The colors on some scans are waaay off. Very strong blues and orange. Hard contrast and very saturated overall. This is problem only on some images on the same roll as the good scans. It’s kind of hard to edit and doesn’t always turn out ok.
Have tried different white balance tips found in the forum. Also to include some black from the borders. Nothing helps or make a difference. Someone on Youtube suggested a heavy crop as a fix but it didn’t work on my images.
I shoot slightly overexposed to get the look I want. This example is with Kodak Gold 200 120 but happens also with 35mm, Gold/Porta 400.
I hope to get some fix to this. I have asked both photographers and labs about this problem and no one can help me. Not sure I want to pay for V3 without knowing it would solve this problem.
Version 2.3, Nikon D800, Tamron 90mm 2.8 marco, Valio complete scanning kit. I have e mailed Nat one of the images. Thanks!
Are you looking for lower contrast and saturation? More like a film/cinematic look?
I usually overexposed negative film for higher saturation…and this seems to work in your case too. That different captures on your film produce different results is understandable, unless all the photos were taken of the same location, time and settings. I see no issue in your camera scanning technique, therefore I conclude that the conversion boosts contrast and saturation to beyond your intentions. It would be helpful to know how you set NLP (first and second tab settings) and to see how the original camera scan looks. But: I also got a few films that convert to high contrast and saturation and no matter how I set NLP, results are similar, if not identical.
Here’s an example (xmas on a beach in Australia, photo taken around noon)
From left: Camera scans with increasing exposure
From top: unedited raw, conversion thereof, white balanced raw, conversion thereof
In general, I see that thin negatives show stronger reactions to NLP’s settings than dense negatives. I could put a lot of trial&error for less contrast and saturation with NLP, but it’s much easier to edit a positive copy.
In my conversion tests, I see negatives that, after white balancing, look yellowish (like the ones above), reddish or violet, and all of them with Kodak film I bought between 30 and 50 years ago. All of them convert, and all of them have their own “character”.
Thank you. Well basically I’m looking for the look that these film types are known for (not expired film). In this photo there is nothing like what I have been working on for the last ten years of film photography. I expose for shadows/overexpose both for overall brightness and colors. The problem now is that one or two frames from the same roll, same time and conditions, get this heavy baked look.
Also that my Instagram followers/film shooters have not seen this crazy look before, nor know how to deal with it, not knowing where the problem lies. Is it my metering, my camera, bad film, development, scanning process, soft ware etc? It almost makes me stop scanning at home/using NLP. I will try and scan the same negatives an a flat bed scanner with Silverfast software. Also let my lab scan them to see if there are differences from the NLP convertions.
I try to save these images by doing the following editing (see image). And the continue in LR. It sort of saves it but I still would like to know why it gets so extreme with NLP on certain images.
I compare mostly to my other NLP results. Also the strange fact that with NLP I get these results sometimes. Results that no film photographer that I know, nor lab recognize. I guess my main question is if this can be adressed or if I should switch to another software. I just need to figure out if in fact the software is the problem.
NLP works as designed, which means that it converts adaptively.
Conversions can be used…
as is - or
as a starting point for further editing to one’s taste.
Now, the adaptive process can react strongly to minute changes in negatives depending on whether the negative is dense, thin, has a narrow or wide range of densities and more.
Also NLP is best when all colours are present, which is the case here.
Again, can you post an ooc camera scan? You might need something loke wetransfer.com.
Here are a few tips for dealing with this type of situation:
First, look at the color balance, and make adjustments. In this case, the auto-neutral algorithm has added too much magenta. The easy solution is to simply take the “tint” slider in NLP and move it towards green to get a more neutral balance. The color balance is just a starting point and easy to adjust. With negative film, sometimes the auto color balances will nail it the first time, and other times, it will be off… this is true even when working in a lab. The good news it that this is a simple quick adjustment!
Second, look at the tonality and make adjustments. Here, the result has too much contrast in it… in which case I’d recommend using the “Linear” tone profile (or, in NLP v3, change the preset to “NLP - Neutral”
So here, with this image, I’ve just changed the color balance, set the tone profile to “linear”, and the HSL/LUT to “Natural”.
This to me looks to me to be a very neutral interpretation of Kodak Gold 200 and has very little adjusted on it.
If you still want something “softer” there are a lot of ways to do that… for instance, here I’ve just changed the blackClip to -11. This is a little softer than I’d expect Kodak 200, but there is a TON of editing latitude in film (and in Negative Lab Pro).
If there is anything else you are trying to do specifically with this image, let me know and I’ll point you in the right direction!
In terms of comparison with other software, the default settings in NLP are a bit more on the “high contrast / high saturation” side, and while this works great in 80% of cases, it will make magnify any color or tone issues with the conversion. In v3, you can set a different default if you’d like (like NLP Neutral), but it is often easier to spot color issues while the contrast is high, and then adjust the tone profile after you see the correct color balance.
Other tips for improving your results:
Use Roll Analysis in v3 to improve the white balance and dynamic range from the conversion
If you do find that one image in roll converts well, use the “SyncScene” feature in v2 (now called “match” in v3.1, currently in beta).
Thanks for the tips. These are all steps I usually do so that’s fine. The mystery to me was why just an image here and there gets this wacky look. I guess I have to live with this happening. I will try V3 and see if that will help me some. Cased closed
The NLP plugin you use is version 3.0.2. You can see this in the NLP panel (where you set colour model, pre-saturation and start the conversion) at the top end of the panel.
NLP camera profiles come in different generations. Each generation was introduced with a NLP release that had its engine changed in a way that made new profiles necessary. Old cameras can have several profiles, new cameras mostly have just one profile.
In short: The version numbers of the NLP plugin and the NLP camera profiles don’t necessarily match - and that is completely okay.
Again, you use the currently official release/version 3.0.2 of NLP.