Comic book drawer & colorist

Workflow

He scans his drawing @ 1000 dpi.
Cleans it with curves/levels filters.
Optionnaly, if it's not already done, he inks the drawing with a simple hard brush (but most artists still prefer to ink with real ink).
He converts it to bitmap, and save it as a copy that will be used by the printer for the black channel.
Undo the bitmap conversion, resize the layer to 300 DPI, set it to multiply composite mode. New layers will be created under that layer.
Create layers for each plan/objects and paint the silouette of that object. Here we have several possibilities depending of available features :
- Group layer + mask
1 mask for shape, 1 layer for base color, other layer used for shading and details (for example one for shadow, one for lightings, one for highlights and effects).
- Clipping mask
1 layer for base color + shape, child layers used for shading/details.
- The Lock layer transparency feature is used a lot, it permits to color with big brush without painting "outside" and without using a mask or selection. Maybe it could be a composite mode too.
When the coloring is finished, remove the Line layer.
Send the Ink and Color image to the editor or printer.

That workflow may not look very artistic, but the color style has to be kept for at least 42 pages. There should be one color per layer if possible, gradients and mixing will be made with pixel transparency or transparency masks. That technical workflow pay off when a color can't be printed, it's easy to modify a color without affecting others colors.

Format & constraints

EPS/PSD/TIFF (and PNG ?)
CMYK.
For French/Belgian comics, the standard is :
1000 dpi bitmap image for lines (sometimes a regular greyscale 300dpi file).

300 dpi CMYK image for colors.
3 - 60 layers depending of the style (The user will merge layers of distant objects/characters when the program is too slow anyway).
Fast color palettes docker (maybe the quick palette from Vera will be useful for this case ?).

Missing features

+++ Preview of out of gammut colors in the image (then we can mix colors and use a more artistic workflow).
+++ Preview of out of gammut colors in the color selector (or a warning, or hide them).
++ Lock layer transparency.
+ Scanning (not that important, because the image can still be scanned outside Krita).
+ Clipping-mask (it make things a bit easier, but technically, you don't need it).