Patrick,
Another update: following your advice that tuning is where I should
start, I made some big progress today, so thanks for the tip!
I started by looking at what values your automatic tuning is setting for
the hardware I'm testing on.
When I first use the device, the initial tuning seems to consistently
get set at:
gain=0x23, dcoffset=0x32, vrt=0x09, vrb=0x0F
If I then repeatedly bring up the device by doing enrollments in fairly
quick succession, the automatic tuning values seem to stabilize at:
gain=0x23, dcoffset=0x32, vrt=0x08, vrb=0x0E
I'm guessing the small difference in vrt and vrb is probably the result
of the hardware warming up.
Then I went through many iterations of manually setting values for vrt
and vrb to see if I could find values that would produce better results.
Eventually I came up with these values, which now have the driver
working quite will on my hardware anyway:
gain=0x23, dcoffset=0x32, vrt=0x14, vrb=0x16
I'm not sure these vrt, vrb values are necessarily "perfect" just yet,
but they are close enough to make the etes603 quite usable.
So on this particular hardware, your current automatic tuning seems to
set vrt a fair amount lower than it needs to be, and vrb somewhat lower
than it should be.
Which of course raises the question of how to correctly automatically
tune the revision of hardware I'm working with without breaking the
tuning that worked correctly on the revision of hardware you originally
developed the driver against.
Out of curiosity, do you know what automatic tuning values you would
typically get when you originally developed the driver?
Thanks for all your help!
-Jason
_______________________________________________
fprint mailing list
fprint@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/fprint
