Calibrating dot gain in screen printing

#Overview

Print an unscreened, uncorrected halftone gradient and compare it to your computer screen: the print is darker, midtones are clogged, highlights have vanished. Nothing is broken — it's the physics of the process. Every halftone dot grows between the film positive and the fabric:

1. At exposure: light bends slightly around the film's dots; the small dots of the highlights don't cure enough emulsion and wash away at rinsing, while dense areas close up.
2. At printing: ink pressed through the mesh spreads beyond the open dot — more so as the mesh is coarser, the ink more fluid, the pressure heavier.

Result: a 50% dot on the film can cover 65-75% on the print. The entire tonal curve is distorted — and distorted differently in every shop. That's why the "standard" compensation values in manuals never quite work: they describe someone else's shop.

#The answer, by measurement

Calibration Flow doesn't ask you to estimate your dot gain: it measures it. The workflow is that of your first calibration, applied to screen printing:

1. Generate the target — 25 patches of known values, from white to black. Export it as a screened positive, like a normal film positive (polarity and halftoning).
2. Pull the target through screen printing — your screen, your emulsion, your ink, your pressure, your substrate. That's the whole point: the target travels through exactly the chain your images will travel through.
3. Photograph the dry print — even lighting, no cast shadow.
4. Import the photo — the app detects the target, reads each patch's real value and plots the drift: what 10%, 20%, … 90% actually became.
5. The compensation curve is generated — it pre-lightens exactly what your process darkens. It applies automatically to all your subsequent film positives, and exports as .acv for Photoshop (Pro mode) if you prepare your visuals elsewhere.

Recalibrate whenever a link in the chain changes: different mesh, different ink, different emulsion, different press. Curves are kept in a library — one per shop configuration.

#What it actually changes

• Highlights exist. Without compensation, everything above ~85-90% brightness prints pure white. The curve gives those areas back their dots.
• Midtones breathe. The "clogging" of the 40-70% range — where faces, skies, and materials live — is pre-corrected.
• Gradients connect. A gradient that "jumps" abruptly into a solid flat is the classic symptom of uncompensated gain.
• The run is reproducible. The curve encodes your shop; as long as the chain doesn't change, the result doesn't either.

#Honest limits

The curve compensates tonal drift, not mechanical defects: an under-tensioned screen, uneven exposure, or a worn squeegee produce variations that averaged measurement can't fix. If two prints of the same target give two very different readings, stabilize the process before calibrating — calibration freezes what it measures, instability included.

#Key takeaways