixing chamber, or lens falloff
Brighter along one edge or in one corner, smoothly	Misalignment
A soft-edged darkening confined to the corners, worst wide open, better stopped down	Lens falloff
A hard-edged darkening at one or more corners, unchanged by aperture	A mechanical vignette

Symmetry and response to aperture are the two questions, and between them they split four causes into four answers.

Likely causes

1. The source or the mixing chamber. A condenser system with the lamp off-centre, or a diffusion chamber too shallow to mix, illuminates the negative unevenly before the lens sees it.
2. Misalignment. Negative stage, lens board and easel are meant to be parallel; when they are not, one part of the paper is further from the lens than another, and both illumination and focus vary across the sheet.
3. Lens illumination falloff. A property of every lens, and Rodenstock plot it for each of theirs: fall-off in illumination in f-stops against relative image height, drawn against the 1 − cos⁴ reference curve.
4. A mechanical vignette. A carrier aperture too small for the format, a filter drawer intruding, a bellows fold, or a hood — anything that physically clips the cone of light.

What is happening: the chemistry and physics

Falloff has a geometric floor that no lens escapes. Off-axis, light arrives at the paper through an aperture that is foreshortened, at a greater distance, and at an angle — and the three effects together produce the familiar cosine-to-the-fourth law that Rodenstock draw as the reference on their own charts. A real lens may do better or worse than that curve, and their charts are the evidence for which; but nothing does better than physics for free. It is the same argument the course already makes about a pinhole: thick-plate vignetting is the extreme case of an aperture that foreshortens.

Falloff responds to aperture and a vignette does not. Stopping down cuts the cone symmetrically about the axis, which reduces the angular spread and therefore the falloff; a piece of metal in the way is in the way at every aperture. That is the single most useful discriminator on this page.

Misalignment breaks symmetry, because it makes the geometry different on two sides of the axis. Everything else on the list is symmetric about the optical axis by construction.

A mixing chamber is an optical component, not a housing. What it is for is to arrive at the negative with the same intensity everywhere, and a shallow chamber or an off-centre lamp does not.

!The confirming test: map the easel with no negative in the carrier

Remove the negative, set the head and lens to the working height and aperture, and expose a sheet of paper to a very short exposure — enough for a light mid-grey. Process it fully.

That grey sheet is your illumination map, and it is the same measurement Part XIV calls the step everybody skips.

Then repeat it with one variable changed at a time:

• Stop down two stops and re-map. Corners that improve are lens falloff; corners that do not are a mechanical vignette.
• Rotate the paper 180° on the easel and re-map. A pattern that stays with the paper is the paper or the processing; a pattern that stays with the easel is the optics.
• Rotate the carrier 180° and re-map. A pattern that follows the carrier is the carrier.
• Check alignment with a spirit level or a mirror, and re-map. A pattern that is symmetric after alignment was misalignment.

Reading the map by eye is enough for large errors. Reading it with a densitometer against a step wedge turns “the corners look darker” into a figure in stops, which is what you need before deciding whether it matters.

Diagnostic questions

• Is the pattern symmetric about the centre? If not, suspect alignment first and everything else second.
• Does it change with aperture? The falloff-against-vignette test, and it takes one extra sheet.
• Does it change with magnification? Falloff depends on the angles, so raising the head changes it; a mechanical vignette at the carrier often does not.
• Is the corner darkening soft or hard-edged? Optics are soft; metal is hard.
• Did it appear after something was changed? A new carrier, a new lens, a bulb replaced with a different shape, a filter drawer left half in.

Corrective action

Align first, because misalignment corrupts every other measurement you would use to diagnose the rest, and because it costs focus as well as evenness.

Stop down. Rodenstock’s charts show falloff at stated apertures for a reason; two stops from wide open is usually where an enlarging lens is meant to work.

Remove the obstruction if the fault is mechanical. A carrier filed out, a drawer pushed home, a hood removed.

Burn the corners if the residue is small and unavoidable, and record how much, so the correction is repeatable rather than improvised.

Prevention

Map the easel once, at the apertures and heights you actually use, and keep the maps. The calibration records worksheet is where they go, and a map is what turns a suspicion into a number.

Re-map after any change to the head, the lens or the carrier.

Check alignment on a schedule. It drifts, and it drifts slowly enough that no single session shows it.