> at second sight, I'm not so sure anymore, whether the gamma
> correction works right:
Please try the following:
edit umax-scsi.c: line 408 (function umax_send_gamma_data)
edit the line (408), change "256" to "length":
set_S_xfer_length(us->buffer, color*256+gamma_DCF2.size); /* bytes */
and move the line about 25 lines down, so that you get:
memcpy(dest, data, color*length);
set_S_xfer_length(us->buffer, color*length+gamma_DCF2.size); /* bytes */
umax_do_scsi_cmd(us->sfd, us->buffer, send.size+gamma_DCF2.size+length*color,
then do
make clean
make
make install (as root)
and please tell me what happends!
> As far as the 10-bit option is concerned, I am not so sure, what
> exactly the effect of turning it on should be. It doesn't cause
> any crashes anymore (these were obviously also only whith
> gtk+970925) but the results look pretty weird. While the button
> is checked, most scans are accompanied by message boxes popping
> up saying either "Error during read: the Operation was cancelled"
> or "cannot handle depth 16". When unchecked again, the very next
> scan will produce another such message, then things are back to
> normal again.
"cannot handle depth 16" is a problem of the frontend that is not
able to use more than 8 bits per value, that may be ok for me,
but "Error during read: the Operation was cancelled"
sounds not so good. Please could you try it again after
you did the change described abouve!
then do
export SANE_DEBUG_UMAX=128
and try it again and please send me the output
Thanks
Oliver
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com