Re: Nikon Coolscan / LS2000 /LS30 ?

From: dave-mlist@bfnet.com
Date: Tue Jul 11 2000 - 16:09:56 PDT

  • Next message: Petter Reinholdtsen: "Re: scanimage feature request (patch available)"

    Hi Andreas,

    Here is some more debug information for you for my LS2000.
    This time I tried scanning the same image twice, once with Scan Mode
    set to color (see http://bfnet.com/misc/gimp-coolscan-debug-color.txt),
    and once with Scan Mode set to RGBI (see
    http://bfnet.com/misc/gimp-coolscan-debug-rgbi.txt).

    The Scan Mode of color seems to work allright, except for the fact
    that sometimes the scan turnes out very blue. The RGBI scan doesn't
    return a picture at all because of a frame error.

    Andreas> Plase send me the debug output (unless it is more than 200kB,
    Andreas> in that case try scanning smaller images).

    The debug output is in these two URLs:
    http://bfnet.com/misc/gimp-coolscan-debug-color.txt
    http://bfnet.com/misc/gimp-coolscan-debug-rgbi.txt

    Andreas> Please indicate:
    Andreas> You SCSI card, SCSI lowlevel driver, sane version, backend version,
    Andreas> what settings you used with the backend and anything that may
    Andreas> give me an idea on how to reproduce your error.

    My system information is below. As for reproducing the error, I do
    the following:

    1. in shell, export SANE_DEBUG_COOLSCAN=128
    2. run gimp
    3. Xtns->Acquire Image->Device Dialog...
    4. Scan

    The result for Color is usually OK, but sometimes it's blue. The
    result for RGBI is an incomplete scan --- before the scan is finished,
    I see a window of mostly black with just a few colored blocks and
    dots.

    Here are some notes regarding the differences between the color and
    rgbi debugging outputs:

    1. Line 244 of the Scan Mode = Color:
            [coolscan] 5/44/0: Command Sequence Error
            [coolscan] sanei_scsi_cmd: returning 0x00000009
       this error does not appear in the RGBI debug.

    2. Line 388 of the Scan Mode = RGBI:
            xscanimage.input_available: bad frame format 5
       this error does not appear in the Color debug.

    Please let me know if there's any more information I can give you!

    SCSI card: Adaptec 2903B (came with LS2000)
    SCSI lowlevel driver:

    I have two SCSI cards, and my LS2000 is connected to scs1:

    [dave@print dave]$ cat /proc/scsi/scsi
    Attached devices:
    Host: scsi0 Channel: 00 Id: 01 Lun: 00
      Vendor: IBM OEM Model: DCHS04Y Rev: 6363
      Type: Direct-Access ANSI SCSI revision: 02
    Host: scsi1 Channel: 00 Id: 02 Lun: 00
      Vendor: Nikon Model: LS-2000 Rev: 1.31
      Type: Scanner ANSI SCSI revision: 02

    [dave@print dave]$ cat /proc/scsi/aic7xxx/1
    Adaptec AIC7xxx driver version: 5.1.28/3.2.4
    Compile Options:
      TCQ Enabled By Default : Disabled
      AIC7XXX_PROC_STATS : Enabled
      AIC7XXX_RESET_DELAY : 5

    Adapter Configuration:
               SCSI Adapter: Adaptec AIC-7850 SCSI host adapter
                               Narrow Controller at PCI 0/19/0
        PCI MMAPed I/O Base: 0xe1001000
     Adapter SEEPROM Config: SEEPROM not found, using defaults.
          Adaptec SCSI BIOS: Disabled
                        IRQ: 11
                       SCBs: Active 0, Max Active 1,
                             Allocated 15, HW 3, Page 255
                 Interrupts: 5281
          BIOS Control Word: 0x0000
       Adapter Control Word: 0x0000
       Extended Translation: Disabled
    Disconnect Enable Flags: 0x00ff
     Tag Queue Enable Flags: 0x0000
    Ordered Queue Tag Flags: 0x0000
    Default Tag Queue Depth: 8
        Tagged Queue By Device array for aic7xxx host instance 1:
          {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
        Actual queue depth per device for aic7xxx host instance 1:
          {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}

    Statistics:

    (scsi1:0:2:0)
      Device using Narrow/Async transfers.
      Transinfo settings: current(0/0/0/0), goal(0/0/0/0), user(25/15/0/0)
      Total transfers 2190 (2190 reads and 0 writes)
                 < 2K 2K+ 4K+ 8K+ 16K+ 32K+ 64K+ 128K+
       Reads: 0 0 0 61 0 2129 0 0
      Writes: 0 0 0 0 0 0 0 0

    Here are the revisions of my SCSI drivers from my RH6.2 kernel:

     * $Id: aic7xxx.h,v 3.2 1996/07/23 03:37:26 deang Exp $
     * $Id: aic7xxx.c,v 1.119 1997/06/27 19:39:18 gibbs Exp $
     * $Id: aic7xxx.reg,v 1.4 1997/06/27 19:38:39 gibbs Exp $
     * $Id: aic7xxx.seq,v 1.77 1998/06/28 02:58:57 gibbs Exp $
     * $Id: sequencer.h,v 1.3 1997/09/27 19:37:31 gibbs Exp $

    I am not specifying any kernel arguments at boot time.

    As for my SANE version, I'm using sane-1.0.2_coolscan043.tgz patched
    with your new coolscan_read_data_block() function from your mail dated
    Sun, 09 Jul 2000.

    My settings for the backend can be found at the following URL:
    http://bfnet.com/misc/coolscan:+-dev+-sgb.rc

    Andreas> Thanks for your help.

    Thanks for the code!
    Dave

    --
    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 : Tue Jul 11 2000 - 16:00:35 PDT