Strange bug in SF 5.5 for Nikon LS-4000

[Andreas]

Since quite a while i had a minor problem that SF absolutely refused to scan a very small part of the slides I tried scanning in HDR-mode with 16x scan. Apparently the AI decided that these were not worth scanning... The funny thing is that the whole is absolutely reproducible. Presently I have two slides that are absolutely impossible to scan. SF always tells me "Scannerfehler 0" (5.5 R14).

If Lasersoft is interested in finding out what the problem is, I'd happy to _borrow_ a slide for analysis, however, as I like it more than SF does, I'd need it back...

After upgrade to R 18 the problem got really bad - it was not possible at all anymore scanning in the mode mentioned above.

After uninstalling and reinstalling R 18 problems persisted. However, after going back to R 14 it's again just the 2 slides that do not work. Everything else is fine.

BTW SF 5.5 R14 is running under XP and does fine (5.2 did not at all) under the new OS except for these slides...

Andreas


<font size=-1>[ This Message was edited by: Andreas on 2001-11-29 15:12 ]</font>

[Andreas]

Really nobody has any ideas on this?

[ilyons]

Andreas,

Ae you using Dust and Scratch removal filter with these images amd a mult scan of 4 or higher?


BTW: i don't think it is possible to use multiscan with 48 bit HDR colour.


Ian

[Andreas]

Dear Jan,
yes, I'm using the optimum for my archiving purposes (5.5 R14):
16x scan, ICE and HDR (48 bit).
However, I never found any hint in the manual that this would not work. In fact, it works fine for 95% of the scans. The ones that make trouble are the ones the LS-4000 prescans (exposure measurings) 2 or 3 times. Apparently it does this in case of images with little contrast sometimes. I assume that silverfast runs into a timeout problem waiting for a response from the scanner which takes longer than usual to focus and to measure the exposure. BTW, the whole works fine in Nikon Scan 3.1 (not under XP, though....
Lasersoft Support told me that the problem would be fixed with release 5.5 18, however with this release the problem in fact really started. Now the settings really do not work anymore. Under 5.5 18 it's really almost impossible to scan a slide using 16x scan, ICE and HDR (48 bit).

I'd be grateful for a comment and help from Lasersoft.


Bye

Andreas

[ilyons]

On 2001-12-06 01:32, Andreas wrote:
Dear Jan,
yes, I'm using the optimum for my archiving purposes (5.5 R14):
16x scan, ICE and HDR (48 bit).
However, I never found any hint in the manual that this would not work. In fact, it works fine for 95% of the scans. The ones that make trouble are the ones the LS-4000 prescans (exposure measurings) 2 or 3 times. Apparently it does this in case of images with little contrast sometimes. I assume that silverfast runs into a timeout problem waiting for a response from the scanner which takes longer than usual to focus and to measure the exposure. BTW, the whole works fine in Nikon Scan 3.1 (not under XP, though....
Lasersoft Support told me that the problem would be fixed with release 5.5 18, however with this release the problem in fact really started. Now the settings really do not work anymore. Under 5.5 18 it's really almost impossible to scan a slide using 16x scan, ICE and HDR (48 bit).

I'd be grateful for a comment and help from Lasersoft.


Bye

Andreas

Try switching Off the Nikon AutoExposure feature . It's a button on the left edge of the preview window and looks light a a light bulb or sun (can't recall which)

Ian

[Guest]

Dear Jan,

thank you very much for your reply.

I'll try this tonight and will let you know about the outcome. However, I'm wondering why the autoexposure is not essential.
Shutting off the auto-exposure just to make the prescan fast enough cannot be a solution.

Anyway, I'd appreciate a bugfix or comment from Lasersoft on this as I think, this is quite a nasty bug (especially in 5.5 R18) and for my understanding and (if it really is a timeout problem) should be correctable (?).

Bye

Andreas

[ilyons]

On 2001-12-06 17:37, Anonymous wrote:
Dear Jan,

thank you very much for your reply.

I'll try this tonight and will let you know about the outcome. However, I'm wondering why the autoexposure is not essential.
Shutting off the auto-exposure just to make the prescan fast enough cannot be a solution.

Anyway, I'd appreciate a bugfix or comment from Lasersoft on this as I think, this is quite a nasty bug (especially in 5.5 R18) and for my understanding and (if it really is a timeout problem) should be correctable (?).

Bye

Andreas

Andreas,

the autoexposure susytem is trying to optimise the dynamic range of the scan i.e maximise the useful data from pure white to absolute black. If the image has low contrast it may well be that you have very poor blacks and no whites (or vice vera) so Silverfast cannot obtain the optimum value. If the problem only occurs on a small number of images it is reasonable to assume that something about those images is not to the liking of this feature. I don't know if there is a bug or not but switching off the feature would give some indication of the likely cause.

[Andreas]

Dear Jan,

thank you very much for the ideas. Unfortunately this did not fix the problem. While indeed Silverfast does not do prescans when the "Autoexposure" is turned off for most slides, the problem slides are exactly the ones it still does a kind of prescan (one that sounds quite strange, BTW) with.
For me, this it's an absolute mystery and, if Lasersoft is interested in finding out what's wrong here, I'd be glad to send a slide.

Apparently, however, Lasersoft is not interested:-(

Anyway, thank you very much for your help, Jan. It was very much appreciated.

Bye

Andreas



<font size=-1>[ This Message was edited by: Andreas on 2001-12-06 21:02 ]</font>

[ilyons]

refused to scan a very small part of the slides

what does this mean? How small is small? Are you setting up a second frame within the master frame to scan the small area or jut reducing the size of the master frame to crop it.

The error message in German - I don't understand it - what does it translate to?

Apparently, however, Lasersoft is not interested:-(

I think you will find that they are interested, but just at present ALL hands are on deck and preparing for ..............


Please explain (again) EXACTLTY what these image comprise (content ), the type of mount (glass/glassless/metal insert) and format. Also can (you may have said already) they be scanned without ICE and Multiscan, etc. Does switching the autofocus OFF help.


Ian

<font size=-1>[ This Message was edited by: ilyons on 2001-12-06 21:21 ]</font>

<font size=-1>[ This Message was edited by: ilyons on 2001-12-06 21:24 ]</font>

[Andreas]

Dear Jan,

thank you very much for your quick reply.
I'm sorry for my explaination.

With a small part, I mean a small fraction. I'd say between 1% and 5% of my slides cause problems. I can scan them when reducing the multisampling to 2x or 4x. However, this not really is an option. Even more, Nikon Scan does these slides without any problems. According to support the error should be fixed with 5.5 r14. In my case it appeared with 5.5 r14... and got dramatic with r18 which I uninstalled because the error was getting non-acceptable.

The content of the slides are mainly photographed herbarium sheets (for my scientific work), just dried plants mounted on paper and one of my favourite landscape photographs (which I like a lot, though it is quite grainy as it was shot more than 10 years ago on older film) shot in iceland. They all have little contrast but are correctly exposed. I tried to reframe them without success. All are framed glassless (as all of my slides). Scanning with manual focus does not help.

BTW, the error would ranslate "Scannererror 0".

Again, I'm really grateful for your suggestions. Thank you very much.

Bye

Andreas

[ilyons]

On 2001-12-06 22:33, Andreas wrote:
Dear Jan,


With a small part, I mean a small fraction. I'd say between 1% and 5% of my slides cause problems. I can scan them when reducing the multisampling to 2x or 4x.

Now I understand it's not the area of a slide but the quantity.

The multisample thing is a known bug. Yes it was supposed to be fixed in r018 but obviously hasn't been in your case.

The content of the slides are mainly photographed herbarium sheets (for my scientific work), just dried plants mounted on paper and one of my favourite landscape photographs (which I like a lot, though it is quite grainy as it was shot more than 10 years ago on older film) shot in iceland. They all have little contrast but are correctly exposed. I tried to reframe them without success. All are framed glassless (as all of my slides). Scanning with manual focus does not help.

BTW, the error would ranslate "Scannererror 0".

Again, I'm really grateful for your suggestions. Thank you very much.

Bye

Andreas

I think it is the multisample bug that has caught you. However, my understanding is that it was combination of 4X and higher along with dust/scratch filter that triggered the problem.

Things "might" improve for youin a few says - I can't say any more!


Ian

[Andreas]

Dear Ian,

I tried around a little more and indeed (like you suggested) the error can be avoided by turning ICE off and keeping 16x multisampling. While I still think that this is not a real option, it is at least a better solution than turning the multisampling off.

Let's hope this issue will be fixed soon.

Thank you very much

Andreas

[LSI_Support]

Dear Andreas,
try SF 5.5.2, which is available now.