CCD scanners and shadow noise, grain anti-aliasing etc

[philipus]

Are all CCD scanners the same in this respect?

The Coolscans have, I think, a problem with noise in shadow areas and scans require quite a lot of post-processing to address this. I have a V ED and a 9000 and the 9000 is better in this respect due to the light source but it is by no means good.

I've had my eyes on iQsmarts but they don't appear often. They're also CCD-based so I am wondering if they would give similar (-ly poor) results.

And what about other scanners, used or new?

I would love to shoot more C41 due to the latitude but I hesitate because of the difficulty scanning and post-processing the negs.

I guess a related question is post-processing tips (and even scan software settings) to reduce the problem.

Thanks in advance for your thoughts/experiences,
Philip

 

[mfogiel]

The most important tip is to overexpose your C41 film, and voila' you will no longer have noise in the shadows. Then you can apply multiple pass scanning, but if you go for this I warn you it can take some time, so you better use the glass carrier irrespective of size, because you risk the heat making the film bulge during the scanning process. Multiple pass scanning averages out readings, so noise becomes greatly reduced.
Third thing you can do, is to get one of these plugins for PS with grain reduction software, and make your own tailor made filter by shooting a target and developing as usual, and subsequently calibrating it with the software, this way you will optimise grain and colour cast.
But the most effective strategy is to overexpose - if it is a ISO 400 film, try to overexpose even up to 3 stops.

 

[philipus]

Thanks Marek for your helpful observations. Three stops (and even two stops) overexposure generally for ISO 400 sounds rather extreme - are you sure?

Would you apply the same to ISO 100 and 200 C41? EDIT: And would you develop at box speed or? Sorry for the stupid questions.

But, fundamentally, I thought noise in the shadows was (at least to a large extent) due to how the Coolscans work. They try to pull every last bit from the shadows. So if this is true overexposing would not be a silver bullet.

I have tried multi-pass scanning before but saw virtually no (as in none) difference. My negs didn't bulge, btw, even without glass.

The idea of a tailor-made grain filter is very interesting. But would it really help? It seems to me that the noise is completely random. Perhaps I misunderstand how such a filter works. Could you recommend a filter?



best and thanks again
Philip

The most important tip is to overexpose your C41 film, and voila' you will no longer have noise in the shadows. Then you can apply multiple pass scanning, but if you go for this I warn you it can take some time, so you better use the glass carrier irrespective of size, because you risk the heat making the film bulge during the scanning process. Multiple pass scanning averages out readings, so noise becomes greatly reduced.
Third thing you can do, is to get one of these plugins for PS with grain reduction software, and make your own tailor made filter by shooting a target and developing as usual, and subsequently calibrating it with the software, this way you will optimise grain and colour cast.
But the most effective strategy is to overexpose - if it is a ISO 400 film, try to overexpose even up to 3 stops.

 

[willie_901]

It can be confusing to think about noise levels without thinking about the signal levels.

Visible noise is a symptom of low signal levels. The same amount of noise is present in highlight regions. Unlike the shadow regions, the highlight regions' noise level is low compared to the signal level.

There is no way to "pull every last bit from the shadows" when the scan is underway. There is nothing to "pull" as the analog signal-to-noise ratio is an inherent characteristic of the CCD data stream. However post-scan, in-scanner software can produce an image with non-linear rendering which selectively pulls shadow regions.

Multi-pass scanning lowers some noise levels by the square root of the number of passes. This means two passes should reduce pseudo-random noise levels by ~ 40%. If this is not the case, then what your attribute to noise must be in the negative or some other type of non-random noise.

Aliasing is inherent to all analog-to-digtal conversions.

mfogiel's accurate and complete advice addreeses both issues.

Perhaps a scanner with newer technology would have a better signal-to-noise ratio?

 

[Dwig]

... The Coolscans have, I think, a problem with noise in shadow areas ...

When scanning postives, yes noise in the "shadows" is common as the signal is the weakest and the background noise in the system is a higher percentage of the total. You should find the opposite when scanning negatives as it is the highlights that are the dense low signal areas.

When discussing either shadows or highlights in reference to scanners or copying with a digital camera you need to first specify whether the original is a positive or a negative.

 

[Pete B]

Three stops (and even two stops) overexposure generally for ISO 400 sounds rather extreme - are you sure?
Philip

I typically do one or two.
http://ukfilmlab.com/2014/04/24/film-stock-and-exposure-comparisons-kodak-portra-and-fuji/
Pete

 

[Dwig]

The most important tip is to overexpose your C41 film, and voila' you will no longer have noise in the shadows. ...

With negatives films, like all C-41 films, the shadows are the thin nearly clear portions of the film and true noise in the digital signal when scanning has little if any impact. It is the highlights, the dense portions of the negatives, where digital noise has its largest impact.

Increasing exposure on C-41 films reduces the size of the grain, or at least the dye clouds where the true grain was. If you carefully examine a large wet print from a C-41 negative you can see the decrease in grain as you look from the shadows towards the highlights. This is easiest to see in B&W prints but is present in color prints as well. Intentionally overexposing C-41 films effectively makes them, on average, finer grain.

 

[philipus]

Thanks very much for the further replies, and Pete also for linking to UK Film Lab which is very helpful.

I'll try overexposing 1-2 stops. But, hmm, for indoors shooting it will become a bit difficult.

 

[Sparrow]

... f2/as-slow-as-you-go works fine if one's resigned to block the shadows, and the fastest film one has

I agree but note he lives in Monaco; I've gone to 4 with ilford's xp 400 even further south than he