Strange Nikon Coolscan LS-8000 color noise

[tony25]

I just picked up LS-8000 scanner. I scanned slide film and I noticed strange color noise. I tried with my other Minolta 5400 II and it didn't have the issue. I am using VueScan and checked Fine mode. It seems to be noticeable when zoom in. Anyone experience the same problem? I would like to make sure that this isn't user error before I am returning the scanner to the seller. Thanks.

 

[Tomro]

I got the exact same problem. Don’t know what it is though or how to fix it

 

[tony25]

I did couple more testing. The above image is Slide. I tested with negative this time.

This is with Infrared off

and this is with infrared on:

Turning on and off on slide (provia above) doesn't help much.

 

[brbo]

Have you tried it with Silverfast (I believe they have a demo version) or Nikonscan?

 

[Tomro]

to me it happens with either SF or NS

 

[brbo]

Do you get the same pattern regardless of the scanning resolution? Does it stay the same if you enable multi sampling (like 2x, 4x..)?

 

[Tomro]

Do you get the same pattern regardless of the scanning resolution? Does it stay the same if you enable multi sampling (like 2x, 4x..)?

as far as I remember it is independent of scanning resolution: I normally scan at 4000dpi, but I remember seeing it also at 2000. Never tried multi sampling

 

[brbo]

Thing is that this looks like some sort of electrical interference (time-dependant based on the pattern) and I'd guess that changing resolution or multisampling would also change the pattern...

Have you tried using the scanner in a different location, with different FW cable...? But then again, the pattern isn't visible on negatives.

 

[Tomro]

while I'm not the OP, I did test the scanner in 3 different locations, one even 300km away from the other. 2 different FW cables, on a Mac mini, a Windows 10 PC.
Pattern has nothing to do with the negatives, unfortunately

 

[brbo]

I meant that the OP shows scans of slide film with visible pattern and a scan of negative film that doesn't have the pattern.

I'm clearly out of ideas, but curious if your scanner was always like that or did it develop the problem with time?

 

[tony25]

Sorry, for the confusion. The pattern is still there on negative but it's less obvious. If you look at the second negative image where I have ICE (infrared setting on Vue) on, the pattern is obvious in the film edge (white area). When I turned off, it looks much better , but if I added sharpen, I start to see it again.

 

[gorbas]

Try multiple scanning pases?

 

[tony25]

um.. I didn't try multiple scanning passes. I did try multi-exposure only to slightly increase the issue. I can try multiple sampling tonight. Just curious, what can be the issue that multiple scanning passes help?

 

[tony25]

I did 8x sampling and it seems to be more obvious:

 

[runswithsizzers]

RE: "I did 8x sampling..."

I have found multiple sampling to be somewhat effecive in reducing noise on underexposed areas of film. The improvement at 2x was noticeable. At 3x there was some additional improvement that was discernable; I did not see anything at 4x to justify the extra time.

However, what you are showing here looks a lot different from what I call noise. When I see noise on underexposed film, the pattern is a random distribution of faintly colored dots. I agree with others who suggest your regular, moire-like pattern seen on the slides looks like it might be some kind of electrical interference.

While it is probably not the cause of your problem, I have noticed different sharpening routines can magnify some undesirable scanning artifacts. If you have good post processing software, my advice would be to not let your scanning software do any sharpening. Chances are, you will have more control and get better results when sharpening with something like Photoshop.

 

[brbo]

If you make multiple scans of the same negative with the same settings do all have exactly the same patter or does it shift left/right/up/down? Then changing multisampling settings (and again doing multiple subsequent scans), do you see exactly the same pattern in the same place. You obviously don't have to scan entire frame, just a small part where the problem is most prominent.

Probably won't help, but if you are desperate enough, clean the scanner inside, unplug and plug again every connector... CS-8000 is easy to get inside. If you have Facebook show your scans in "Orphaned Scanners" group. Groups.io group for coolscan scanners if not very frequented anymore.

 

[cramej]

For those having difficulty seeing the pattern:

 

[tony25]

Based on what I understood about electrical noise, it should be more gaussian random. I can see pattern for each motor stepped here (so, I incline to think that this is not really noise). I joined Nikon Coolscan facebook group last night (they actually have very good info around these scanners). One of the post stated that this may be due to CCD wearing out, though, the pattern seems to not exact what I and cramaj have.

Updated: Look like the group has some fix by replacing A/D capacitors. In case, you have problem and handy with electronics.

 

[Tomro]

I t

Based on what I understood about electrical noise, it should be more gaussian random. I can see pattern for each motor stepped here (so, I incline to think that this is not really noise). I joined Nikon Coolscan facebook group last night (they actually have very good info around these scanners). One of the post stated that this may be due to CCD wearing out, though, the pattern seems to not exact what I and cramaj have.

Updated: Look like the group has some fix by replacing A/D capacitors. In case, you have problem and handy with electronics.

I too am member of that group but didn't see the post. Could you elaborate what should or could be replaced?

 

[tony25]

@Tomro There is PDF (LS-8000 Zig-zag banding - cause & cure). This is the link to the thread.

 

[Tomro]

@tony25 thanks so much! While I can't access the linked file in your post I was able to find it under the name on Facebook.
Seems straight forward, just need to find the capacitors and learn how to solder

 

[BMbikerider]

Rather than show us a hugely enlarged sample can we see if we can detect anything at normal image size say no larger than A3.

My 35mm Nikon scanner will give sharper results at up to A3 than those you have shown and any colour shift is down to operator error. We really need to se the whole scan rather than a tiny portion of it. What is shown here is like posting an image of a underexposed area where the remainder of the image is perfectly OK

Is the perceived problem something more basic like the medium you are scanning was originally taken in a light which was not a correct colour balance. Was it slide film or colour negative film, that may be the problem not the scanner. Was there a mix of daylight and artificial light from one or more sources?

 

[koraks]

What is shown here is like posting an image of a underexposed area where the remainder of the image is perfectly OK

The kind of noise shown here is problematic by all means. It's not supposed to be there, and it will show up on larger prints or digital enlargements.

Is the perceived problem something more basic like the medium you are scanning was originally taken in a light which was not a correct colour balance.

Have a look at the pattern; the conclusion that it was electrical noise likely from the stepper motors interfering with the sensitive A/D conversion circuitry sounds really plausible to me. I'd replace the power supply decoupling caps at the A/D chip(s) and check all electrolytics as well.

 

[Tomro]

Are there any other online stores in Europe other than mouser that offer the parts needed to be replaced?

 

[koraks]

I mostly use:
Conrad
TME.eu
AliExpress
And sometimes local suppliers but those generally don't ship outside of my country.