[sane-devel] HP ScanJet 7400c on Mac OS X (TWAIN-SANE interface v. 2.0, sane-avision backend)

[Moritz Heckscher]

Hi all,

I'm trying to get a HP ScanJet 7400c to work on a Mac OS X platform  
using SANE as provided by Matthias Ellert's TWAIN-SANE packages, and  
I'm having some problems.

Do others also have problems with this particular scanner (not  
necessarily on Mac OS X), or does it usually work fine?

The SANE backend I'm using is the newest (1.0.18). The OS recognises  
the scanner correctly on the USB port.

"sane-find-scanner" usually detects the scanner correctly:
> found USB scanner (vendor=0x03f0 [hp], product=0x0801 [hp scanjet  
> 7400c]) at libusb:002:002-03f0-0801-00-00

I say usually because sometimes it doesn't. I haven't figured out a  
pattern, but I believe it takes a few minutes after plugging in for  
the scanner to be recognised (although its display indicates long  
before that it is ready). Either that or I need to unplug and replug  
the power and USB connection. As I said, I haven't found a pattern.

"scanimage -L", however, takes very long (almost a minute), and also  
doesn't always successfully find the scanner. I've enabled debugging  
output with "export SANE_DEBUG_AVISION=7" in the shell. In a  
succesful procedure, the scanner seems to be recognised almost  
immediately:
> [avision] attach: Inquiry gives mfg=hp, model=scanjet 7400c,  
> product revision=0.8S.

A lot of communication then follows between the scanner and the PC.  
This takes very long, as described above, and finally ends with (in a  
succesful case):
> [avision] sane_get_devices:
> device `avision:libusb:002:002-03f0-0801-00-00' is a Hewlett- 
> Packard ScanJet 7400c flatbed scanner
> [avision] sane_exit:

I'm attaching the complete log for a succesful case to this message  
(HP7400c-succesful.txt).

So my question is: Is it normal for "scanimage -L" to take this long?  
I've used another scanner with the gt68xx backend before, and there  
the command worked almost instantaneously. And is it normal for the  
command to sometimes work and sometimes fail (although "sane-find- 
scanner" has detected the scanner)?

In cases when it doesn't work (although "sane-find-scanner" has  
detected the scanner), there seems to be a problem with attaching:
> [avision] sane_init: config file line 24: trying to attach USB:`usb  
> 0x03f0 0x0801'
> [avision] attach:
> [avision] attach: opening libusb:002:002-03f0-0801-00-00
> [avision] attach: open failed (Invalid argument)

I'm attaching the complete log for an unsuccesful case as well  
(HP7400c-unsuccesful.txt).

Once the scanner is found by "scanimage -L", I still have problems.  
Usually I can scan from the command line with "scanimage", but I  
can't in GUI applications. Usually the GUI just hangs and doesn't react.

I've debugged the TWAIN-SANE interface (version 2.0) a bit and have  
found the problem to occur in line 832:
> /tmp/TWAIN-SANE-Interface.build/SaneDevice.cpp:832: failed  
> assertion `status == SANE_STATUS_GOOD'

I've successfully used the scanner *once* in "Image Capture.app", in  
all other cases it failed, and it never worked in "Graphic  
Converter.app".

My initial guess was that the problem could be the long delay  
described above. But maybe it's a problem in the TWAIN-SANE interface  
with getting some options of the scanner (buttons? ADF?) to work?  
Unfortunately, my debugging capabilities aren't that great, so I  
couldn't track this further.

Any help is highly appreciated!
-Moritz

P.S.: I intend to give it a try in a few days on a machine with  
Ubuntu once I have access.

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: HP7400c-succesful.txt
Url: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20070805/976fb7fa/attachment-0002.txt 
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: HP7400c-unsuccesful.txt
Url: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20070805/976fb7fa/attachment-0003.txt