LaserSoft Imaging

Hello,

several weeks ago I reported a phenomenon related to silverfast AI 5.5 software which is still very annoying:

In high resolution scans (not in previews), about to 20 stripes (single color per stripe, 1 pixel width (x direction), about 1000 pixels high (y direction) of different colors and random positions appear (also visible after reloading such scans after saving).

My systems:
-----------
'new PC': ASUS A7V Via KT-133 board, 1000 MHz Athlon Socket A, 3x256 MB Infineon RAM PC-133 cl2, Win98, Graphics board: Elsa Erazor MX board (Elsa or vidia drivers).

'old' PC (for comparison): Gigabyte board, BX440 chipset, Celeron 333A, 2x128 MB PC-100 RAM, Graphics board: Matrox mystique (PCI)

These stripes are only seen with the new PC when scanning with AI 5.5.0r14 (as photoshop plugin or twain module (but not with Nikon Scan or Nikon twain import into photoshop). Same phenomenon with Kodak Imaging (42 to 24 bit twain import).


I excluded the following suggested causes:
--------------------------------------------
1. Win98 is not the cause:
a) AI works well with the same software (win98 and imaging programs) on the old PC. b) Windows 2000: still stripes

2. RAM is not the cause (new PC equipped with RAM from old PC: still stripes appear)

3. ISDN card (Fritz card, PCI) not the cause: installed in both PCs

4. Graphics board not the cause: Stripes also when Matrox card is used in new PC

5. Fire Wire card and driver not the cause: used in both PCs

When scanning with Nikon Twain, scan sounds continuous, no mouse action possible

When scanning with AI 5.5 (twain or PS plugin): scan stop and go (every second), mouse actins possible (opening new files etc.), but obviously not related to position and appearance of stripes (stripes even without mouse movement).

The stripes may be related to differences in handling of interrupts or other processes on the new computer (Nikon seems to take over the control).

I see this problem being related to AI 5.5 (possibly in connection with hardware).


Any ideas?

Guenter

Dear Guenther,

usually the scanning programs do not "take
over" the PC when scanning. This sounds to
me like an interrupt conflict (no mouse action possible....).
I'd check whether there are any conflicts with the firewire card.

Bye

Andreas

don't turn on the scanner when film is inside, hope this could be the cause, scanner needs to see perfect black when turned on.

Guenter:

You have the Nikon Jaggies. They are caused by discrepancies in the way the drivers are able to control the stepper motor. Ed Hamrick's v**s**n works around this in an undetectable manner. From what I understand, Nikon licensed the code from Ed, and used it in NikonScan 3.1.

I do wish LSI would have some comment on this problem, since (along with the NegaFix/Nikon blues) it renders the software useless with Nikon scanners.

Les

Since yesterday Nikon Scan v3.1.1 is available for downloading from Nikon's European site

http://www.nikon-euro.com/nikoneuro2/do ... d_107e.htm

Dear Mr. President_LSI, dear AI5.5 / LS400 users,

upon my posting (see first mail ofthis topic) I received some answers, but none of the proposals could cure the bugs.

Bug: LS-4000 system. Small stripes (single pixel width, about 300 to 1000 pixels high, different colors) appear with AI 5.5 (not with NikonScan).

My system: details see first mail (Athlon 1000, Via133 chipset, FireWire, AI 5.5.1r03). Even with the newest AI 5.5 release (no matter if installed prior to or after NikonScan 3.1.1 installation on a pristine Win98 system) still shows this behaviour.

I tried AI5.5 as Photoshop plugin (PS 5.5 or PS Elements) and as Twain module (PS / Elements, Kodak Imaging). In any case, the Nikon Twain module did never cause any stripes, but the AI software consistently (!) caused stripes.

Mr. President: could you please comment on any differences between AI5.5 and NikonScan in:

2. control of LS-4000 hardware

2. data transfer handling

3. handling other tasks running during scanning? The way the software handles e.g., mouse events seems to be substantially differnt between AI 5.5 and NikonScan 3.1.x. For instance, mouse double clic into PS / Elemeins window causes opening of file selection window during an AI 5.5 scan, but never during a Nikon scan (but remember: even without touching the mouse stripes appear with AI 5.5)

This problem is very annoying, and LaserSoft should solve it, or I would like to give my AI 5.5 software back for refunding (or get a HDR version to do the 16 bit raw scans with Nikon 3.1.1).

Guenter

Guenter,

You have the stripes, I have problems with slides and the SA21, furthermore there are problems with the SF200.
We probably all use Nikonscan, Silverfast and v**s**n. I'am very fond of negafix but thats about everything right now. I'am very pissed about the results on slides, never mind the colormatching blabla, the interface is no good the control over the scanner is *********! And the software is very expensive! Ed Hamricks v**s**n is only $45,- and conrols the scanner without a glitch. I have a LS2000 running without any problems on SF and the LS4000 is since it was supported by SF having problems with the SA21 and slides, while Vscan and Nscan never showed these problems. Why is SF many times more expensive as Vscan still not able to fix these problems which do not occur in V and Nscan.
SF hire Ed Hamrick to solve your problems!

Xander

Dear all,
please check SF ver. 5.5.2

Dear Mr. President,

since half a year, the problem of stripes in scans (in every scan) exists (see my previous messages).

Any release of Nikon Scan 3.1 (3.1.0, 3.1.1 and 3.1.2) works fine (no stripes).

I also installed every new release of Silverfast AI 5.5 (up to 5.5.2r08), and with any of these releases I have the same problem (Photoshop plugin and Twain)!

The problem may be related to the chipset/processor (Via KT133 / AMD Athlon 1GHz; for more details ,see my previous mails.

Did Lasersoft check the scanner/firewire card with such a system??


I hope this problem will soon be solved!

Guenter

Dear Lasersoft team,

after replacement of the FireWire card the scanner works fine with SF AI 5.5.2 and Nikon 3.1.x (the Ratoc card which came with the scanner was exchanged for an EXSYS EX-6450 card).

But, nevertheless, the old card worked well with Nikon 3.1.x.

Guenter