Nikon LS4000, AI 5.5, Dual Processor is Unstable

[Michael Gay]

I've not found any problems using Nikon Scan 3.1.2 with the LS4000 and a dual Athlon 1600 on a Tyan MP S2460 board and I've scanned about 20 negative films. Most were batch scanned to disk at full resolution from the Nikon window, but some were scanned, a few at a time, using the scanner under Photoshop 6.0

Note that this is the third release, not the second, the original 3.1 crashed and so did the next release 3.1.1

Silverfast 5.5 crashes and I now presume that this is because it uses earlier Nikon software.

Incidentally I received good support for Nikon Scan from Nikon in Switzerland, but they didn't seem to think that Silverfast was their problem. They told me that the company wrote all its own software.

[schweikl]

No, Nikon Scan 3.1.2 is simply crashing during starting. None of the earlyer versions was so useless to me.
About Nikon Sevice: I don't know about their quality. I just can say it took months until I got a reply asking for my system configuration and since then I didn' hear any more.....
This is exactly 2 weeks ago now. I am expecting to get an answer in June.

Btw. I am in direct contact with Eric and it's looking good.

[LSI_Flyvbjerg]

Dear All.

I implemented a workaround, that avoids the chrashes. As Richard Schweikl suggested I force SilverFast and the Nikon SDK to use only one CPU. This might result in slower scan performance, but it works. With my dual PII 400 MHz system, a 1350 dpi scan with the LS40 took exactly as long as with two CPUs. Mr. Schweikl reported that a 4000 dpi scan with a LS4000 took 16% longer time (3:15 minutes instead of 3:00). I think that should be acceptable.

Updates will be available on our update site soon. The version number of SilverFast with the fix will be v5.5.2r08 or higher.

For user, who have an older version and don?t want to upgrade to v5.5, the fixed version will be v5.3.2r08.

Eric.