Uncorrectable colour issues with old Fuji negatives

[overnight5603]

Negative Lab Pro 3.1.1

Epson GT-X970 (V750 Pro) / Silverfast SE Plus 9.2.8

Scan settings: 2300ppi 48 → 24 bit TIFF with ME and iSRD on

Adobe Lightroom Classic 15.2 Build [202602111402-ec4112e8] / Camera Raw 18.2

Hi everybody, newbie seeking assistance with an intermittent but persistent colour conversion problem with old Fuji negatives.

I’ve been using NLP for a few weeks and it has been an overwhelmingly positive experience so far.

The thing that originally motivated to buy the NLP licence in the first place was difficulty getting Silverfast’s in-built colour conversion to handle the occasional strip or set of strips.

But this past week I’m running into the same trouble with NLP as I had previously with Silverfast trying to work with some Fujicolor HR 100 negatives from the 1980s. It was Fuji negatives giving me trouble before NLP too.

Basically, with affected negatives, the entire frame is polluted with a sort-of colour cast that I can only describe as dehydrated radioactive urine. It can be managed with great difficulty, but there is no way that I can find to get the image actually looking right.

I converted non-Fuji negatives from the same event as the most recent problematic strips, taken at the same time under the same conditions, and NLP handled them perfectly.

What am I doing wrong and how do I do it right instead?

Or is this a matter of the negatives degrading in a way that’s difficult to address with the kind of uniform whole-of-image editing I’m doing in NLP and Lightroom? I’d like to avoid labour-intensive Photoshop work to correct these if at all possible.

Unconverted TIFF files: Transfer.it

Thanks in advance for any help or advice.

[Digitizer]

Welcome to the forum @overnight5603

In your post, you wrote

@overnight5603

It was Fuji negatives giving me trouble before NLP too.

@overnight5603

the entire frame is polluted with a sort-of colour cast

@overnight5603

I converted non-Fuji negatives from the same event as the most recent problematic strips, taken at the same time under the same conditions, and NLP handled them perfectly.

…and this looks like the problem is in the negatives. The cast seems to have a slight gradient too. Uneven casts, low contrast subjects, mixed lighting and more can throw off NLP. Sometimes it helps to crop the image differently (e.g. to get rid of burnt areas or to balance colours) before converting. Under these conditions, a trial-and error approach can help get better conversions or at least some experience and build one’s lessons learnt and best practices.

NLP converts, based on what the negative offers. Some negatives can take a lot of effort to correct after the conversion, others are close to un-recoverable and I then convert those to B&W.

NLP is a great time-saver and provides easy-to use starting points. Occasionally, it can’t handle the material fed into it - and it doesn’t know what look the user expects.

Your edits produced a fairly usable image imo. Some colour work in Lightroom can work miracles, be with local or global edits. Alas, there is no free lunch in NLP either. But as long as most of your negatives or scans work out in NLP, there’s a discount.

[winslow]

FWIW I have found that Fuji degrade more significantly than Kodak emulsions. Normally I see this with rolls shot after their expiration date but it would make sense that perhaps the dyes also degrade after processing.

[greene881]

As Digitier suggested, as an alternative, you can get some nice B+W images from these negs.

[Mark_Segal]

Actually, I think your title for this post is not quite correct. Having experimented with your down-loaded tiff file of the Fuji negative, I don’t come to quite the same conclusion. Tricky, yes, but “uncorrectable”, not so much.

Firstly, I tend to agree with Digitizer that the fundamental problem may have something to do with the film image itself. Perhaps the dyes are not standing-up very well to whatever storage conditions they endured over time, I don’t know, but the tell-tale sign is that the colour distortions do not appear to be a uniform cast, but rather have different errors depending on the colour, which is kind of what one would expect from dyes that misbehave differently depending on colour and cause.

Secondly, you sent us a TIFF file. Unless the image was photographed or scanned in one of those older devices which produced TIFFs rather than raw files and therefore you don’t have a raw file, I would appreciate if you could send us the raw file because who knows what could have happened in the conversion from raw to tiff.

Thirdly, there is a strategy for approaching problems like this which starts from the general and drills down to the particular. Firstly, for the initial Edit settings in NLP, select the group that provides the most overall neutral mix available from all the options (see my “pre-treatment JPG for this one:

The fact that this starting point is not too bad is revealed by the near-neutrality of the shirt collar worn by the gentleman in the middle. Also the gray hair on both men is not too bad, and the green leaves look quite natural. The stand-out issue is the suits of the two gentlemen on the far left - they are lime, when they should probably be white (I’m guessing, I wasn’t there, but normally one doesn’t wear lime suits to weddings). This can’t be corrected with general edits in LR without messing up the rest of the photo which looks much better. So we go to local editing using masks in Lightroom. Using the brush tool, I created one mask for the jacket tops and another for the pants, and then adjusted the Color sliders within Lightroom’s masking toolset until I brought them as close as I could within a reasonable amount of time to look kinda white. Then using the linear gradient tool I created another mask for the right side targeting the gentleman just into the lady, in order to neutralize what looked like a slightly greenish cast, which could have come from the trees, again using the Color sliders. Where I left it looks like this:

Oh - then back in NLP I added a bit of saturation. Those suits on the left are not totally white - more time spent I could have cleared out that slight magenta cast they have, but you get the point - big improvement from where we started.