Scanning Gum for RPN / ChartThrob negs

I've got a new inkjet printer with a new inkset and I'm experiementing (happily) with using a bleach step in my gum prints. In other words, I need to create all new digital neg curves.

I'm working (for the first time) with the RPN & ChartThrob method. Seems to work great. Now (finally) my question:

When working with a bright color of gum (say bright yellow or red), should I convert the scan to grayscale before reading it in ChartThrob?

In Platinum or Palladium I'm getting real logical curves. In gum I'm getting curves that need a fair bit of "smoothing".

Thanks,
Tom

First things first. It R-N-P (Ratios are Not PDN) ;^)) [Hey, I need a clever recursive acronym. And a wink and a nod to Mr. Clay Harmon]

Gum can be tricky to curve. But your question is a very good one. How best to represent pure colour on a scale for purposes of calibration. Let's take Yellow as the example. Yellow, by itself, is really "hard on the eyes" as far as trying to calibrate it visually. So by all means you want to examine it as a greyscale where your eyes can make some sense of it.
Pure yellow can be made in RGB space by adding 255 red and 255 green and 0 blue. Blue is the complimentary to yellow. When examining a yellow gum ChartThrob scan if you change channels in your levels histogram all, or most, the information is found in the blue channel. The combined RGB channel, the normal view will add in information from the other channels but the bulk of the information will still be from the blue channel. So does it matter? Not so much is my answer. You could run the curve twice to prove it once as a greyscale made from the combined RGB channels and then another just made from the greyscale and see if there's much of a difference but since you're dealing with "relative" changes between the steps I'm guessing that the changes on both would be proportional in their respective spaces.
~m

tomf said:

I've got a new inkjet printer with a new inkset and I'm experiementing (happily) with using a bleach step in my gum prints. In other words, I need to create all new digital neg curves.

I'm working (for the first time) with the RPN & ChartThrob method. Seems to work great. Now (finally) my question:

When working with a bright color of gum (say bright yellow or red), should I convert the scan to grayscale before reading it in ChartThrob?

In Platinum or Palladium I'm getting real logical curves. In gum I'm getting curves that need a fair bit of "smoothing".

Thanks,
Tom

Click to expand...

Thanks for the guidance. A wiser person than I would only change one variable at a time..... having lots of "fun". I'll start by just doing a fair bit of smoothing to the gun curves and see where that gets me. I'm "wondering" if my scanner has too much of a "texture fetish" with the watercolor paper I'm using for gum. If I still have "issues" I'll try a friend's scanner and see if I get more even results.

An update..... Calibrating is going well and (to somewhat answer my own question) converting the scan to grayscale or using color for Gum negs in ChartThrob's analyze doesn't seem to matter. The curves are virtually identical.

Indeed, as mkochsch suspected, the one truly "odd" curve that needed the most smoothing was my primary yellow for tri-color work. It is a very strong and very saturated yellow. I think the main problem is my scanner. The red channel is clipping in some of the yellow squares.