I made a coating of 6% gelatin (grocery store generic brand), 1% sorbitol and 5% McCormick's yellow food coloring (FD&C 5; a.k.a. tartrazine) as an extinction dye.
This produced a coating that is quite orange, somewhere between a yellow and orange b&w contrast filter. I suspect that it will be very good for keeping the relief shallow, and the addition of dichromate will further the stain.
I'll try to post an image soon.
After exposure, it will be etched in hot water (shooting for 115°F), cooled in room temperature water (68°F), cleared in sodium bi-sulfite and then "set" in 91% isopropyl alcohol.
All but the last step are common to the carbon process, but it is recommended by Friedman and others to dehydrate the gelatin relief, thus immediately "plastering" it down on the base. This helps in the all important drying step to limit any exudiate, which basically means no dissolved gelatin will be squeezed out while it dries, leaving gelatinous material that could cause streaks and spotting in the dye print. An alternative to the alcohol is sodium sulfate.