[sane-devel] Scanning from fujitsu ScanSnap S1500 Error during device I/O

Simon Matter simon.matter at invoca.ch
Fri Feb 4 19:08:51 UTC 2011


> On Fri, Feb 4, 2011 at 12:16 PM, Simon Matter <simon.matter at invoca.ch>
> wrote:
>>> No, Markus has uncovered a real bug. The S1500 chokes when doing high
>>> resolution duplex scans. The backend reads data from both sides of the
>>> paper, and buffers the back side. Unfortunately, we read too far ahead
>>> on the backside, and the scanner has problems.
>>
>> I have no idea if this could be related but I have recently tested an
>> S510
>> which one of our users was using at home. He said the scanner doesn't
>> work
>> fine anymore. I thought of course you can not compare to the bigger ones
>> like fi-6230 bunt when I tried to scan duplex at - I think it was 300dpi
>> -
>> it was veeeeeeery slow. It felt like being 1 or 2 minutes and that
>> really
>> seems to much :)
>>
>> What I remember is that it took half of the A4 paper quite fast and then
>> waited a long time, scanned a bit more, waited and so on until the page
>> was finished. The output file itself was looking fine.
>>
>> Any ideas if that's related somehow, at least I guess the S510 and S1500
>> may be from the same "family".
>
> I think perhaps there is a relationship. You said "doesn't work fine
> anymore", which implies that it did work fine, and then something
> changed. Perhaps an upgrade from sane-backends 1.0.20 to 1.0.21?

Hm, all I remember is that initially we bought a number of scanners, most
were fi-6230 and fi-6130 but also a single S510 just to test. I remember
that the S510 was slow compared to every other model, but I don't remember
it being so slow. You are right, we upgraded from 1.0.20 to 1.0.21 since
installing it. I have no idea when it started to be slow because the
device was used in a home office and I think it didn't scan much there.

>
> I have made some changes in the duplex handling code in the past year,
> which appear to resolve Markus' problem in the git version. They might
> also improve the pausing issue you described. I still do see some
> pausing at high resolution with the fi-6110, but nothing like minutes,
> only 10-20 seconds.

Well, I really don't know exactly how much time it took to scan the A4
page in duplex, but it felt like 1 to 2 minutes. Even if it's less, I
don't think it's normal.

>
> Perhaps one of you guys could test a current git snapshot? we are
> preparing for release, and I'd like to fix it first.

Good idea, I'll prepare an rpm anyway to test with our others scanners.
I'll let you know how it works.

Regards,
Simon




More information about the sane-devel mailing list