Little green dots.....

Hi, I've used a number of prosumer scanners over the last decade. I've used all of them with VueScan. However, amongst all of my color negative scans, there appears to be little green dots in the shadows, and in fact there seems to be a lot of shadow noise in my scans. The only constant across the decade is me and VueScan. Multipass and multisample seem to do nothing, which surprises me, because if it is random noise, you would think it would average it out. As this is negative stock, that means the noise is in the light areas of the negatives.

Any idea what I am doing wrong?

Provide a scan sample for examination.

Have you consistently used the same film type?

Do you need to see into the shadows? A common problem is that people think they should be able to see everything in the shadows. Increase the contrast and let the shadows go black. The noise will diminish. If you want to see into shadows then you need to get the exposure right to expose the film for them. Not saying there might not be an issue with your scanner, but sometimes people are trying to push their source material in ways that just do not work.

So this is an unknown color negative film type taken on an unknown camera (doing family archiving and not my usual real photography). It was scanned on the PrimeFilm XA at 5000SPI, and reduced in size with 2x2 averaging. This is also the product of 4x multiple scans and multi-exposure (which I guess doesn't affect the shadows on negative film). Just about every one of my scans with black in it shows this. The fact that the green noise remains as visible as it does after the 4x sampling and the 2x2 averaging leaves me wondering if it is my workflow, VueScan or the scanner.

When I adjust contrast to move the noise into the black, much detail is lost in the rest of the picture. They are standing on a maroon carpet, which now looks black. They are in front of a dark wood bar, and it is now black.

I suggest you look at the negatives with a good loupe. There may be contaminates embedded in the emulsion from processing or a type of mold has sat in from storage.

James, I would not be so sure that the scanning process would not create the green dots. I would first do some tests to rule the scanning process out before concluding that is not the issue. I would turn off multi exposure and multi-scan for one image and then try multi-exposure and multi-scan separately for two more exposures and the last exposure would be multi-scan and multi-exposure together. If some have the green dots and some do not, that wold narrow down the issue. If they all have the same issue then you at least eliminated the issue being related to the software combining all the separate scans properly together.

 

Thought this would be worth a shot.

The green speckle isn't the only issue. The sample has an overall blue-magenta cast and the tuxedos are brown.

 

I think you need to tweak the colour and tone curve by hand, and not rely on the scanning software to get it right.

 

"When I adjust contrast to move the noise into the black, much detail is lost in the rest of the picture..... They are in front of a dark wood bar, and it is now black."

Using the curves tool is far more refined than a crude contrast slider.

There's also falloff from on-camera flash to take into account, which will render background objects darker than those closer to the camera.

 

I've taken the liberty of adjusting the picture using the simple editor on my smartphone. "Autocolor" got rid of the magenta cast, and a tweak with the curves tool boosted the highlights and suppressed the shadows.

 

A full-blown image editor would have allowed much better correction than the above, but for a few seconds work on my phone I think the result's not too bad.

Too late to edit.

 

The fact that a fairly strong magenta cast went unnoticed leads me to suspect that your monitor might be in need of calibration. A green bias on the monitor would exaggerate any green speckling in the scan.

Hi, thanks to everyone so far. I did notice the magenta cast but didn't correct it or any other adjustments for the picture I posted as I wanted people to see what is coming out of the scanner/vuescan combo with nothing more than scratch and dust removal. I didn't even think that the magenta cast could affect the shadow noise.

So while I try a few of your suggestions, can I get people to agree that (whatever the cause) that kind of noise is not "normal" for a color negative? Basically, I don't want to hear "sorry kid, that's the best you can get from a 35mm negative". I see this kind of noise in my older scans (taken with a variety of scanners over the decades) taken with pro gear and better-than-consumer film. I suspect it is something in my workflow, as the only other constant is Vue Scan, which I know people are getting better scans from.

Once I finish the family happy snaps, I plan to rescan my old amateur landscape and portrait, as well as all of my new medium format film that I started shooting agan a few months back. For that, I had intended to start making linear raw scans to feed into ColorPerfect. My fear is that these medium format scans will be as bad as my 35mm scans.

I'm not seeing scanner noise in the colour corrected tuxedos. All I'm seeing is "normal" film dye clouds, or "grain" if you like, from the combination of fast film and a small format.

 

Scanner noise occurs on the pixel level, and looks very similar to hot pixels or pumped up ISO in a digital image. That's not what I'm seeing in the above sample.

 

Having said that, you haven't let us see the raw scan at 100%. The downsampling will have distorted the appearance of the grain structure or noise - if there is any.

 

Why not scan a negative that you know has been shot on slow (100 ISO) film and been properly exposed and developed? Then scanned straight with no oversampling or manipulation after scanning.

 

That's the only way we'll be able to judge the scan quality.

Joe,

Thanks. I'm not quite sure what box my photos are in. I recently moved and still have many boxes to go through. Having said that, I unpacked a Soviet Leica copy and bought a few rolls of Ektar 100. I'll run some film through it this weekend. Give me a week to get it developed and I'll repost. Thanks again,

James

With scans I think it pays not to do too much pixel peeping as film may produce all kinds of artifacts - Newton's rings, slight uneven density patches, uneven sharpness over the frame, noisy shadows and blown out highlights. I think this is a particular issue with color negative film as there is no reference, so you may try and look for things that you think should be resolvable and aren't within the dynamic range of the scan. With slide film you can see what the scan color should match and what is realistic to find in the highlights and shadows.I already see improvements in your and Rodeo's tweak where the contrast is increased and the color adjusted.