LaserSoft Imaging

Hi,
I finally installed 6.0.1.r19 after a boatload (not my first choice) of problems. Unfortunately, when scanning negatives I get mostly blue scans (no red or green channel). Sometimes the scan works, or I might get a red and blue channel but no green. This is only in negative scanning. I remember this coming up somewhere else in the forums and reading about a new .dll to fix it. Is this true? Otherwise, what should I do to get the negative scanning to work?

Another thing is the response speed. Maybe it's related to the same problem but the program is unbelievably sluggish. It takes as many as 2-3 seconds for input commands to register. Very difficult with sliders etc. (McAfee is off).

Negative scanning is also slow compared to version 5. Positive scans seem to take about the same time as in version 5.

This is a 600 mHz PIII running Win98SE with a Twain driver and ~370mB of ram.

NMoss

Soory about this but I can't seem to edit my last post. I'm using a 600mHz PIII with Win98SE and the Twain driver with ~370 mB of RAM
NGM

I have the same problem with this release of the Silverfast. I hope there is a solution.

Yes, I hope they fix it too although this appears to be a very capricious problem. Today I removed the program and reinstalled it but the problem remained. This is probably a negafix issue so I masked different regions of the image with more or less lightness but it always failed in the same way. Then I moved to a new frame and it scanned fine! Then I returned to the original frame and it scanned in fine too! This was after three days without coming close to a proper scan. It looks like a classic example of the programmers' nightmare going on here.
NMoss

nmoss wrote:Yes, I hope they fix it too although this appears to be a very capricious problem.

This is the behaviour I've been seeing with Cyberview and Filmscan 35, and it's why I ditched Cyberview for Silverfast. Please say they haven't broken it

Does reinstalling r10 get it back working?

The updated DLL is probably the SASPI32U.DLL that support will send out on request ... the latest builds should have it (r10 seems to have the right one, so r19 certainly should!)

Mike

Hi there,

I spent a while checking out the problems this weekend and came up with some fairly consistent findings. Firstly, it's clear that the program is very unstable, probably due to a bad and unpredictable reaction to specific features in each negative. It was always sensitive to black regions but this is a bigger problem. Firsly check "options" and make sure that the area used for color correction is set to avoid 10% or more of the edge. There was a problem with framing in the early versions of release 5 which left a black stripe at the top of each image. This was corrected in r22 but is back again. Avoid this area for color balance etc by the offset described above.
Another early version 5 problem was a yellow cast down the right side of the image. I had my RMA from PIE and was about to send it back when release 22 came out and the yellow cast went away! Now it's back.
The scratch and dust removal doesn't work correctly for masked areas. Only masked regions inside the last active "pane" of the scratch and dust window are corrected.

These are some of my current fixes to get some utility out of the program

1) If the scan comes up blue in the preview screen move the negative a little bit and try again.
2) If clouds etc come out pink, check whether the mask has been moved. I found that if the the scanned area on the preview is reduced after the image is set up colors can get very washed out and blue in the final image. When the the scan is set back to 100% things work fine again.

3) If the preview comes up OK then changes to violent combinations of green and yellow after the scanner head has retracted, "reset" the color histograms and things usually fall into place.

4) Sometimes the scan in the preview window looks good but the final image colors are messed up, too light and too blue. So far I haven't been able to fix this but I did notice that the color corrections were way out of the normal range in order to produce the normal looking preview and I think these settings are being passed on to the final scan. This is a very variable problem, each time the image is scanned a different set of color corrections are needed to make the colors look correct.
Stick at it LaserSoft

NMoss

nmoss wrote:I spent a while checking out the problems this weekend and came up with some fairly consistent findings. Firstly, it's clear that the program is very unstable, probably due to a bad and unpredictable reaction to specific features in each negative.

...snip...

1) If the scan comes up blue in the preview screen move the negative a little bit and try again.

NMoss

This is very consistent with what I'm realising. That moving the negative about can have a huge effect on the scan. When scanning a negative, the mechanical operation of the scanner is different, and resulting picture quality different, when the negative is shifted up or down a little. This is repeatable, now I've realised it's doing it.

Set to 3600 dpi (scanner is 1800dpi)

E.G. 1 :- Load a dark (overexposed) negative into scanner, closing frame down. "Zero" all controls and preview. Get a "banded" preview (horizontal stripes) and grainy picture. Repeat rescan, same. Hit scan, same is delivered to TWAIN app. Scan and preview are performed immediately after a quick warm up and travel down to top of neg. Sprockets are visible in the top of picture (nearly all of the sprocket). Bottom of picture is missing.

E.G. 2 :- SAME NEGATIVE - pushed slightly up ( about 1mm) by pressing against the edge of the neg and the plastic "sprung frame". Zero all controls still. Hit scan, no banding (or very slight), and much less grain.
Scanning is not performed immediately, but after a short delay. The head travels to the top of negative, and then moves back and forth very slightly, then scans at a LOWER speed. This is for scan or prescan. And the sprockets are not in view, and also the whole frame is captured.

I had about 6 goes trying to align a negative so that it would get a proper scan last night, and finally got it. You can't just load the negative and hit scan, it's not reliable enough ...

It's especially bad if you're near the end of a neg strip, don't let any light in "around the edge". That screws it up. It's also bad if the nice people at Kodak have put a dark stripe along the edge of your film through the sprockets. That affects the scan -- I know this because I had a troublesome one like this, and scanned it all upside down.

Can someone from Lasersoft answer WHAT Silverfast/Negafix is doing looking around the sprockets, and what is it trying to achieve?? Taking light levels? Colour balances of the "mask" because that's an area where you can guarantee it's mask only, and no image? Alignment of frame (i.e. where is top of picture, by finding the sprockets)

If we understood the process, maybe we could work with it, rather than against it.

Mike.

Yes, I made an interesting observation last night. A negative at the end of the strip gave a very bad scan after a reasonable looking preview. However, if I progressively pulled the negative out of frame (leaving more and more uninterupted light path through the scanner)the final scan of what was left of the image got better and better. This means that the program was doing well when a lot of unfiltered light was passing into the scanner, the opposite of what happens when a lot of opaque film is in the frame.
NMoss

I was very excited to see that the adjustment of the negative could fix the blue or green image that I receive on prescann, but I have an aditional problem... How do you move the negative when the scann head is frozen in place and the program has locked up?

Using a pIII 600 clocked to 1k (no problems on anything else with 70 images scanned) using 1800u and AI 6.0, 384 ram ,photo-shop 7 with more than enough hard drive to spare.

Problem... after reinstalling both photo-shop and AI 6.0 I continue to get the blue and green prescanns followed by a [b]total lock up [/b]of PS which locks the scann head in place. I can except the adjustment of the negative for better results but what is the total lock up about? I was able to scann many images before this became an issue and now I can not scann anything except by using the drivers supplyed by PIE.

[i][size=24][b]HELP!!![/b][/size][/i]

I haven't had that kind of lock-up for a long time, but when it happened I just switched off the scanner and rebooted the computer. The scanner then got its act together and pulled the scan head back into its shell.

A lot of the color problems have been fixed in the latest version of Silverfast (with no announcement or notice from LSI) and negative placement is no where near as critical as it once was. However, there is still a problem at the end of the strip that takes some fiddling. Reversing the film so that the cut end is on the left is often effective.

Recently I have encountered a new problem; SilverFast tells me that there is no scanner attached to the computer. I discovered that this only happens when I switch on the scanner after the computer has been booted. If the scanner is switched on before the computer everything is fine.

NMoss

Win 98SE / Twain / ~390 mb

Thanks for your reply but I have frequently rebooted and flipped the switch on the scanner... two things reset my scanner head 1 running a complete scann (causes complete lock up now PS plugin) so make that one thing resets the head using the original software.

If the answer is to reboot and reset than this is not even close to a working solution. I have the newest verson 6.0 AI yes the full version. I should have opted for the less expensive version as the bit depth is well in excess of a printers capability, then maybe I wouldn't feel so bad.

best wishes,

B

After digging through the FAQ section on this site I found a solution to the lock up .....delete the pref folder and start all over ...something seems to corrupt the prefferences. Have your serial nuber handy as you will need it.

As far as the blue and green screens that I had been recieving it is directly associated with any part of the negative not filling the scann window. Green comes when you scann with the lid up and blue with the lid down. When I adjusted the short negative (cut short in processing 35mm yes end of the roll) I still received the blue prescann.

SOLUTION
I added a strip of black paper next to the negative to cover the unexposed section of the negative (in between the next image) and found that the scann result was what I expected....a good scann with the extra part of the negative which was formerly black appearing absoulutely white.

Update>>>It seems that this may have as much to do with the negative lifting off the scann bed as anything and may not be a software issue but a hardware issue with the negative not being held in place properly at the end of the negative strip


Gotta get back to work I lost to much time figuring this one out

Bruce