Why exactly 1 minute delay before scanning with xsane?

gchandler · 12-31-2016, 01:15 AM

I have a Canon FB630U scanner which works fine in a Windows XP VM in Linux Mint 18. However, when I try to acquire a preview or scan an image in Linux Mint 18, there is a sound from the scanner and then nothing for exactly one minute, at which point the scan is completed successfully.

Because the delay is constant and precisely timed, I am thinking the delay might be caused by some configuration setting somewhere, but have no idea where I might find it.

Can anyone suggest why this delay is occurring and how I might stop it? It makes the scanner almost unusable for more than a single page scan.

ferrari · 12-31-2016, 01:29 AM

Likely built in (intentionally or unintentionally) to the SANE driver behaviour for this device.

From 'man sane-canon630u' I note that it is mentioned

Quote:

NOTES

Due to Canon's unwillingness to provide scanner documentation, this
software was developed by analyzing the USB traffic of the Windows 2000
driver. So things like the calibration procedure I kind of made up; it
seems to work for my scanner. If you have complaints, let me know.

Quote:

SEE ALSO

sane(7), sane-usb(5), sane-find-scanner(1)
http://canon-fb630u.sourceforge.net/

AUTHOR

Nathan Rutman

A bug report may be the best course of action here
http://www.sane-project.org/bugs.html

ferrari · 12-31-2016, 01:33 AM

After reading through the man page, I also noticed

Quote:

/tmp/canon.cal
The calibration file used to normalize pixel brightness. This
is calculated every time the scanner is first used after it has
lost power. Deleting this file will force recalibration.

So, it does seem that it does a calibration after every power up. It's a shame that it can't write to ~/.config/canon.cal (or similar) and store this persistently. I still suggest a bug report for this reason.

gchandler · 01-01-2017, 07:33 AM

Quote:

So, it does seem that it does a calibration after every power up. It's a shame that it can't write to ~/.config/canon.cal (or similar) and store this persistently. I still suggest a bug report for this reason.

The repeat of calibration would only occur if power is lost to the device, but this delay I refer to occurs at every consecutive page scan.

But I will read those man pages you mention carefully and if I can't figure a fix will submit a bug report. Thanks for your follow up. It's appreciated.