[sane-devel] Re: Re: Re: epson 3490 - transparency unit problem

[patrick deflandre]

Levente Novák a écrit :
> On Tue, 2005-11-15 at 21:13 +0100, Oliver Schwartz wrote:
> 
> 
>>I've now adapted the quality calibration code for the 2480 and enabled 
>>it for the 3490. It's a pitty that you solved the problem already ;-) 
>>otherwise it would be nice to know if quality calibration solves it, 
>>too 
>>
>>The code is in CVS now, it should show up in the snapshots tomorrow - 
>>please test.
> 
> 
> I've tested the code with TPU and 3200 dpi (where I noticed a slight
> vertical banding at high contrast enhancements before the code change).
> It works well, at least I hear the scanner doing the calibration and the
> resulting image is better than without (banding almost invisible).
> 
> Ah yes, and 16 bpp colour at 3200 dpi works equally well!
> 
> Thanks for your efforts!
> 
> Levente
> 
> PS: today evening, I'm planning to run some thorough tests. Which
> resolution/bit depth/size combinations should I try?
> 
> 
I confirm the 3200 dpi works well for me, and there isn't any 
interlacing problems. I apologise for my previous message which was not 
enough clear.

In fact, it seem's that if I do a 2400 dpi scan which does not work, 
then the following 3200 dpi scan are bad. But it works again if I 
restart xscanimage.

But for me the 16 bit depth doesn't work. If I try to open the file with 
gimp, the following message is displayed :

PNM : invalid intensity maximum value


Patrick.