QTR-Quadtone RIP

Hi Tom

Tom Moore wrote:
> .. It just occurred to me that the underlying
> assumption (that the C/M and LC/LM densities are close enough to permit
> swapping) implied that one could make a normal (i.e. not split tone) warm
> curve, linearize it and then create a cool curve by simply swapping inks.
> That means less work and therefore interesting - to me at least.
> 
I think you could. As I said the densities are similar albeit not 
identical. For example, with my set up, at 2880dpi, 2100 on PJ Alpha, 
55% ink limit

C = 1.30
M = 1.35
LC = 0.78
LM = 0.81

> What puzzled me was that the warm and cool curves (for HPR, for example) are
> different. I got off the rails a bit when I mentioned toner curves - it was
> late and I was tired. However, between the warm and cool curves, the
> partitioning values are slightly different and the linearization values are
> quite different. The curves for EEM have similar differences.
> 
The different linearisations are probably a consequence of the different 
crossover points. There's nothing sacred about those points. In fact I 
would suggest that these curves are not optimal because the 
linearisations are rather acute - for example the cool Photo Rag curve 
linearises 50 at about 30 - quite a shift. I'm not sure which axis is 
input & which output. I usually try to get the curves a little closer to 
ideal before linearising. But that's a different subject.

> My question was, therefore, have you (or anyone else) actually tried doing
> this - either to create a split-tone curve in the manner you suggested or to
> create a warm from a cool curve without re-linearizing.
> 
The same principle applies to the UT-FS & UT-FSN sets (and I would 
imagine R2W & R2N although I've not used them). I've created many 
split-tone sets with UT-FS/UT-FSN on the same principles. If you're 
interested to see them let me know. They're unlikely to be of any 
practical use as I wouldn't imagine anyone would duplicate my particular 
combination of inks & ink positions.

Howard