You are not logged in.

#1 2018-01-19 18:19:11

Michael Gruenstaeudl
Member
Registered: 2014-12-03
Posts: 51

[SOLVED] Distorted scans - Sane 1.0.27 breaks CanoScan LiDE series

For several years, I have successfully used the Desktop scanner CanoScan LiDE 200 with my Arch Linux system. Upon a system update in 2017 (pacman -Su), the images produces by this scanner are distorted (for lack of a better word). See https://pasteboard.co/H3Gfmts.png for an example. Particularly annoying is the central black vertical bar, that happens to be somewhat translucent.

Several factors lead me to believe that these distortions are based on some driver or SANE configuration issue:
First, the very device works flawlessly on Ubuntu and Windows systems.
Second, older and newer devices of the CanoScan LiDE series work flawlessly on my current Arch system.
Third, the distortions appeared immediately after the said system update.

I would appreciate suggestions as to which  driver or SANE configuration files the system update (or I) could have inadvertently modified to produce said distortions (and recommendations to fix the issue).

Last edited by Michael Gruenstaeudl (2018-01-22 15:39:17)

Offline

#2 2018-01-19 19:21:35

pypi
Wiki Maintainer
Registered: 2014-04-22
Posts: 250

Re: [SOLVED] Distorted scans - Sane 1.0.27 breaks CanoScan LiDE series

This has been reported upstream and appears to be a problem with the genesys driver ( https://bugs.gentoo.org/635348, http://www.sane-project.org/sane-backen … #S-GENESYS, https://alioth.debian.org/tracker/index … tid=410366 ). Your best bet is building from source (https://git.archlinux.org/svntogit/pack … kages/sane) with the patch mentioned in the Debian bug reverted.

If you haven't built packages using makepkg before, you have some reading to do - check out https://wiki.archlinux.org/index.php/Makepkg and https://wiki.archlinux.org/index.php/PKGBUILD as starting points.

Offline

#3 2018-01-22 15:36:39

Michael Gruenstaeudl
Member
Registered: 2014-12-03
Posts: 51

Re: [SOLVED] Distorted scans - Sane 1.0.27 breaks CanoScan LiDE series

pypi wrote:

This has been reported upstream and appears to be a problem with the genesys driver ( https://bugs.gentoo.org/635348, http://www.sane-project.org/sane-backen … #S-GENESYS, https://alioth.debian.org/tracker/index … tid=410366 ). Your best bet is building from source (https://git.archlinux.org/svntogit/pack … kages/sane) with the patch mentioned in the Debian bug reverted.

If you haven't built packages using makepkg before, you have some reading to do - check out https://wiki.archlinux.org/index.php/Makepkg and https://wiki.archlinux.org/index.php/PKGBUILD as starting points.

Yes, the solution proposed by pypi worked perfectly. I wrote a patch (see below) to revert the changes that were introduced by the problematic patch referenced above (i.e., https://alioth.debian.org/plugins/scmgi … 956c8ce242), inferred the md5sum for my new patch, updated the PKGBUILD file and installed via makepkg -si.

The patch to revert the changes that were introduced by the problematic patch is as follows:

--- a/backend/genesys.c	2017-04-09 03:40:32.000000000 +0200
+++ b/backend/genesys.c	2018-01-22 16:04:42.731907170 +0100
@@ -2069,12 +2069,9 @@
   /* turn on motor and lamp power */
   dev->model->cmd_set->set_lamp_power (dev, dev->calib_reg, SANE_TRUE);
   dev->model->cmd_set->set_motor_power (dev->calib_reg, motor);
-
-  /* if needed, go back before doin next scan, by using rewind, registers and
-   * slopes table are kept intact from previous scan */
-  if (dev->model->flags & GENESYS_FLAG_SHADING_REPARK && dev->model->cmd_set->rewind)
+  if (dev->model->flags & GENESYS_FLAG_SHADING_REPARK)
     {
-      status = dev->model->cmd_set->rewind (dev);
+        status = dev->model->cmd_set->slow_back_home (dev, SANE_TRUE);
     }
 
   status =

Thanks to pypi. Marked as solved.

Offline

Board footer

Powered by FluxBB