[Scribus] [announce] CMYK Patch gets image support

[Martin Costabel]

David Purton wrote:
> Hi folks,
> 
> I finally uploaded a first up attempt at getting CMYK images working.
> 
> http://marshwiggle.net/scribus/
> 
> There is a tarball containing test files and what I say is the "correct"
> output for comparison. 

Sorry for coming late into this discussion. I also apologize for not 
knowing anything about color management, profiles and such things.

But I have the suspicion that the participants in this discussion are 
all using little-endian systems, so my experiences with the patch on the 
bigendian MacOSX/Fink version of scribus might be useful. In short: they 
are not pretty.

I have looked at the files in cmyk-patch-test.tar.bz2 and I am seeing 
the following:

A. On unpatched scribus-1.2.1 with standard rgb and cmyk profiles:

1. If I open a new document and either import cmyk.eps or create picture 
frames and load any one of cmyk.eps, cmyk.tif, or rgb.tif, I get 4 
absolutely identical results, except that the imported cmyk.eps shows 
white vertical separation lines in the middle of the squares. The 
printer preview looks also almost the same. The colors are not identical 
to any of the cases in the screen.pdf or printer.pdf files, they are 
somehow "brighter".
2. If I open the cmyk.sla file, the last 3 cases where files are said to 
be loaded into picture frames, look identical but very different from 
case 1, but the first series that is said to correspond to imported 
cmyk.eps looks completely wrong: For example the first square that 
should show two shades of cyan, has a cyan half and a red half and so 
on, all squares show two different colors in the two halves.

B. On a patched scribus-1.2.1.cvs:

1. In a new document, the imported cmyk.eps, the picture frames with 
loaded cmyk.eps and with rgb.tif still look identical and almost 
identical to the unpatched case A. However, the picture frame with 
loaded cmyk.tif is completely wrong: Yellow becomes magenta, black 
becomes cyan, whereas cyan, magenta, red and green in the first row and 
blue in the second row all become white.
2. If I open the file cmyk.sla, the first series (imported cmyk.eps) 
looks for the first time similar to what it should, namely every square 
shows two shades of the same color. The other 3 series show different 
behaviors: The cmyk.eps and rgb.tif look almost like they did in A1, but 
the cmyk.tif has blue instead of light brown, red instead of black and 
black instead of red, and various shades of green instead of cyan, 
magenta and blue.

In conclusion: I hope that this patch does not make it into the official 
scribus sources until it is fixed for bigendian systems.

If I can make specific tests, please tell me, but I will not have the 
time to chase endian bugs in the sources myself this time.

-- 
Martin