stripes in Avision-USB scans Was: [sane-devel] Re: Weird vertical stripes on Microtek ScanMaker X12USL (sane-1.05)
Wed, 14 Nov 2001 13:44:16 +0100
Is it possible that the stripes in the Avision-USB scans appear, because we do
no yet do any calibration? (The SCSI ones do not need / and even to do support
it). - But they are really black and very regular ... ???
Hm. Now when I watch to these two pictures they are not as black as the ones I
got via eMail. Seems to be clear that the USB scanners _simply_ need software
calibration and software color pack (color channel displacement).
I will do this in the next days ...
On Mon, 12 Nov 2001 06:25:15 +0100 (MET)
Novak Levente <email@example.com> wrote:
> On Sun, 11 Nov 2001, Michael Thomson wrote:
> > Hi Henning
> > Thanks for your tips - I've successfully built & installed
> > sane-backends-1.06 and sane-frontends-1.06 as well as xsane-0.80 now.
> > Sadly I'm still getting weird stripes, but I found a previous discussion on
> > this and I've mailed the folk involved to see if they made any progress.
> > To answer your question, sane-1.05 came precompiled with my distribution:
> > Yellow Dog Linux 2.1 from Terrasoft (www.yellowdoglinux.com)
> Usually if you get vertical stripes in scanned images, it means that there
> is no (or there is a wrong type of) shading correction. Every CCD or CIS
> is composed of photoelements slightly different in characteristics, so
> usually the scanned images are not directly converted to images but
> corrected for lamp and sensor unevenness.
> I had similar problems with Microtek Phantom 636cx, until Karsten Festag
> corrected the shading function in microtek2. Since then, everything is
> working fine.
> The only strange thing in your image is that the stripes are not dark like
> in my case but coloured. It seems to me it is a wrong (buggy?) shading
> correction rather than a non-existant one. Or a communication problem
> between the scanner and the computer which corrupts the data?
k33p h4ck1n6 René
René Rebe (Registered Linux user: #127875 <http://counter.li.org>)
Anyone sending unwanted advertising e-mail to this address will be
charged $25 for network traffic and computing time. By extracting my
address from this message or its header, you agree to these terms.