One thing I am still struggling with a bit is whether I should be trying to get the image to look close to my desired end state in NLP itself, or if that’s unrealistic and I should expect to always do some final editing outside of NLP. My experience thus far is that it’s pretty hard to get what I’d consider a “final” result with NLP only, and I also find the controls somewhat fiddly. And I can almost always get things looking the way I like quickly in Photoshop and/or Lightroom. Then again, I’ve been using Photoshop since 1993 and Lightroom since it was introduced, so I’ve had a lot more time to get used to those controls. So I don’t know if I should continue to put a lot of effort into refining my NLP skills so I can get closer and closer to the desired final result, or if I’m better off just getting it “close enough” and doing the rest of the edits elsewhere.
Curious if there’s an agreed-upon best practice, and what others do.
Fair question. My definition of “perfect” for this discussion is “exactly the way I want it to look.” Which is of course completely subjective, but I’m just speaking about whether I can get things looking precisely they way I wish them to look.
If it makes it easier, we can strike the word “perfect” and re-phrase the question as “do you aim to do all of your edits in NLP or do you expect to do more tweaking outside of NLP before you consider yourself ‘done’?”
And yes, I always export 16 bit tiffs for any further editing. I don’t even use Lightroom Classic for anything except NLP, so it’s not even a positive copy in Lightroom. I literally export a file and import it into either Photoshop or Lightroom (non-Classic) to finish the edit.
If we understand Photography as a creative art, subjective is all it is.
NLP can do a lot of things with its selections and sliders. Sometimes, the second tab tools will do, sometimes, re-converting with a different colour model, pre-saturation, border buffer and white balance setting gets closer to target. Going that route means a lot of iterative testing which can help to understand NLP’s ways and limits, but then again, the source image plays an important role too.
The lot of options in combination with and dependence on the images leads into a maze that changes like Hogwart’s staircases. And although easy to use, NLP is hard to master imo. Therefore, I tend to do an(one) initial batch conversion and continue only with the image(s) I really want to show, continue with either the options of NLP or the tools applicable to the TIFFs.
This is my approach for dealing with negatives taken under a wide variety of lights and subjects, which makes each capture an individual with or without special needs. Hence, NLP sets the stage for easier culling and provides a set of tools for customising, but a set that is just short of doing it all. EEMMV (everyone else’s milage…)
I shoot both digital and film, although I only started with film about a year ago. One of the reasons I enjoy film is exactly this: it embraces imperfection. The little flaws, limitations, and surprises make it feel more real to me.
When I first started using NLP, I spent a lot of time tweaking settings and chasing the “perfect” conversion. Over time, I moved in the opposite direction. Today, almost everything is set to the defaults, and I saved those settings as my standard workflow.
My goal is simply to get what the film itself gives me. That’s the interpretation I’m interested in. Everything beyond that belongs more to the digital world, where you can keep editing endlessly - all the way to AI-generated imagery if you want. Nothing wrong with that, but it’s also very easy to lose sight of the point where photography ends and graphic design begins.
So for me, NLP is mostly about translating the negative, not perfecting the image.
Without getting too philosophical about “is there even one true interpretation of a negative?”, my view and experience is that NLP is already doing a ton of interpretation (even setting aside any variables introduced by the light source and camera or scanner), and it doesn’t always seem completely deterministic on a given frame.
I’ve been shooting film for over 30 years, so I’ve done everything from darkroom printing to operating a minilab (pure analog) to working with just about every type of scanner that has been produced (the first I used was a Leafscan). I have seen a broad spectrum of how images can be interpreted. And however I would internally think of a “straight” representation of a color negative, the default output from NLP rarely comes close to it. That doesn’t mean the default looks bad (though sometimes it does…), but it definitely is very much an interpretation, and not one that tends to be neutral or similar to other interpretations in my experience.
All that said, it’s definitely helpful and interesting for me to hear that some folks find the default output to be pleasing without any editing at all! I don’t think I can personally get there, so for me it’s more a question of whether I do the adjustments in NLP or in Photoshop.
I observe that converted images can look different depending on WB and crop settings. I found that sometimes, it’s beneficial to NOT white balance a negative or to crop things off that contain exceeding amounts of dominant colours or tones.
For ease of use, NLP can be used like a contact sheet: It reveals what’s in an image, but it’s not necessarily the targeted rendering. And while many things can be corrected with NLP’s tools, other tools exist that can be needed, so why bother correcting with NLP - if a positive copy has to be created sooner or later anyways?
Imo and (all) things considered, what we do and how we do it is more of a personal taste/choice- than a “right thing to do” decision.
I aim to get a 16 bpc tiff with all the data & reasonably good colour quickly & then do whatever work is needed in Pshop. I look on the NLP file as Ansel Adams did a negative & the final P’shop file as a print. The negative is the score. The print is the performance.
@graveladvisory Given that you are not a novice in field of photography and have lot of real experience with film, I suspect that something is off with your setup - either light source or exposure during the capture or something else. For originals with common, everyday content NLP provides very decent results out of the bat. (Yes, the shots may have blue skies which are often characteristically NLP-style, but that’s minor.) At which point i save file as 16 bit tiff and continue working on it in LR. This is mostly because NLP reverses tone curve slope and the traditional behavior of LR sliders becomes counterintuitive.
As far as i understand NLP uses some sort of heuristics to invert colors with idea that RGB channels get balanced properly only if wide gamut of colors is represented in the picture. So the bottom line is to find and preserve presets which typically produce good results for you and apply those presets consistently and resort to extra editing when picture deserves it.
In the ‘Suggested backlight sources’ thread @graveladvisory posted that simply inverting the file gave better results than NLP. I wonder if NLP isn’t optimised for RGB light sources and so is thrown out of kilter in some way. In another forum there is an alternative RGB light source that also combines white light, all adjustable. He uses NLP, always has and gets excellent results, ,but has found that the white light element is very beneficial.
The amount of editing control one has in LR on an NLP converted raw file is actually quite limited relative to what LR can do on original digital files made from a digital camera without NLP. If you wish to make edits well beyond what NLP can provide, then saving the file as a TIFF and working it in either LR or PS is the way to go.
For me, it is much more efficient and effective to edit in Lightroom as much as possible. I spend just enough time in NLP to get a rough approximation of where I want to go, then I export a Positive and switch to Lightroom for my final editing.
Significantly, I don’t believe it is possible to make local adjustments in NLP. Almost every photo can benefit from local adjustments, and that requires some sort of a tool that can select part of the image; as far as I know, no such tool is available in NLP. This limitation means NLP is not capable of making complete and final adjustments to my negatives.