[sane-devel] Scannning trouble with HP 2520A

[Peter Kirchgessner]

Hi Brett,

does it make any difference if you select different speeds ?

You can also try to play around with the SCSI buffer size by changing 
environment variable SANE_SG_BUFFERSIZE.

--Peter

Brett (Mare) Henley schrieb:
> I'm still having trouble with what looks like some kind of
> data corruption when I scan wide areas with my scanner using
> relatively medium to high scan resolutions.
> 
> The problem seems to become worse when I use the higher bit depth
> aka 10 bits vs 8 bits per colour. It also gets worse if I increase
> the resolution or scan a wider area.
> 
> The way this manifests is displayed by lines of what looks like noise or
> interrupted data about 2 pixels tall all the way across the scan. The progress
> bar also pauses in its progress several times before it completes. The
> lines are horizontal. When the scan is done, it seems I can piece the image
> back together using gimp if I want.
> 
> I posted about this problem last week and didn't get a response so I'm hoping
> to hear from someone.  If you like I'll send a sample scan to whoever needs
> to see it. Don't really think sending it out on the list would be such a cool
> idea.
> 
> Below are some specifics for your consideration.
> 
> OS: FreeBSD 4.6.2
> Proc: 550 Mhz PIII, 512 MB ram
> SCSI:
> aha0 at port 0x330-0x333 irq 15 drq 6 on isa0
> aha0: AHA-1542CF FW Rev. C.0 (ID=45) SCSI Host Adapter, SCSI ID 7, 16 CCBs
> (the ATA controller is disabled on the motherboard this is the second IRQ and
> drq I've used. Neither had any conflics)
> 
> Scanner: HP 2520A (AKA HP scanjet 3C)
> scanning res 300 DPI - 1200 DPI (1600 DPI lets me get a clean scan for about
> a 1/4" in area. :) )
> 
> I've seen this problem with every front end I can find, include hptool etc.
> I've checked the scsi controller and cabling and tested the scsi subsystem in
> every way that I can.
> 
> I'm not against the idea that there's a chance the scanner could have a problem
> but it sure seems to work great in every other way I can find.
> 
> thanks in advance.
> 
> Mare
> 
> 
> 
> _______________________________________________
> Sane-devel mailing list
> Sane-devel@www.mostang.com
> http://www.mostang.com/mailman/listinfo/sane-devel
> 


-- 
Peter Kirchgessner
http://www.kirchgessner.net
mailto:peter@kirchgessner.net