Skip to content

RIP and halftoning

The RIP (raster image processor) converts continuous-tone images into 1-bit fire/skip patterns. Colour conversion, ink coverage and dithering at this stage determine much of the final print quality.

The processing chain

  1. Resample. The image is scaled with a high-quality filter so its pixels line up with the print grid, target size in mm times DPI.
  2. Colour separation. RGB becomes the head's ink channels, which come from the head profile rather than being fixed: six on c6n90, four on c4n180. Neutral content is drawn from the black channel, grey-component replacement, instead of stacking C+M+Y, which wastes ink and shifts colour. Supply an ICC profile if you have one; without one a UCR/GCR conversion is used.
  3. Tone compensation and ink limiting. Printed density is not linear in requested coverage, because dots overlap and spread. Midtones are pulled down before halftoning, per-channel scales balance the inks, and a total-ink cap keeps the substrate from flooding.
  4. Halftoning. Each channel is dithered to 1 bit per pixel: Floyd–Steinberg error diffusion by default, or blue-noise masks, or ordered dithering.
  5. Pass slicing. The bitmaps are cut into interleaved passes (see Swaths and passes) and written out as the RIP payload the encoder reads.

Why resolution changes tone

Drop size stays the same when DPI changes, but grid spacing gets smaller. At 360 dpi, the pitch is about 70 µm; at 630 dpi, it is about 40 µm. With the reference drop size, a drop covers roughly three grid cells at 630 dpi. The same requested coverage therefore deposits more ink per area, and coverage above roughly 30% approaches solid.

Calibrate tone separately for each resolution. Reduce coverage as DPI increases, using measured per-channel curves where possible.

Calibration philosophy

To calibrate a print, print a target, measure the result and feed the measurements back into the RIP. Calibration parameters stay in the RIP; the firmware fires the patterns it receives.

The full workflow is in the rip-encoder module's docs/RIP_USAGE_GUIDE.md.

For the packing step that follows the RIP, see Printhead data protocol. For the CLI, see the RIP & Encoder component page.