On 2011-11-21 00:56, Olaf Meeuwissen wrote: > > No clue what capability 32829 is, but the scanner protocol command that > selects whether you use the flatbed or TPU option uses the same values. > That is, 0 will do a flatbed scan, 1 a TPU scan. > >> I'm not versed in c++, so I wonder if such a capability could be >> included in SANE? > It could, but I seriously doubt it should as it would drive the scanner > outside of its spec. > > Thanks very much for your reply.
I suspect that the scanner has a TPU capability limit is set lower since Epson must have assumed that people are not interested in scanning the full area in TPU mode (you are meant to use their fixtures to scan negatives etc. which all fall inside the standard TPU capability range). However, under Windows it is indeed possible to extend it. In fact the reason I found out about the 32829 was because I monitored what settings were changed when I launched Epson's own scanning interface (Two more did, but I never found any functionality connected to them). And changing this value to 1 does the trick/allows full size scanning. I've gotten many images that way, and I don't see any significant quality drop outside the standard max-range. Swithing it back to 0 and you get the smaller scanning area - all other settings equal. The setting could mean something like "use full sensor width/length". Unfortunately for me, my use of the scanner demands that I succeed in getting the full image. So I would very much appreciate having this feature included or directions for how I could build it myself. Else, I would have to surrender back to Windows and TWAIN again. Also, my brute-force solution to the lamp-issue below is by simply cutting power to the scanner. This requires me to work in Linux. So I'm in a bit of a tight spot here. Best regards, Martin Zackrisson