I made two different scans of the same image that should be at least similar. (The film here is grainy and dirty, but that’s beside the point.) The bottom one was from a DSLR scan that included several strips of film. In that case, I set the white point as advised and the cropped way down to the individual frame before converting. This was a success in my view - a (grainy low res) image of a beach scene with what looked like typical/good color. In the other one, I made a single, high res scan and used the same flow after - set the white point, cropped a little and converted. The color is really odd. Dark in a way that can’t be corrected, orange sand, etc. I think I had something odd happen with a beach scene some time back. I double checked my work.
Welcome to the forum. You say that bottom one was a DSLR scan “that included several strips of the film”, so presumably a kind of contact print? The top one is “a single high res scan” so presumably you got your camera closer somehow so that the negative filled the frame. Are you able to give any details about your setup and the exposure for each. If you look at them together is the exposure for this particular negative more or less the same, not darker/more dense or lighter/thinner?
Thanks for the help. The bottom one was what you might call a contact print. I didn’t use any special process. For the conversions, I zoomed in and the results looked good to me. For the full frame I used a canon 100mm 2.8 macro, for the “contact print” I left the camera in the same position and switched in a canon 35mm lens. For both I masked off bright areas around the image with black cardboard. I set each lens to F 11 and used the histogram, setting the exposure time to where the clear areas of film - “highlights” from the perspective of the dslr - were a safe margin from clipping. I’m attaching the two unconverted images. The values in the two images aren’t that different: clear film is RGB 92, 89, 75 in the “contact” sheet image that converted well, and 89, 83, 64 (darker) in the high res image that didn’t. There is one difference that might be significant though. The exposure time for the high res images was 3.2 seconds, whereas for the contact sheet it was 1.0 seconds. I don’t know much about reciprocity with DSLRs, but there may be an issue there.
Thanks, not so much reciprocity as light lost at the higher magnification with your 100mm Macro. The fact that you are using a different lens and a very different magnification might limit the usefulness of Roll Analysis with these two but it would help for other images from the same roll taken with your Canon 100mm Macro. To me the left hand image looks under-exposed but NLP is very forgiving when it comes to exposure.
Reviewing the initial images, I found that the upper had some diamond shaped pattern, visible in the darker “sand” portion of the image. The two scans show fairly low contrast und different overall brightness as noted by @Harry.
I tested low contrast/low density/narrow histogram negatives with different lighting (Enlarger, Kaiser Plano, iPad…) and found that these react very strongly when converted with NLP’s adaptive algorithm. During the conversion, the width of the histogram is stretched (depending on darkest and brightest parts of the image) and this “amplification” picks up minute differences and makes them easily visible, even though NLP is tolerant to variations in exposure of the scans.
I propose/advise you take a series of captures, taken at the same aperture and iso settings and different times to produce scans of e.g. -1, 0, +1, +2 and convert them like shown here.
Also, it can help to deviate from standard procedures occasionally. Some of my negatives get me a better starting point when I don’t white balance them. NLP has soo many ways to tweak the conversion and converted image that it makes sense to (systematically) test different settings. This does not mean that you’ll find the silver bullet for all of your images though.
Hmm. Am I getting things backwards? The lighter image at left shows the highlights (clear film plus base tint) as brighter → more exposure. The low contrast is concerning. There may be stray light getting in, so I should probably try to eliminate that.
What I’m saying is, that, starting from the negative, the gear has some influence on how conversions will look, specially with scans that have narrow histograms (low contrast).
Stray light makes a difference too, but I found that some stray light is tolerable, even if best practices are against it.
The point is to try conversions with different colour models, pre-saturation and border settings etc. Often enough, I find this to improve conversions, even with images that have histograms like this:
So, no matter your scans, NLP can pull images out of them, but not necessarily in an easy and repeatable (in the sense of images with similar histograms) way.
If you like, you can share the two scans (the original RAW files) so that we can try to “make them work”. As a new member of the forum, you best share the files with something like wetransfer.com or a share (Google drive, etc.)
I explored a little more. Changing exposure in the digital capture didn’t seem to make much difference. The two images above were from the same capture. The difference is that in one case, I cropped down to a small area around the figures (shown above) before doing the inversion. This gave results closer than to what I’d expect. The dark sand at the bottom seemed to mess up the algorithm for the conversion in the top one. There are a lot of other issues to look in to here - the crosshatching, which may be from my Kaiser light tablet. There’s also the vignetting. I don’t think this camera (Hexar AF) has major issues here, so it’s something else to thing about.
Thanks all for the help with this. DSLR scanning has had a really steep learning curve so far. There’s a lot to be said for an Epson flatbed. It has less resolution with 35mm film, but a lot more with large format.
More data: photographed my Kaiser slimlite at using the same height and settings as I’ve used for my scans, then radically increased the contrast in photoshop. That explains the cross hatch. I think that the vignette, if it has any effect, would make the corners brighter when the results are inverted. More experiments to come.
Kaiser slimlite plano tablets have this pattern, I got it with a similar experiment too. The pattern does notmatter though in most cases, but it can show as we’ve seen in the original post.
It helps to move the negative away from the backlight, specially with thin, low contrast images.
The falloff is most probably caused by the lens and can be, if unwanted, compensated with (a) circular mask(s).
