Bug with Silverfast 6.0 2r.31? (negafix & color cast)

[harry shin]

1. First of all, thanks to LSI for working on correcting the problem with the last version of silverfast 6 (SDR problem)--> everything is working great now and I really like the SDR function; excellent.

2. The SDR problem was solved with the current version (6.0 2r.31), however I think there's now a NEW problem. When scanning the SAME negs that I have scanned several months ago, I'm now having problems with getting a mild-moderate reddish color cast to basically all of my images. I was thinking that it was perhaps due to a particular image, but not the case. I use negafix, and I've been using it for quite some time and thus I don't think that's the issue. Can someone at LSI check this problem out because it's basically driving me nuts. I also read a recent thread on the polaroid section about a similar type sounding issue re: IT8 callibration and potential bug with the current version of silverfast.
Again, the only reason I believe that there's some type of bug with the current version is because I'm getting a mild-moderate reddish cast to the SAME negs I scanned several months ago--and at that time, I did NOT have such a color cast--using an older version of silverfast 6.

3. Here's my data:
a. Mac OS 9.1
b. Polaroid 120


thanks for the help. Harry Shin

[harry shin]

Addendum to the above original post:

1. Yes, there is a bug with the current version of silverfast.
2. What I did was delete the current version of silverfast. Then I loaded the previous version (Silverfast 6, 2r20)--> I then scanned the reference negative I've been using for the past few days--> it scanned perfectly with negafix (ie no reddish color cast). Thus, I have to assume that there's some problem with the current silverfast version.
3. Now of course I could use the 2r20 version, but the problem with that one was the bug involving SDR. So, it would be greatly appreciated if someone could produce an updated version so that a) SDR will work properly b) negafix / color balance issues will work properly.


Thanks for all the help. Harry

[LSI_Kratzenstein]

Could you try to change the tolerance slider on the bottom in the expert part of Negafix. We set there more tolerance per default in the current version, to prevent some orangemask problems with some high noise scanner.

Kristian Kratzenstein

[harry shin]

1. in re: changing the tolerance slider, I did read about the suggestion to do that by the president of LSI. The bottomline is that it does NOT help the color cast with the current version, because something is wrong with the current version (yes I did try moving the tolerance slider). And again to reiterate, I believe that there is something wrong with the current version because last night I removed the current version, then installed the older version (I think it was 2r20)--> then scanned the SAME negative and no longer had the color cast issue. THUS, it seems to highly suggest that there's some type of bug with the current version of silverfast which is causing some color cast problem.


thanks. harry

[LSI_Kratzenstein]

When this problem cames up only with Negafix, then there is the solution. Use the tolerance to the left, use auto again (to calculate the orangemask again). We did change this in the current version.
Is the red cast also in dia scans ?

[harry shin]

1. Thanks for the suggestion; I will try it tonight however I have doubts that this will solve the problem, because I tried playing around with the tolerance slider and all it did was make the problem worse. Nonetheless, I will try it in a more systematic fashion and then report the results tonight.
2. Question: Is there any way you could elaborate in more specific terms and steps on how to optimally use the tolerance slider? The reason I ask is because I read LSI president's comments on how to use negafix and the tolerance slider (it's in the negafix discussion folder). I followed those steps exactly--> NO change with the color balance problem. However, it seemed that in your comments, one is first supposed to move the slider and check on / off / on the "auto" button again? I'm not sure how this is supposed to work and the president's comments never addressed this. Furthermore, from what I recall, version 2r20 does have the tolerance slider as well, however as stated before, I've never had to use it because I never had such color balance issues as I'm having with the current version.

Thanks. Harry

[LSI_Kratzenstein]

The problem is on scanner with higher noise. There the tolerance slider gives more control over the quality of the orangemask. You get more neutral pictures. So it was set for all scanner on a higher default.

[harry shin]

1. I did what you suggested--basically, it didn't help at all. However, here are the details of what I found:

a. With version 6.0 2r31, I scanned the reference negative, moved the auto tolerance up and down, on / off etc... It did NOT improve the color balance issue. Moving the tolerance slider to the right, did not improve the color casts--it just seemed to blow out the highlights. By the way, within the histogram tab, I also played around with the color cast slider--> did not improve the color cast problem either.
b. With version 6.0 2r20, I scanned the reference negative, NO reddish color cast problem. Thus, again from what I can tell, it appears that there IS some type of bug / problem with the current version of silverfast (6.0 2r31). The only reason I would like this fixed is because version 6.0 2r20 has a bug with SDR. Thanks. Harry

[JackConnick]

I'm having much the same problem. Using Negafix with Fuji Relia film.

Very orange cast, not even close, tolerance contril has no effect. I can dial it in a bit with the red slider control, but then it gets dark and I have to boost the whole neg to about 2.5.

I have fixed the scans in Photoshop with a click on auto levels, but this isn't the best solution. I had no problem with v20 with the old Negafix files. I think it is related to the Negafix new files.

Are the old files available still? I'd rather use them and don't have a copy.

This needs fixing.

Jack