An experiment of sort (color from B&W)

I remeber reading a long time ago about various photographers who would shoot the same scene with 3 seperate emultions, usually chemistry thats sensitive to red, green and blue, so that when the three glass negatives were projected in a projector with filtered glass the results would be a color image (magic latern I think they'd call it)

I'm thinking What if I can imitate the same with panchromatic film?

I have an 80s Nishika 3-D camera that takes four half-frames per shot (two frames total, with 4 shots) thats used for a stereo camera. What if I filtered each lens, Red, Green, Blue, and Clear (reference) Course I know the slight seperation in distance may affect the image when trying to merge them due to the principles of a stereo camera, but how would panchromatic film behave to colored filters and would it replicate results similar to how the past photographers tried color from B&W glass plates.

I once experimented with direct separation inspired by the two-colour Technicolor process . Involved a camera on a tripod, some panchromatic film, and "absolute" red and green filters. Absolute meant that the red passed only red and absorbed everything else; ditto with green. An absolutely still subject is also a must, so in the experiments, I was limited to shooting table-top fruits and objects and the rare instances when my nephew was in his 'rest' mode.

The film was reversal-processed, and then mounted in slide frames. The results were just two B&W slides with strange tonalities and absolutely no colour. To view them in colour, the slides were put in two slide projectors which were arranged so that their beams superimposed. No matter what we did, those Ekta projectors refused to cooperate- their beams never really jived inspite of having matched lenses. Without the filters (same filters used for the camera), a plain B&W image was seen on the screen. It also looked somewhat diffracted since the two super'ed images didn't exactly meet.

Then the red filter was placed over the "red" slide. A semblance of a coloured image started to be seen. The reds in the picture now look red. For some reason (perhaps due to some weird visual perception phenomenon) the non-reds seemed to become coloured as well. Likely due to accomodation, when the eye started compensating for the mostly red picture seen in a darkened room. When the green filter was placed over the green slide image, the scene started to look right. It was now in colour!

Two colours then seemed to be right enough. The lack of blue wasn't really a handicap. Fruits, flowers and even human skin looked natural enough. But there were no real blues- the really blue objects came rather grey, and since blue objects really reflected some red or green, these came in various shades of greenish blue. Cinecolor (the cheaper, 2-hued competitor of 3-strip Technicolor) movies and cartoons looked real enough to many viewers in the 1940s...some of these cartoons can still be seen on TV today.

I should repeat those experiments. Those experimental slides were lost in school. That was long before I had scanners...

Jay

In theory what you describe should work; however, as Jay details, it'd be tedious at best to get it to work in practice. The main reason I'm replying, though, is to provide a link to the Empire That Was Russia Web site at the Library of Congress. This site has a collection of color photos by Sergei Mikhailovich Prokudin-Gorskii (1863-1944) that were taken close to a century ago using multiple plates and filters. The LoC digitally re-combined the images, producing some amazingly modern-looking color images -- at least as viewed at computer screen resolutions. I used one of these photos as my desktop wallpaper for quite a while.

They used to make tricolor cameras back 50+ years ago for just such color photography. Then they used dye transfer, carbro or bromoil to make color prints.

Nothing new under the sun. Just do a search and you will find more than you ever wanted to know about the subject.

PE

srs5694 said:

In theory what you describe should work; however, as Jay details, it'd be tedious at best to get it to work in practice. The main reason I'm replying, though, is to provide a link to the Empire That Was Russia Web site at the Library of Congress. This site has a collection of color photos by Sergei Mikhailovich Prokudin-Gorskii (1863-1944) that were taken close to a century ago using multiple plates and filters. The LoC digitally re-combined the images, producing some amazingly modern-looking color images -- at least as viewed at computer screen resolutions. I used one of these photos as my desktop wallpaper for quite a while.

Click to expand...

That LoC link is what got me thinking.

Also what I'd probally try to do , and I'm not aiming for perfect alignment is to take these four half frames, scan them in, and assign them as color channels in photoshop. At best what would happen is the objects closest to the camera would be best aligned, and as object went further away the colors channels would seperate further. And the experiment would be to find where the 'sweet' spot would be for the 4-lens stereo camera so that that object would be in closest-to-good color, and the rest would slowly become kind of a tri-color mix being seperated. And the 4th frame could be an unfiltered 'reference'. Tho if I were to do so, I'd wonder where I can find some 'absolute' filters, possibly square in about half an inch to an inch in diameter.

This is the camera I speak of


an example out of it

(try to cross your eyes on the middle two (easiest), its a cool effect)

Another one

(lil harder, gota cross til you see 3 of em, with the middle being 3Dish)

So yea thats the camera I'm thinking about utilizing for the experiment as it has 4 lens that fire simutaneously.

Use whatever you can find OR request/buy a filter swatch book from Tiffen or Roscoe that has sample color gels.

Yeah, precise filters would be keen, but what makes you think the pioneers had those? As long as you use the same brand filters for taking and projection, you are good to go.

A 4 lens stereo camera will not work.

Each picture is different and the final negatives will therefore not register due to the horizontal displacement. The result is severe color fringing.

A camera for this type of work has a single lens with a beam splitter to get identical color separations. The beam splitter / color filtration method uses standard panchromatic film to achieve the R/G/B separation negatives which can then be used to make dyed separation positives for printing.

PE

You can also solidly lock down any camera and shoot three shots with successive RGB filters; of course this kind of kills any action shots...

The only way to get action shots is with one of the tricolor (one-shot) cameras which has the three filters for color seperations built into them as PE pointed out. The one thing that you have to be aware of if you try and shoot the three negatives in a view camera is that each negative will have to be processed at different times to allow for correct calibration of the resulting seperation negatives or the exposures can be adjusted with ND filters to allow for all three seperations to be processed together which would then allow for roll film to be used. TMAX 100 is probably the best film for this for anyone trying this but the combination of ND and color seperation filters will result in TMAX having an effective ASA of about 6.

Gord

Just FYI: I think it was (is?) standard practice in the movie industry to create separation negatives in b/w off of colour films - because the archival quality of b/w is so much better than colour. (Some colour films used to start fading after 6 years - in the dark!)

Simon

Photo Engineer said:

A 4 lens stereo camera will not work.

Each picture is different and the final negatives will therefore not register due to the horizontal displacement. The result is severe color fringing.

A camera for this type of work has a single lens with a beam splitter to get identical color separations. The beam splitter / color filtration method uses standard panchromatic film to achieve the R/G/B separation negatives which can then be used to make dyed separation positives for printing.

PE

Click to expand...

As mentioned before I'm not looking for perfect alignment and what I said might happen is as items get close to the camera towards the 'sweet spot' the mor enormal it would appear, were as the further away the more frindge as you say.

Wasn't the original Polaroid color, a two color process as well. Apparently when the eye is given the correct information at two wavelengths it is able to fill in the middle colors itself!

Simon.Weber said:

Just FYI: I think it was (is?) standard practice in the movie industry to create separation negatives in b/w off of colour films - because the archival quality of b/w is so much better than colour. (Some colour films used to start fading after 6 years - in the dark!)

Simon

Click to expand...

Yes, it is still the preferred way to do it, BUT sequentially on the same strip ( 3000 foot roll for a 1000 foot original, NOT as individual RGB matrices as the original Technicolor camera shot.

Differing gammas and shrinkage rates across rolls of stock make it very hard to recombine discreet RGB rolls into a coherent color image.

Frank

Kino said:

Yes, it is still the preferred way to do it, BUT sequentially on the same strip ( 3000 foot roll for a 1000 foot original, NOT as individual RGB matrices as the original Technicolor camera shot.

Differing gammas and shrinkage rates across rolls of stock make it very hard to recombine discreet RGB rolls into a coherent color image.

Frank

Click to expand...

Frank;

Special films on estar support are made for making color separation from negatives. These B&W separations are archived at the major studios and at George Eastman House. They are relatively easy to print from. After all, remember, Technicolor and Dye Transfer used this type of process.

And, to answer another post, the first color films were indeed 2 color. They eye adapted to this and supplied the 'missing' color, albeit poorly. Early Technicolor was an example.

PE