[sane-devel] how should you determine a good gamma correction to use through the backend?

[Christopher Marshall]

I have been using Sane for about two months now with a CanoScan LiDE 80 and getting great results.

I have been wondering, however, about how best to enhance my images after scanning.

They have been coming out fairly dark with no corrections.  I have dealt with this in two ways.

For scanning text documents, I set the backend --contrast argument to 30 and have no complaints.

For photographs, however, setting --contrast to 30 is a poor choice compared to using the GIMP to
brighten the midtones with the image->color->curves tool.  I add a control point in the middle of
the "value" curve and move it up and to the right and the image gets very close to the original
with this.  However, I can still see differences in the colors when comparing the image on the
screen to the photo I scanned.

I just read up on gamma correction and it has hit me that I was approximating a gamma correction
the way I was using the gimp curves tool in the value channel.

So what I am wondering is this.  What is a reasonable procedure for figuring out a good gamma
correction to give to the backend when scanning photographs?

Might it make sense to print a test page with solid blocks of red, blue, and green colors spread
across the 0,255 range, scan that page with sane,  and then use the Gimp to get the exact values
of the colors, then calculate the best gamma correction from that?

Has anyone ever done this?

Chris Marshall


__________________________________
Do you Yahoo!?
Yahoo! Finance: Get your refund fast by filing online.
http://taxes.yahoo.com/filing.html