I am using a double core Pentium processor and set the NumRenderingThreads = 2 When I run the tiffsep device to generate the color separations I get invalid results. For some color I get all black, for some all white, and for some only part of the objects in the PDF. This happens on several files with more complex structure, images with spot color on top, images with transparency.
Created attachment 4254 [details] TestSample On this test sample tiffsep fails with Multithreading.
Created attachment 4404 [details] SimpleTiffSepTest.pdf
Created attachment 4405 [details] sep.s0.tif Created using: gswin32c -sDEVICE=tiffsep -dBATCH -dNOPAUSE -dTextAlphaBits=4 -dGraphicsAlphaBits=4 -r144 -sOutputFile=sep.tif SimpleTiffSepTest.pdf Ghostscript v8.63
This works for me with 904. Since this was reported some time ago and was against 8.63, it undoubtedly has been fixed. Note that the bug report was sort of suspect since it mentions 'multithread', but the command line given in comment #3 didn't specify -dNumRenderingThreads=2.
(In reply to comment #4) > This works for me with 904. Since this was reported some time ago and was > against 8.63, it undoubtedly has been fixed. Note that the bug report was > sort of suspect since it mentions 'multithread', but the command line given > in comment #3 didn't specify -dNumRenderingThreads=2. I have missed to specify the NumRenderingThreads as I copied my default usage flags. But the problem was indeed happening on multithreaded usage. Since this and other problems with PDFs with transparency I have stopped using Ghostscript. If I gather some time will check again rendering results.