We’re getting closer to a feature freeze for the darktable module and in this last minute I have to air a question (or possibly an idea).
What is the correct workflow for printing a Spektrafilm developed file? Simulating the paper in Spektrafilm and then sending the file to printing via lab on an actual paper seems wrong.
I print my files via lab on Fujifilm Crystal Archive, which incidentally happens to be one of the paper profiles offered by Spektrafilm. I wouldn’t want to “double-paper” the lab print.
Could there be a way where I develop my raw image in the Spektrafilm darktable module twice. One version with the paper profile for on-screen viewing and one without the paper profile for sending to the lab? If I use the Crystal Archive profile for the on-screen viewing version and Crystal Archive paper for printing, they should end up looking roughly the same.
I guess I’m looking for state between checking the scan the film (skip print) checkbox and adding the print profile (working with the orange base is terrible) — like a neutral print ready state. I don’t have the correct words for describing exactly what I mean, but I hope you will understand.
Is this even technically or mathematically possible?
The problem with this is that without the paper, we just have a negative. And to view that negative you would need to then invert and white balance it separately, similarly as you would for real scanned negatives.
There isn’t really a print profile that gets applied to a neutral pre-print positive image, but rather modeling of the actual exposure that happens by projecting light through filters and the negative on to the paper and how the paper chemically responds to that.
Thanks @niklasiivari, what I suspected. What about creating a theoretical paper profile for a paper that doesn’t exist, that is as technically neutral as possible? Let’s call it “Spektra Neutral”.
You can technically hack this in, I tried it before (in python though) - just generate a linear print paper profile and deal with the consequences (edit: you could “hack” it, but it has been quite some time ago and I guess things have changed in the code, I now cant get it to work at all haha)
I guess that the “correct” way would be to build and include a whole separate scanning section for the negative stocks, something that inverts the negative, applies white balance, gives you some real world based controls…, I am sure that this was something that was considered and will eventually be done.
am curious to know what would happen if you sent the file to your lab.
since fujufilm’s CA is the only in production type C (chromogenic) paper made, every RA-4 wet-print made is on this product. If the roundtrip analysis is accurate, then your onscreen image could be considered a soft-proof.
Aren’t we talking about two completely different types of paper, though? One is a paper for accepting printer dyes and the other is for making optical prints with an enlarger from a negative, am I mistaken about this?
The paper being simulated in Spektrafilm is a fundamental part of the picture making process. The paper being printed on in the lab from a digital file is just a blank white surface prepared for accepting inks/pigments, even if they carry the same brand name. All I really care about is being able to match the print to the image I see when I processed it. I don’t think skipping the “paper” step in Spektrafilm is the means to this end.
Not at all. No ink involved. Look up Durst Lambda or Fujifilm Frontier digital C-type laser prints.
They take a digital image and use red, green and blue lasers or LEDs to expose conventional light-sensitive photographic paper that is then developed using RA-4 chemical processing.
Fujifilm Crystal Archive can be used both for conventional optical enlarger printing and for digital exposure systems such as Lambda.
Hmm… Perhaps something interesting could be extracted from an ICC profile used for soft proofing Lambda or Frontier prints on Crystal Archive.
On the other hand, there must still be something I’m missing about your original idea, so I’m still left confused. Are we talking about a lab that prints uncorrected scans of actual negatives? Then I understand sending them the image without the simulated printing step.
If we’re talking about a lab that prints regular digital images, we’d still be talking about a printing process with some kind of profile used in their own machines, only it’s something that translates RGB values into whatever they need for the lasers instead of what they need for the dyes. It still sounds to me like the regular printing challenge of using profiles in our workflow to get what we see on our screen to match what we get with the resulting print, which doesn’t really change whether we’re using Spektrafilm or not.
I’m sorry, I feel like I’m missing something basic. I’m not good when it comes to advanced color profiling.
My point was that the ICC profiles from the lab should isolate or reveal how the print step (for Fujifilm Crystal Archive) alters the colors/image — @arctic, @hanatos and @piratenpanda. Somehow subtract that difference between file and physical paper print and we could perhaps create a neutral output option or preset (that does not simulate the paper) in the Spektrafilm module.
We are talking about a lab that prints a file (on Fujifilm Crystal Archive using a chemical process) created using Spektrafilm.
Yeah, that would be cool but that again requires that we can somehow neuter the paper emulation step in Spektrafilm. Or else we would be “applying the paper simulation” to the file that is later used for an actual physical print on the same paper.
Every film profile has its own special adjustments for every print paper, they are quite wonky on their own, Andrea wrote about this somewhere. So with the linear paper profile I made there were some wild color casts and shifts. But since then there has been some changes in where these adjustments are saved so now a linear profile on its own doesnt work anyway and I do not want to deal with an LLM to fix that
Anyway, I have been experimenting with negative spektrafilm output and negadoctor in darktable and I think it is a very viable solution.
it may be that the founding concept of this spekt* is as an effects/simulation system meant to reside in electronic space exclusively.
i am just recently greping thru the code in effort to verify the emulsion chemistry subs.
it may be posible to simulate C-printing from the 70s using current digital-C …
was that the ICC profiles from the lab should isolate or reveal how the print step (for Fujifilm Crystal Archive) alters the colors/image — 