[sane-devel] genesys backend calibration

Robert A. Schmied uwppp at flash.net
Wed Jun 6 19:57:40 BST 2018


genesys backend expert(s)

is there any public documentation relative to the design
and implementation on the genesys backend calibration functions.

i've got a canoscan lide 200 (as i've written before) and the
default calibration results in a better central area about 3 cm
wide in each scan line -- i can provide any scan data requested,
just give me the parameters to use.

i've found that changing the 'lines used in shading calibration'
near line 2424 in genesys_gl847.c is assigned like this:
    dev->calib_lines = dev->model->shading_lines;

from the default value of 50 to 200 and then perform a calibration
cycle using a blank sheet i get a substantially better result in
subsequent scans.

so i'm curious about how the genesys backend calibration functions
where developed including the default values that define the
calibration chart.

since i have yet to try to disassemble the scanner to actually
see the calibration chart which might (or might not) exist under
the top edge alignment guide i'm thinking it might be faster,
better and safer to implement a special scan of that area itself.
then i thought i wonder if something like that has already been
implemented ...

i'm thinking scan from x=0 to x=max with ylines at home or parked to
ylines = 3 cm (or some other convenient value).

this should probably be performed using an existing calibration file
and performed only as a test case.  it is even possible the various
image files generated with envvars SANE_DEBUG_GENESYS_GL847=255
SANE_DEBUG_GENESYS=255, SANE_DEBUG_GENESYS_LOW=255 actually do some
of this, but is undocumented ...

i can provide these files as well if requested.

comments, active discussion welcome/desired.

aloha

ras



More information about the sane-devel mailing list