Summary: | -dFillOrder does not work for color tiff output devices | ||
---|---|---|---|
Product: | Ghostscript | Reporter: | Mike <mlungu777> |
Component: | General | Assignee: | Ray Johnston <ray.johnston> |
Status: | NOTIFIED FIXED | ||
Severity: | enhancement | Keywords: | bountiable |
Priority: | P4 | ||
Version: | 0.00 | ||
Hardware: | PC | ||
OS: | Windows 2000 | ||
Customer: | 631 | Word Size: | --- |
Attachments: | Patch to add BigEndian option to TIFF devices |
Description
Mike
2009-03-05 12:41:02 UTC
According to TIFF 6 spec:
> FillOrder = 2 should be used only when BitsPerSample = 1 and the data is
> either uncompressed or compressed using CCITT 1D or 2D compression, to
> avoid potentially ambiguous situations.
Currently, Ghostscript produces either monochrome 1 bit/pixel TIFF that
already honors FillOrder or 8 bit/sample TIFF where this flag is not applicable.
Please provide more information about the desired TIFF format or attach a sample
TIFF file with the required properties.
Perhaps, you just need to generate big endian format (aka 'MM') on a little
endian host ?
Alex, you are correct that is what is needed: generating big endian format (aka 'MM') on a little endian host. Created attachment 4844 [details]
Patch to add BigEndian option to TIFF devices
Attached please find a patch to add "BigEndian" option to the TIFF devices.
Use "-dBigEndian=true" to generate TIFF files with big-endian byte order
("MM"), and "-dBigEndian=false" to generate TIFF files with little-endian byte
order ("II").
When this option is not specified, native byte order will be used.
I guess there are TIFF readers that may not be able to handle MM vs. II endian issues, but since we didn't reject this request immediately, I have committed this patch. Thanks (and bounty) to Wendy. Since the MM/II order doesn't change the order of most of the raster data this doesn't really seem like a worthwhile change, but the patch itself demonstrates the correct method of adding parameters to a device, so it is a good example. Change committed as rev 9719. Changing customer bugs that have been resolved more than a year ago to closed. |