Would this be colour negative or colour transparency?
color negatives! thanks
1 Like
Thanks, ChatGPTās response isnāt necessarily logical. Whilst itās true that LEDs donāt have a continuous spectrum your panel is very popular as you say. Also an enlarger is printing on to RA4 paper with its own particular colour response and specifically it is designed to work with the orange mask of colour negative. Might it be possible to process in NLP in a way that would prevent these out of gamut reds?
I have tried all sorts of conversion settings, including linear flat, pre-saturation set to 1, etc. I think it may be more of an issue with Epson ink gamut, as I dug some old Imacon scans and also have reds going out of gamut rather easily. But it did get me concerned about the spectral integrity of my light source compared to other options. I had read about spiky red frequencies before and am wondering the best way to ensure I get a balanced scan.
Thinking aloud really but if you had too much red from the spiky led panel wouldnāt you end up with too much cyan in your inverted processed positive? Certainly worth trying a different light source but tungsten might be a bit on the warm side when dealing with the orange base. Maybe flash just as an experiment?
Iām going to test out flash as well, even though it is a clunkier workflow. Also considering just popping for an expensive Neg Supply light, but it might not make as big of a difference. It may make more sense to use the cheap light for general scans, and then rescan the important ones with flash before printing.
I agree about the clunkier workflow, panels are just so convenient. However if you can rig something up with flash as an experiment, maybe bounced off white card or āstyrofoamā, then you wonāt get a purer light source. You could look into that new CS-Lite+ Spectracolor as well. For colour negative if you are thinking about expensive NS high CRI panels then there are also RGB options becoming available , still a bit experimental though and not necessarily recommended for use with NLP.
Iām also wondering if it such a bad thing to end up with out of gamut colours in a āscanā, who knows but you may even get more with a high CRI panel or flash. The alternative would be to only get colours that fit into the Epson colour gamut, isnāt it just a question of how best to deal with out of gamut colours when printing?
https://www.dpreview.com/forums/threads/tips-for-printing-out-of-gamut-colours.4742072/
Ran a test with a flash and a small softbox, and the inversions were definitely cleaner than with my panel. The output gamut issue is indeed separate and not really a problem in practice. Going forward, I will continue to scan with a panel, and then will go back and flash-scan anything important enough to print. It is just too slow to use flash on everything, unless I get a proper monolight with modeling lamp.
1 Like
Now that I mentioned a modeling lamp, Iāve been thinking about implementing something like that. I found a normal size flash with a modeling lamp, so it makes it easier to use on a desktop. Has anyone use one of these (link below), and was it easy to focus with the modeling lamp @ 2.8? I donāt think the modeling light shuts off when the flash is fired, but Iām guessing the difference in brightness between the led and the flash would be big enough to effectively eliminate any exposure from the led.
Not super lively around here latelyā¦
I am looking into the Jackās Big Scanlight vs flash, and am curious if anyone has compared? Just in terms of sheer color quality for negatives. I am aware of the workflow differences.
I recently switched to the big scanlight and the results are well worth it compared to a 95 CRI white light. Much better saturation, vibrance (sometimes a bit too much) and color depth with more accurate skin tones. Reds and greens are particularly enhanced. I think the concerns about NLP not being āoptimizedā for RGB lights is a bit overblown. I get better consistency than with the 95 CRI, with only a few head scratcher conversions here and there.
I have not tried flash, though. The workflow considerations with flash wouldnāt make it feasible for me.
There are two big problems with the flash workflow. One is that most people attach the neg to the light source rather than the lens. Iām not sure why this became the standard, when it makes more sense to attach the neg to the lens. I bought the very flawed Valoi Easy35 when starting out, and removed the bad video light they put in there. Now I can I just push that up against much better light sources and it gets great results. The larger light sources manage to wrap around the shadow-casting thumb indentation the ādesignersā at Valoi erroneously jammed in there. Not a good purchase economically, but it worked out, since no one else makes a usable lens mounted neg transport (thereās the Nikon, but the holder makes it clunky).
The other is seeing what youāre doing to advance film and focus. If I were to continue with flash, I would probably pick up a Godox mini monolight and mini softbox. They have some with modeling lamps that automatically shutoff when the flash fires, like any standard monolight.
I ordered a Scanlight, so weāll see how it compares. Flash results are very good, but I expect the RGB scans will match or exceed it.
Did you have to do any tweaking of the RGB channels to get optimum results with the Scanlight?
I have done some very cursory tests of adjusting individual color channels but it seems that up to a certain point NLP just brute forces the curves to give nearly identical results. After that the colors get totally wonky. I imagine individual tuning is more applicable for manual conversions.
One workflow issue I have had is that the RGB light creates a vignette with a tint shift that does not occur with white light. It may have something to do with how the camera sensor interprets the information. I have to compensate for this with a custom radial mask or FFC or else I get reddish vignetting in the corners. I discussed this on another NLP thread and someone mentioned that elevating the film plane farther from the light source resolved the issue. I still havenāt determined what exactly causes this so if anyone has insight Iād appreciate it. I bought some NR glass to elevate the film plane and hopefully that works because it is quite annoying. That being said, even with this error the scans are superior. YMMV
So do you use the panel as a white light source, or do you adjust the colors on the panel itself to negate the orange mask?
I use the RGB mode with all colors at 100% brightness which itself neutralizes the orange mask due to the blue bias of the RGB light combination. I have not tried tweaking individual color channels to more accurately negate the mask because results are great left as is. Would be a hassle with the variation in mask color for each film. Creating presets could help though, thatās possible in the scanlight software.
Bonus is that you no longer need to white balance in LR. Conversions look better when skipping this step. Makes converting different emulsions at once much faster.
1 Like
Thanks. Really looking forward to trying this out. I have been working on getting better manual inversions in PS so as not to rely on NLP/Negmaster/Colorperfect so much, and I expect this will make things a lot better.