Experiences with Scan Dual II and SANE

From: Clark Case (ccase@rocketmail.com)
Date: Fri Feb 16 2001 - 14:09:18 PST

  • Next message: j.steindlberger@gmx.de: "Re: HP5300C - USB-problems?"

    I found a fellow Minolta Dimage Scan Dual II / Linux
    user, one Jose Paulo Monitinho de Almeida
    (moitinho@civil.ist.utl.pt) who had had some limited
    success in getting his scanner to work with SANE, and
    I've been trying to duplicate his efforts. I thought I
    would post the results I've gotten so far, and try and
    get some suggestions on moving forward.

    I'd seen posted that the Scan Dual II is actually made
    by Avision, and used SCSI commands sent over the USB
    bus. The post said it might be possible to use the
    usb-scsi driver that is being developed and a modified
    avision backend to drive my scanner. So, I downloaded
    the 2.4.1 kernel source and the patches posted on the
    HP 5300C development site
    (http://www.neatech.nl/oss/HP5300C).
    The patch for the config module config file didn't
    work cleanly, so I manually edited that file. The make
    file patched fine. The usb-scsi.c file didn't
    contain the vendor ID and product ID for my scanner
    (0x638 and 0x26a,
    respectively), so I added them. I changed the
    references to HP in the avision backend patch to
    Minolta. I compiled and installed the kernel and
    modules, booted with the new kernel, and compiled the
    SANE stuff. After turning on the scanner and doing a
    "modprobe usb-scsi", I had my scanner in
    /proc/scsi/scsi. After doing an "export
    SANE_SG_BUFFERSIZE=32768", find-scanner was able to
    see my scanner. I made a link from /dev/sga to
    /dev/scanner, and, after waiting for the green light
    to stop blinking on the front of the scanner, ran
    xscanimage. As this is a film scanner, not a flatbed
    scanner, I reset the size from A4 dimensions to 1cm x
    1cm, hit scan, and nothing happened. I had set the
    SANE_DEBUG_AVISION to 10, so I was able to see some
    debugging messages. It had apparently sent the scan
    command, but the scanner didn't do anything.
    Xscanimage hung up at this point.

    I also tried scanning an image in Windows, and then
    booting into Linux, while leaving the scanner on and
    the film caddy in place. Jose said that that had
    worked for him, but I didn't get any different
    results. I've also been experimenting with VueScan. If
    I just freshly boot into Linux in insert the usb-scsi
    module, vuescan just hangs when I try to run it. If I
    do the "scan an image in Windows" trick, vuescan
    starts, but clicking on the Preview or Scan buttons
    doesn't do anything. Trying to exit caused the program
    to hang up, and tons of messages from the usb-scsi
    driver to get sent to my syslog. This probably happend
    when I killed xscanimage as well, I just didn't check.
    Also, when I try to reboot after trying to scan, my
    system can't disable my usb mouse, and I have to do a
    hard reboot. Perhaps I should be doing this with a
    regular mouse?

    Well, that's where things stand now. If anyone has any
    suggestions for moving forward, I'd certainly
    appreciate hearing them. I've already email Ed Hamrick
    with my experiences in VueScan, to see if he has any
    ideas there. It would seem that the avision backend
    was written with flatbed scanners in mind, and
    probably doesn't have the commands built in for
    transporting the film caddy, focusing,
    etc. We will probably need information from Avision or
    Minolta for doing this stuff. Does anyone have any
    contacts in these companies that may be able to
    provide the data we need?

    Thanks,

    Clark

    __________________________________________________
    Do You Yahoo!?
    Get personalized email addresses from Yahoo! Mail - only $35
    a year! http://personal.mail.yahoo.com/

    --
    Source code, list archive, and docs: http://www.mostang.com/sane/
    To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com
    



    This archive was generated by hypermail 2b29 : Fri Feb 16 2001 - 14:04:58 PST