Hi Ralph, Ralph Sontag schrieb: > I have an Canon CanoScan LiDE 500F. > > I've added in backend/genesys_devices.c a section as a copy from > the LiDE 60 - ok, now scanimage list the device: > >> scanimage -L > device `genesys:libusb:001:050' is a Canon LiDE 500F flatbed scanner > > But it don't work.
The LiDE 35/40/50/60 are very similar(not to say they have identical hardware). Your LiDE 500f on the other hand has some differing properties. > I have a Windows 98 running in qemu. The scanner works in this emulator- > environment, an I got some debug-messages with usbsnoop: > > http://www-user.tu-chemnitz.de/~sontag/sane/usbsnoop.log.1.gz I only decoded the above log. It looks like a calibration run. First the driver does some tests on the registers, then on the memory. After this, it does the calibration. This should be enough to make the scanner deliver data using scanimage. > http://www-user.tu-chemnitz.de/~sontag/sane/usbsnoop.log.2.gz > > This are scans with color, 300 dpi. > > > _After_ working with qemu also scanimage will produce output: > http://www-user.tu-chemnitz.de/~sontag/sane/scanimage-scan-grey.png > > scanimage then also produces log-files: > http://www-user.tu-chemnitz.de/~sontag/sane/scanimage-debug.log > > Without initialisation under W98, the scanner don't work with scanimmager > - so it seems, there is at first a problem during initialisation. > It hangs at > [genesys] sanei_genesys_read_register (0x41, 0xf0) completed Please try the attached patch. It may already get scanimage past initialization. If not, please also send a debug log without win98 initialisation. > At second the data is wrong: the picture shows only noise. This is to be expected. The AFE(Analog FrontEnd) chip is not initialised correctly and the gl841 probably latches digital data at wrong times. To further complicate this, your scanner uses a CCD-chip, while mine has a CIS. > I have not enough experience to interprete the log-files, so I ask, what > to do next. If you want to, i can send you some perl scripts and instructions on how to use them. I don't understand all details of the protocol, so there are some "unknown"s. > I will try to get some debugging-output with Windows XP and will try to > start usbmon for debugging with linux. At least i don't need logs from linux. The sane backend already dumps most of the relevant information, and does a lot things different to the windows drivers. Logs from win98 are as good as from winxp. Regards, Pierre -------------- next part -------------- A non-text attachment was scrubbed... Name: canon-500f.diff Type: text/x-patch Size: 3312 bytes Desc: not available Url : http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20060506/2ca88d11/canon-500f.bin From i...@quantum-sci.com Sat May 6 21:15:57 2006 From: i...@quantum-sci.com (Quantum Scientific) Date: Sat May 6 21:17:38 2006 Subject: [sane-devel] Negs Scanning Pure Red In-Reply-To: <87lktfaqmc....@frigate.technologeek.org> References: <200605061000.55912.i...@quantum-sci.com> <87lktfaqmc....@frigate.technologeek.org> Message-ID: <200605061415.58789.i...@quantum-sci.com> Thanks. This does help a bit. But it seems this pure red for very black is intended as a warning that blacks are crushed. I know they're crushed, and I need to recover them as best as possible. This red crap just makes for hours of repair work to remove it. On Saturday 06 May 2006 12:40, Julien BLACHE wrote: > Quantum Scientific <i...@quantum-sci.com> wrote: > > Hi, > > > Running Debian Linux, Gimp for photo editing, and XSane with the > > Microtek2 profile for scanning, and Sane 1.0.15. > > You can try to upgrade SANE (and XSane maybe) using the Sarge > backports maintained by Aur?lien Jarno. > > See <http://people.debian.org/~aurel32/> for instructions. > > JB. >
