Legacy ImageMagick Discussions Archive

IM 6.5.9-10 Q16 Mac OSX Tiger

I was trying to read a pdf file provided by another user: http://www.martes.de/test.pdf

when I do

identify -verbose test.pdf
Unknown device: bmpsep8
Unrecoverable error: undefined in .uninstallpagedevice
Operand stack:
defaultdevice
identify: Postscript delegate failed `test.pdf': No such file or directory @ error/pdf.c/ReadPDFImage/638.


I have looked at this page:

http://www.gnu.org/software/ghostscript/devices.html

...
bmpsep8 Separated 8-bit CMYK .BMP file format, primarily for testing
...

and it says to do

gs -h

to get a list of devices and bmpsep8 is listed on my system.


Is this a problem with IM or with GS? Using GS 8.61

Other PDFs I can read and process with no problem. I also get the same problem reading eps files as well. Is this associated with cmyk bmp files embedded in the pdf or eps?

ImageMagick uses the bmpsep8 device for PDF's with colors in the CMYK or ICCBased colorspaces. If your version of Ghostscript does not support this device or if there is a bug in Ghostscript you will get an exception thrown. In the mean-time, for a workaround, add -colorspace RGB *before* the PDF image file name on the convert command line.

magick wrote:ImageMagick uses the bmpsep8 device for PDF's with colors in the CMYK or ICCBased colorspaces. If your version of Ghostscript does not support this device or if there is a bug in Ghostscript you will get an exception thrown. In the mean-time, for a workaround, add -colorspace RGB *before* the PDF image file name on the convert command line.

My GS reports that it does have bmpsep8 device. But I should have tried adding -colorspace RGB as I know that is needed when reading CMYK pdf. But when I do that, I get this

convert -colorspace RGB test.pdf test.tiff
**** Warning: File has an invalid xref entry: 2. Rebuilding xref table.

**** This file had errors that were repaired or ignored.
**** The file was produced by:
**** >>>> itext-paulo-155 (itextpdf.sf.net-lowagie.com) <<<<
**** Please notify the author of the software that produced this
**** file that it does not conform to Adobe's published PDF
**** specification.


So I assume the input pdf is not valid and have reported that back to the original topic. Nevertheless, it did produce a valid looking result!

Thanks for the help.