Public Access
Left alone, coverage follows whatever the photo happened to be, which came out around 44% and printed heavier than wanted — thermal dots spread as the paper heats, so they land fatter than they look on screen. The brightness offset needed to hit a target is solved for rather than searched: error diffusion preserves mean tone, so the first guess is analytic and up to three correction passes handle the clipping. Measured 33.6% across normal, heavily darkened and heavily brightened versions of the same photo, in under 6 ms, so badges now print at a consistent density regardless of lighting. Set targetCoverage to null to opt out.