[sane-devel] xsane vs. scanimage when using HP ScanJet Pro 2000 s2

[Ellis Cohen]

Summary:  how to make 'scanimage' produce the "clean" output that I get when using 'xsane'.

Details:

I have an HP ScanJet Pro 2000 s2.  Connection is via USB 3.

Running
 * Fedora 33
 * xsane-0.999 (c) 1998-2013 Oliver Rauch
 * scanimage (sane-backends) 1.0.31; backend version 1.0.31
 * HP Linux Imaging and Printing System (ver. 3.20.11)
Context is having documents of varying heights (e.g., 8.5x11 and 8.5x14) in the document feeder.

*When I use 'xsane':*
**everything "just works" -- I get mutiple 'tiff' files of different sizes (on the order of 1-2M), each containing a suitably-sized image of each document.  I can view those images with 'gimp'; the only diagnostic (from gimp) is
Non-conformant TIFF: extra channels without 'ExtraSamples' field
and a query about how gimp should process the extra channel.

*When I use 'scanimage'*:
if I don't specify '-y <value>', it appears to default to the max for this scanner -- 3098.8mm.  The output files are about 25M in size; the output of 'file' on the result is
 * TIFF image data, big-endian, direntries=16, height=36600, bps=0, compression=none, PhotometricIntepretation=RGB, orientation=upper-left, width=2550
If I try to view the 'scanimage' result in 'gimp', I get the diagnostic messages
 * Bogus "StripByteCounts" field, ignoring and calculating from imagelength
 * Read error at scanline 3248; got 0 bytes, expected 7617
Message repeated <lots> of times
-----------------

*So, my question:* 
What parameter(s) can I pass to 'scanimage' so it will "auto-size" its output (as 'xsane' is clearly doing)??

Thanks for any suggestions!

Ellis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://alioth-lists.debian.net/pipermail/sane-devel/attachments/20201231/9bc24ecc/attachment.html>