Cameron Harris schrieb: > On 10/30/05, Pierre Willenbrock <pie...@pirsoft.dnsalias.org> wrote: > >> >>Thanks for the log. >> >>Please try the attached patch. >>I don't know if it helps. The backend is trying to read 1560000 bytes >>from the scanner, but libusb does not provide that much. We need to find >>out where the data is lost. The patch addresses one possible cause. >> >>Regards, >> Pierre >> > > Ummmm.. well... good and bad news. > > The patch made it get past where it was stopping before, and now loads > of rubbish appears on the terminal when I do scanimage. I ran xsane to > see if it was scanning anything, hit acquire preview.
The "loads of rubbish" actually are the requested image. scanimage outputs on stdout per default. > > The scanner made a few noises, then went "beeeeeeeeeeeeeeeeeeeeeeeep" > until the scan was done. While it was beeping, the scanner head was > lit green. Did the head move? xsanes default is to scan in grey mode, which is translated into a monochrome green scan. > > The image produced was just grey lines on a black-ish background. One > good thing is that if I open the scanner lid mid-scan, it seems to > have an effect on the image appearing on screen. > > Sane Screenshot: http://htsc01.b3ta.org/scan.png > > Where it gets lighter is where I opened the lid, and darker again when > I closed it. > I guess the calibration step failed. Each scan should look like this: - (optionally)move head home - blink lamp - move head away from home - blink lamp - move head home - (maybe) blink lamp - move head away from home - blink lamp - move head home - do a short scan - move head home - do the real scan If the head moved in the "short scan" part, please compress and send me black_white_shading.pnm. That image should be generated while using scanimage with SANE_DEBUG_GENESYS=255. > I'll try getting it onto a Windows system and actually making sure the > scanner works properly :|, maybe I could get a USB snoop while I'm at > it, just in case. > Regards, Pierre