Additive or Subtractive Colour?

Hi all, for my year 12 major work I am hoping to explore historical colour processes.
I am interested in doing very similar things to Sergey Prokudin-Gorsky, using a three colour method. I am a complete newcomer to anything analog to do with colour, and was wondering if I should shoot my negatives through additive (RGB) filters, or subtractive (CMY) filters for printing?
If I shoot through RGB, do I print with RGB or CMY?
If I shoot through CMY, do I print with CMY or RGB?
Does it really matter at all?
I've overthunk it and got myself into a bit of a pickle.
Thank you for any advice!

He used the seperation process, in which three seperated images were taken (in his case even in sequence).

Such a process uses three continous segments out of the spectrum. Thus the additive range. Though in this context one does not speak of additive, as this term is applied for an image reconstruction part.
They typically are called seperation filters. The respective Wratten designations are 47B/58/25 respectively 47/61/29.

Prokudin-Gorsky used equipment designed by researcher, scholar and photographer Adolph Miethe.
Prokudin-Gorsky become known by the internet, same as Autochrome. In both cases they are typically depicted as unique, though being only representatives amongst others of a certain process.

Shooting 3 separate negatives might be a bit difficult. You could try shooting a professional color negative film and scanning it to make RGB color separations. Then put it back together on a computer.

If Kodak still made dye transfer products you could spend a lot of time and money.

You could shoot your film scan it, do a separation in PS. then you could adjust each channel. Make a print of each channel and the combined print, with inkjet. You could even print RGB channels on clear film and assemble for a full color transparency.

We are speaking as well of a historic workflow as we are in an analogue-only forum. Advise for a hybrid workflow is not apt here.


That guys back then had better equipment than we today, should not keep us from trying.
A good camera mounting and the ability to keep the three sequential negatives in register should be enough for the start. What then rests is the time-parallax, the non-register due to subjects movements (eg. leaves in the wind, people moving).


What to do with the gained negatives would be the next question. Miethe used them for printing albums which then got published, but he also reversal processed/copied them and employed a three-slides, additive projector for screening.

Making the 3 gelatin matricies for a dye transfer print will be extremely difficult for your project. I'd suggest shooting B&W and processing it as POSITIVE then projecting the images with 3 slide projectors and the same filters over the projector lens you used for the initial exposures. That should be pretty easy to accomplish. That should be very similar to what Sergey Prokudin-Gorsky did in his lifetime.

Maybe the mods can remove the unwanted posts in the 100% analog section.

If this is a student project, rather than spending hundreds of dollars on optical glass separation filters, I'd recommend students obtain something like this (used to be able to get it for free) and place these over the lens and projector.

You're not going to kill a charging rhinoceros with a few toothpicks for a spear. You haven't overthunk it. It's going to get a thousand times more complicated if you expect quality results. Look up Andy Cross down under and see if he offers classes. He's conversant with digital, Ciba, Dye Transfer, and Tricolor Carbon. If you want to do a tricolor slide show you'll need three aligned projectors, each one fitted with a different RGB filter. Use the same filters over the camera lens. Try 29 red, 58 green, and 47 blue with TMax 100 film. It will take a lot of experimentation to get the film curves to all match. Helps to own a b&w transmission densitometer; and that's an understatement. If you want to learn traditional RGB printing methods, set aside 10 to 20 years.