On Mon, Sep 19, 2005 at 09:07:22PM +0200, Julien BLACHE wrote:
> "Belgabor" <belga...@gmx.de> wrote:
> 
> > I also have to thank for this patch (it came on the day I boght the scanner
> > without knowing about the issues). It is almost perfect as it works nice for
> > resolutions up to 1200 dpi. At 2400 and 3200 I get a combing effect. Every
> > other pixel column is offset by several pixels. It looks a bit like an
> > interlaced PAL TV picture (only the combing is vertical rather than
> > horizontal). 
> > Hints appreciated (also if someone could confirm this. The error isn't there
> > in windoze, so it doesn't seem to be a hardware problem).
> 
> The backend needs to perform line distance correction on the image as
> for the 2480/2580 and an increasing number of scanners.
> 
> The high resolution are obtained by combining the data from 2 slightly
> offset sensors (lower resolution, cheaper), hence the offset you see.
> 
> JB.

Here is an image showing the effect we're talking about.
This is a scan of a 5x5mm area at 2400 in Lineart mode.
Note that the ratio is also wrong -- the x dimension off
by a factor of approx. 1.333333.

The combing also shows up in res. 3200.  Ratio at that res.
is correct.

bjd
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 2400-combing-ratio.pnm
Type: image/x-portable-anymap
Size: 38191 bytes
Desc: not available
Url : 
http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20050920/ed1c5f77/2400-combing-ratio-0001.pnm
From mattias.ell...@tsl.uu.se  Tue Sep 20 11:49:43 2005
From: mattias.ell...@tsl.uu.se (Mattias Ellert)
Date: Tue Sep 20 12:19:09 2005
Subject: ? about [sane-devel] Epson GT 15000 not recognized by SANE
In-Reply-To: <20050916044845.HGTX2438.fed1rmmtao06.cox.net@merlin>
References: <20050916044845.HGTX2438.fed1rmmtao06.cox.net@merlin>
Message-ID: <1127216983.19823.15.ca...@ellert.tsl.uu.se>

On Thu, 2005-09-15 at 21:48 -0700, Kent Braun wrote:
> I hope you don?t mind me asking. I?m trying to find out if there is a
> way to get the Epson GT-15000 to run on OS X. What OS are you running
> it under and how is it working?

> Kent Braun

Most of SANE works on MacOS X, unless it uses something very OS
specific.

E.g. USB scanners supported by the microtek2 does not work, since this
relies on a Linux kernel module.

Also closed source backends that link to propriety libraries usually
don't work, since these libraries are usually not available for MacOS X.
This is the case for the epkowa backend.

At the moment the epkowa backend is the only backend supporting the
Epson GT-15000 according to the SANE database, which means it is not
supported by SANE on MacOS X at the moment.

I don't know the status for planned support for this scanner in any of
the open source backends.

        Mattias Ellert

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 2260 bytes
Desc: not available
Url : 
http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20050920/b07f0366/smime.bin
From jsm...@suse.de  Tue Sep 20 15:02:09 2005
From: jsm...@suse.de (Johannes Meixner)
Date: Tue Sep 20 15:02:33 2005
Subject: [sane-devel] sane-frontends 1.0.14 xcam.c using uninitialized
        variable
Message-ID: <pine.lnx.4.58.0509201652490....@wotan.suse.de>


Hello,

I noticed in sane-frontends 1.0.14 that in src/xcam.c
the following code uses the uninitialized variable fd:
-------------------------------------------------------------------
static void
load_defaults (int silent)
{
  char filename[PATH_MAX];
  int fd;

  DBG (DBG_debug, "xcam, load_defaults: enter\n");
  if (make_default_filename (sizeof (filename), filename,
      dialog->dev_name)
      < 0)
    return;
  if (fd < 0)
    return;
  load_settings (filename, silent);
  sanei_load_values (fd, dialog->dev);
  DBG (DBG_debug, "xcam, load_defaults: exit\n");
}
-------------------------------------------------------------------

Kind Regards
Johannes Meixner
-- 
SUSE LINUX Products GmbH, Maxfeldstrasse 5      Mail: jsm...@suse.de
90409 Nuernberg, Germany                    WWW: http://www.suse.de/

Reply via email to