[sane-devel] Canon MP700 support (partially OK)

[Nicolas Martin]

Le lundi 19 juillet 2010 à 23:11 +0900, Gernot Hassenpflug a écrit :
> Dear list,
> 
> I'm currently testing a Canon PIXUS all-in-one MP700 which is similar
> in structure to the MP710 I had until recently (power failed, thrown
> out).
> 
> I am using CVS, as shown by "scanimage -V":
> 
> scanimage (sane-backends) 1.0.22git; backend version 1.0.22
> 
> 
> Results:
> ------------
> 
> Here the apparent capabilities as exposed to Xsane & scanimage:
> 
> Modes: Color & Grayscale (no lineart, i.e. B/W)
> 
> Resolutions: 75, 150, 300, 600, 1200 dpi (appears to be correct for
> max resolution for this device)
> 
> I could successfully test color all resolutions. 300dpi I went up to
> A4 size, then 600 & 1200 half a page and 1/4 page to save time. All
> worked flawlessly.
> 
> For Grayscale, I could work perfectly with all resolutions *except* 1200 dpi.
> 
> Problem: for 1200dpi, the scanner starts to scan, perhaps 0.5 cm
> movement, and then it hangs, light dimly on, and power button no
> longer responds. With scanimage I can use Ctrl-C to cancel the scan
> and the scanner recovers properly:
> 
> ^Cscanimage: received signal 2
> scanimage: trying to stop scanner
> scanimage: sane_read: Operation was cancelled
> 
> Looking at the partial image scanned at 1200dpi shows that it is
> probably not in the right format, as there are diagonal stripes to be
> seen (the image as such is not recognizable).
> 
> I had a look at pixma_mp730.c and it is not clear to me if this is a
> known error or owing to untested resolution & mode perhaps? I can
> state from memory that I did not have this issue with the MP710 which
> seems to follow the same protocol.
> 
> On the other hand, the man page for the pixma protocol lists the
> Multipass MP700 and 730 rather than PIXMA MP700 & 730, so perhaps
> there is some slight difference between the models depending on where
> they are sold? PIXMA, PIXUS, Multipass...?
> 
> Logs:
> --------
> 
> The log is too large to attach here, so I am sending a tgz version to Nicolas.
> 
> The log shows that suddenly the scan values become "ff" and in fact
> the USB logs with SANE_DEBUG_PIXMA=21 shows that the last line does
> not complete scanning, but immediately goes to some kind of error.
> This repeats (see below) until I press Ctrl-C and cancel the scan.
> Recovery is good and I can use the scanner without having to power it
> off and on again.
> 
> [pixma]  00003d70:86 80 7b 6f 6a 6a 6d 72  7a 78 7a 78 74 77 7f 7e
> [pixma]  00003d80:7e 75 6d 69 6e 77 78 7a  79 75 74 76 7d 7d 80 7b
> [pixma]  00003d90:75 70 6d 71 78 78 7a 78  79 7c 82 83 84 77 75 6f
> [pixma]  00003da0:6b 70 75 78 7a 76 74 6f  73 77 79 72 6d 54 3a 35
> [pixma]  00003db0:30 30 ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
> [pixma]  00003dc0:ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
> [pixma]  00003dd0:ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
> [pixma]  00003de0:ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
> [pixma]  00003df0:ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
> [pixma]  00003e00:ff ff ff ff ff ff
> [pixma]
> [pixma] OUT  T=7.369 len=10
> [pixma]  00000000:d4 20 00 00 00 00 00 c0  04 00
> [pixma]
> [pixma] IN   T=7.372 len=6
> [pixma]  00000000:06 06 00 00 00 00
> [pixma]
> [pixma] OUT  T=7.475 len=10
> [pixma]  00000000:d4 20 00 00 00 00 00 c0  04 00
> [pixma]
> [pixma] IN   T=7.478 len=6
> [pixma]  00000000:06 06 00 00 00 00
> [pixma]
> [pixma] OUT  T=7.581 len=10
> [pixma]  00000000:d4 20 00 00 00 00 00 c0  04 00
> [pixma]
> [pixma] IN   T=7.584 len=6
> [pixma]  00000000:06 06 00 00 00 00
> 
> Any comments and things to try most welcome,
> 
> Regards,
> Gernot Hassenpflug

Unless I'm wrong, 1200 dpi is not offered basically by Canon's drivers
on their Pixma models, so I guess the scanner may or may not work
correctly using these resolutions. MP700 is also a rather old model, and
it looks performances at greater resolutions have been improved somehow,
on more recent models. 

However, sudden change in scanned data ressembles to some hardware
discrepancy, did you try maybe with another computer, is this
reproducible ? 

Nicolas