[gs-devel] Colour Separation output

[Ken Sharp]

[resent, omitted gs-devel initially...]

At 03:27 10/01/2008 -0800, you wrote:


>email size, but have the PDF inlcuded in the zip) With these results, 'I' 
>am expecting a black output, and what I mentioned in the original e-mail, 
>but am now thinking that what if this is an RGB PDF,

I opened the file with Acrobat Prodfessional 8 and used the Preflight tool. 
All 4 images are RGB, as is the text, and the two filled objects. Looks to 
me like this is in fact an RGB PDF file.

>  would this mean the RGB would be processed at CMY, and please correct me 
> here if this is wrong?

No, the RGB will be processed as RGB. However, if you print to a CMYK 
device, the RGB needs (obviously) to be converted to CMYK. The method for 
this conversion is given in the PostScript Language Reference Manual. It is 
also affected by the undercolorremoval and blackgeneration in the graphics 
state.

To summarise the spec, when converting from RGB to CMYK:

Cyan = 1.0 - red
Magenta = 1.0 - green
Yellow = 1.0 - blue

However, because (as you know) equal values of CMY do not usually produce a 
good black, normally tackled by subtracting the smallest of C,M.Y from each 
of the CMY values and making that the black channel. The full conversion is 
given as:

C = 1.0 - red
M = 1.0 - green
Y = 1.0 - blue
K = min(C,M,Y)

Cyan = min(1.0, max(0.0, C - UCR(K)))
Magenta = min(1.0, max(0.0, M - UCR(K)))
Yellow = min(1.0, max(0.0, Y - UCR(K)))
Black = min(1.0, max(0.0, BG(K)))

UCR is the undercolorremoval procedure, BG is the blackgeneration 
procedure, which are defined as PostScript procedures.

So... depending on what you have UCR and BG set to, the result will be 
different. Possibly this is not correctly set for your expected output.


However...
I suspect the confusion is really due to the ProcessColorModel of the 
device. I'm no expert here, hopefully Ray can help further, but I notice 
that if I create a CMYK TIFF (using tiff32nc), and split the resulting file 
into channels in Photoshop, I do indeed see a black channel. Whereas If I 
just use tiffsep I get correctly named files where the black channel is 
empty and the gray values in each separation do not quite match the 
separated composite.

It looks to me like the tiffsep device is preserving the original RGB as 
CMY data, and not converting it into CMYK, even though it is naming the 
files as CMYK and inverting the RGB values. Presumably this is deliberate, 
so that RGB files can also be separated.


                             Ken

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
A company is known by the people it keeps.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -