[sane-devel] Super CoolScan Slide Feeders

[Alexander Oelzant]

On Sat, Aug 27, 2005 at 12:20:13PM +0100, Nigel Feltham wrote:
> I cannot help with info on the SF-100 but the LS-1000 scanner itself works 
> fine under SANE. 

Excellent, I just won one by accident on you-know-where (including the
SF-100, we shall see ...) :)

> The only Nikon film scanner to avoid is the original LS-10 as it's a real pain 
> to get it to work (the reason I gave up and upgraded to LS-1000) - I have one 
> and it's impossible to use reliably on either windows or linux unless using 
> an ISA SCSI controller (so needs an old PC) and is unsupported by SANE. 
> It does work intermittenyly under Vuescan on Linux on a PCI card (according to 
> Nikon's website it doesn't work at all under windows on PCI so one point for 
> Linux there) but only for actual scanning - automatic calibration fails. The 
> reason appears to be that it only supports SCSI 1 data speeds and PCI cards 
> try to transfer data too quick (they only support SCSI II and later) and 
> corrupt it.

My Tekram dc315 provides a parameter for selecting the speed:

> /sbin/modinfo dc395x
filename:       /lib/modules/2.6.11/kernel/drivers/scsi/dc395x.ko
author:         C.L. Huang / Erich Chen / Kurt Garloff
description:    SCSI host adapter driver for Tekram TRM-S1040 based
adapters: Tekram DC395 and DC315 series
license:        GPL
vermagic:       2.6.11 preempt K7 REGPARM gcc-3.3
...
parm:           max_speed:Maximum bus speed. Default 1 (0-7) Speeds:
0=20, 1=13.3, 2=10, 3=8, 4=6.7, 5=5.8, 6=5, 7=4 Mhz (int)
...

the last three of which seem to me quite SCSI-1, but the old asr saying
about SCSI chains and the necessity of three terminations (one being 
the black goat sacrificed to please the gods) still very much applies

> You may want to email the makers of Vuescan ( http://www.hamrick.com/ ) about 
> support for the Slide feeders  - the software's not free but does support 
> many scanners (mostly SCSI) that SANE doesn't as they have proprietry info 
> that the manufacturers won't give SANE coders access to (due to NDA's that 
> mean the resulting Source cannot be distributed).

Yes, I could still try to rever^Wuse that if all else fails.

thanks,
   Alexander

-- 
Alexander Oelzant (Durchlaufstr. 7/4/5, A-1200 Wien)
alexander at oelzant.priv.at aoe at fsinf.htu.tuwien.ac.at
       ex-internic, ripe, bofh, priv.at: !ao418
            +43 1 3500929 +43 676 84441065                                  McQ