A standard digital negative is printed on a transparency as a grayscale image — each pixel has a precise tonal value, from pure white to deep black. On most processes, that works very well: the sensitized paper absorbs the chemistry, the exposure transmits the tonal transformation, the final print is clean.
On some pigment chemistries — typically carbon and Aquaprint — solid flat grays cause a problem. Pigment laid down as a thick, uniform layer tends to bleed during exposure, or to lift off during clearing. The final print shows dirty transitions at the edges of each flat. Dither solves this by turning the flats into dot patterns that simulate gray at normal viewing distance, without laying down a continuous pigment layer.
#How it works
Three positions:
Off (default): your negative is exported exactly as the app computes it — a solid grayscale image, no screening. This is what suits most processes (cyanotype, platinotype, bromoil, gumoil, resinotype).
Auto: the app screens your negative automatically. The screening used by default is an AM halftone (amplitude modulated) — the classic magazine printing technique, where the size of each dot varies with the tonal value. The screen (the dot pattern) is rotated 54 degrees to avoid visual artifacts.
On (forced): same screening as Auto, but the app can't disable it automatically even when the generation mode doesn't suggest it.
In some rare configurations, AM halftone screening can fail or produce a poor result. In that case, Calibration Flow automatically falls back to Floyd-Steinberg — an error-diffusion algorithm dating from 1976 that produces a grainy pattern rather than a regular one. You don't have to choose between the two algorithms — the app handles the choice based on what works.
Whatever the mode, screening is applied after the final composition (so after adding the frame, registration marks and background). Screening applies only to the useful image, not to the layout's graphic elements.
When the generation mode is Color, the Dither option is hidden automatically — screening a color image makes no sense in this workflow.
#Why it matters
For carbon (carbon transfer). Pigmented gelatin VP N°05 tends to lift off during transfer when laid down as a thick, uniform flat. A screened negative produces micro-zones of pigment instead of continuous layers, which makes the transfer more stable and preserves tonal resolution in transition areas.
For Aquaprint (gum print). Pigmented gum arabic VP N°04 suffers from the same problem — an extended dark gray flat can either bleed during clearing in hot water (40 °C), or create a visible break at the edges. Screening breaks the flats into more forgiving patterns.
For consistency across a series. If you produce five Aquaprints from the same source image, consistent screening gives you five prints with the same grain. Without screening, the look of a gray flat can vary from one print to the next depending on the gum's viscosity at the precise moment of brushing.
#When to avoid it
For cyanotype. Iron-cyanide chemistry has no flat problem. A standard, un-screened negative gives a better result — tonal transitions are finer, and screening would only add visible noise at very high print density.
For direct-contact exposure processes (platinotype, palladium, Van Dyke). Same — the chemistry absorbed into the paper doesn't suffer from flats.
For bromoil, gumoil, resinotype. None of these processes requires screening to work. Screening can be tested if you want deliberate grain, but it isn't the answer to a recurring technical problem.
If you haven't observed flat problems on your prints. If your Aquaprints come out clean without screening, don't force it — it would add grain with no visible benefit. The Auto setting lets Calibration Flow decide, but you can always force Off to test without screening.
#Worth remembering
| Item | Value |
|---|---|
| Default algorithm | AM halftone (amplitude modulated, screen rotated 54°) |
| Fallback algorithm | Floyd-Steinberg (error diffusion) |
| Setting positions | Off, Auto, On |
| Application | After composition (frame, registration, background included) |
| Hidden automatically | In Color mode |
| Relevant for | Carbon, Aquaprint (pigment chemistries) |
| Not needed for | Cyanotype, bromoil, gumoil, resinotype, platinotype |
#The test
Generate two versions of the same negative — one with Dither Off, one with Dither Auto. Compare them on screen at 200% zoom: the Off version shows solid grays, the Auto version shows a pattern of fine dots. Print both on transparency and expose both on sensitized Aquaprint paper. Compare the two prints: on the Off version, look at the tonal transitions and the edges of the dark flats. If you see bleeding or breaks, screening would have helped. Otherwise, keep Off — you'll get a better result with no added grain.