On 11/08/2013 18:38, idle wrote:
> Hello,
>
> thank you for your replies.
>
>
> This model is listed as 'basic' in sane-backends 1.0.22. Try
> upgrading
> to at least 1.0.23, or a recent git snapshot.
>
> allan
>
>
> I'm trying now with the git snapshot:
> scanimage (sane-backends) 1.0.24git; backend version 1.0.24
>
> Unfortunately, the result is the same.
>
> Hello,
>
> if updated version doesn't improve the scan quality. You may
> also test if it is related to improper warming up by doing several
> scans. Clearing calibration (with the 'clear calibration'
> option)cache between each scan, so a calibration happens each time.
>
> Regards,
> Stef
>
>
> Thank you for the suggestion, I tried a series of 10 scans, clearing
> the calibration after each one, and the result is interesting. The
> first scan of the series is visibly lighter than the rest (although
> still too gray), the second and third are each slightly darker than
> the previous one. All the following scans are the same as the third.
> So there is something going on. The stripes are still there, though.
>
> Here are the first three scans:
> http://i.nahraj.to/f/s5o.jpg
> http://i.nahraj.to/f/s5p.jpg
> http://i.nahraj.to/f/s5q.jpg
>
> Eva
Hello,
it seems gain calibration isn't good, resulting in poor shading
calibration. Could you send the debug log generated when debug traces
are enabled by:
export SANE_DEBUG_GENESYS=255
export SANE_DEBUG_GENESYS_LOW=255
export SANE_DEBUG_GENESYS_GL646=255
Some *.pnm debug files will be created, please also send them.
Regards,
Stef
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20130811/911e4ca7/attachment.html>
