> Abel Deuring <a.deuring@satzbau-gmbh.de> is also getting
> some impressive speedups with his SnapScan using larger
> buffers and "pseudo" command queuing.
I'm working with a Sharp JX250, not an Agfa Snapscan :) But the
remaining is right: the usage of (pseudo) command queueing, large
buffers, a separate reader process and shared memory for the data
transfer between the reader process and its parent avoids many, if not
all, carriage stops with higher resolutions, at least with the Linux box
to which my JX250 is connected (P166, 64 MB, Adaptec 2940). On the other
hand, it seems that the JX250 has not the largest internal buffer, or a
poor buffer management. The scan time decrease from more than 20 seconds
to less than 10 seconds (A4, grayscale, 400 dpi) may be a somewhat
special situation, compared with other scanners.
Abel
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com