[sane-devel] Fujitsu fi-5110eox scan area ignores pagewidth option?

[m. allan noah]

On 5/19/08, spammed <spammed at get2net.dk> wrote:
>
>  Hello SANE people,
>
>  I have a Fujitsu fi-5110eox ADF scanner that I use on Ubuntu 8.04 with
> scanimage and the fujitsu backend (both version 1.0.19). A couple of
> questions:
>
>  1.
>  The 5110 has a couple of paper guides that are adjusted symmetrically, so
> that the paper is centered with respect to the scanner's vertical center
> axis. According to the sane-fujitsu man page the pagewidth option is used by
> the scanner to determine centering of scan coordinates.

correct.

>  Am I correct in assuming that as long as you correctly set the pagewidth
> option, (tl-x, tl-y) = (0, 0) is always at the top left corner of the
> inserted sheet?

correct.

> Because that is not what I get, as seen from the attached
> example of scanning an A5 page:
>
>  scanimage --pagewidth=148 --pageheight=210 -t 0 -l 0 -x 148 -y 210
> --resolution=150 --source='ADF Front' > out.pnm
>
>  It looks as if the horisontal reference is at the extreme left of the
> scanning area.

so it appears. either this scanner does not actually support the page
size params, or there is a bug somewhere in the backend. both are
possible, since this is a lower-cost, non-twain scanner, and i'm only
human :)

>  2.
>  According to scanimage --help, the scanner accepts a --pageheight of up to
> 863.489 mm. Then why is the max value for -t and -y reported to be only
> 279.364 mm? (This value corresponds to 'letter' format?)

the hard-coded default in the backend is for a letter pageheight, but
if you set the height to something else, the t and y value max should
change too.

try this scan, all on one line:

SANE_DEBUG_FUJITSU=30 scanimage --pagewidth=148 --pageheight=210 -t 0
-l 0 -x 148 -y 210 --resolution=150 --source='ADF Front' 2>eox.log >
out.pnm

then compress eox.log, and send that to me directly. if we are sending
proper data to the scanner but it still does not work, you might have
to get a log of the windows driver in action, so we can see if the
machine uses some other technique. if the windows driver makes no
attempt to prune the scan area, then we will have to simulate in
software. yay.

allan
-- 
"The truth is an offense, but not a sin"