Opticfilm 120 Banding Issues

[eurekaiv]

Anyone know what the deal is with banding in Opticfilm 120 scans? I have an older non pro model, picked it up used for $600 so it has served me well but I recently started getting banding in my b&w scans. I've probably scanned 100 rolls in this thing over the last couple years and this is the first I've seen of it outside of some very mild banding in some color shots. My b&w scans were always quite good. Something seems to have changed however.

 

[fs999]

Never had banding with my OpticFilm 120. You could send it in for a repair. In Germany they have a good service, I don't know how it is in the US.

 

[eurekaiv]

There's a long (and old) thread on flickr about this... no solutions given, just some hunches that it's a hardware problem.

Plustek OpticFilm 120

All about Plustek OpticFilm 120

www.flickr.com

 

[George Collier]

My understanding of banding is that it occurs in areas of very gradual (slow) gradation (or color) change, in very smooth areas, like skies.
The gradation change is so slow, or gradual, that you see an edge form when crossing over from one pixel value to the next, like a set of steps vs a ramp.
I'm not sure it is the scanner at fault. Film grain or any other detail in the original can obscure or interfere with the visibility of the bands, so you don't often see them in 35mm film scanned to a high scale ratio, etc. I have seen them in 4x5 film scans only, since the grain is so fine.
My solution is to scan everything at 16 bit (all black and white negs for me), It produces many more potential values for a pixel to have. And I think it has to be done at the point of scanning from analog to digital (at the scan) rather than just taking an 8 bit depth file and converting it to 16 bit in the Pshop Image menu. Even at 16 bit, with a 4x5 image, with a clean winter sky, I can get them if I start to do too much graded curve adjustment, which is applied after scanning.
I'm not sure my explanation is physically accurate, but this seems to be how it works in my experience.

 

[eurekaiv]

My understanding of banding is that it occurs in areas of very gradual (slow) gradation (or color) change, in very smooth areas, like skies.
The gradation change is so slow, or gradual, that you see an edge form when crossing over from one pixel value to the next, like a set of steps vs a ramp.
I'm not sure it is the scanner at fault. Film grain or any other detail in the original can obscure or interfere with the visibility of the bands, so you don't often see them in 35mm film scanned to a high scale ratio, etc. I have seen them in 4x5 film scans only, since the grain is so fine.
My solution is to scan everything at 16 bit (all black and white negs for me), It produces many more potential values for a pixel to have. And I think it has to be done at the point of scanning from analog to digital (at the scan) rather than just taking an 8 bit depth file and converting it to 16 bit in the Pshop Image menu. Even at 16 bit, with a 4x5 image, with a clean winter sky, I can get them if I start to do too much graded curve adjustment, which is applied after scanning.
I'm not sure my explanation is physically accurate, but this seems to be how it works in my experience.

Thanks George for the thoughtful reply.
My workflow really hasn't changed (I scan B&W with Vuescan as a 16bit greyscale and 5300dpi which I then scale down for sharing) and the artifacts showed up in my images only on this last roll of film. Maybe the roll had a problem but it looks just like examples I saw on the thread complaining of the issue. FWIW, this was a roll of Tri-X vs the Delta 400 I more commonly shoot but my negatives looked perfect on my LED lightbox. I'm a bit hesitant to post examples of anything on photrio anymore so I'll continue to describe this issue as best I can and if I can make it to the darkroom before xmas (this is my ultra busy time of year with work related stuff) I'll try printing one to see if I get anything noticeable on paper, just to rule out the actual negative.

 

[George Collier]

"I'll try printing one to see if I get anything noticeable on paper. . "
Do you mean try enlarging the neg on silver paper, or making a digital print? Shouldn't show up in a wet print unless you can see it in the negative (banding is a digital thing, not analog.)
Also, remember that banding occurs in the output device (display, or paper if digitally printed) if it doesn't support the gradation.
Just in case, has your computer display changed recently?
Trix should be grainy enough (though not like the old days) to create enough noise to mask banding. What was the developer? Is the grain sharp?
One last thing - be sure (in PShop or whatever editing program) that the file is 16 bit, and has been from the scan, not converted later.