16bit greyscale not working on big endian
Posted: 2014-07-16T07:21:09-07:00
Hello,
16-bit greyscale images (fits or tiff) are not handled properly on Solaris/SPARC which is big endian. On Solaris/x86 (little endian) and Linux/x86 there is no problem. Also 8-bit greyscale images are not affected.
The problem was introduced somewhere between imagemagick 6.7.3.10 and 6.8.8.-7, it is also still present in 6.8.9-2.
Edit: I also found the problem in 6.3.4 Q16. All affected versions were "Q16" versions, the versions that worked, weren't "Q16" versions.
To reproduce the problem:
- view a 16-bit greyscale image (fits or tiff) using display
- convert a a 16-bit greyscale image to a 8-bit format (e. g. as png) and view it
In my case the fits-image is displayed (or converted to) completely black and the tiff-image to grey noise.
Regards
Kester
16-bit greyscale images (fits or tiff) are not handled properly on Solaris/SPARC which is big endian. On Solaris/x86 (little endian) and Linux/x86 there is no problem. Also 8-bit greyscale images are not affected.
The problem was introduced somewhere between imagemagick 6.7.3.10 and 6.8.8.-7, it is also still present in 6.8.9-2.
Edit: I also found the problem in 6.3.4 Q16. All affected versions were "Q16" versions, the versions that worked, weren't "Q16" versions.
To reproduce the problem:
- view a 16-bit greyscale image (fits or tiff) using display
- convert a a 16-bit greyscale image to a 8-bit format (e. g. as png) and view it
In my case the fits-image is displayed (or converted to) completely black and the tiff-image to grey noise.
Regards
Kester