View Full Version : How to test film for in-camera colour separations

David Nash
11-Mar-2005, 14:00
I'd like to know how to go about testing a film for making in-camera colour separations.

What should I be looking for when looking at spectral sensitivity and density curves?

I'm thinking of using Bergger film that has a linear straight line, but the spectral density chart looks like three skewed humps, with increasing peaks at 450-nm, 575-nm, and 690-nm.

I do have a densitometer that I got a while back, and which I've never used.

I have Lee tri-colour filter set 25, 58 and 47B. When I phoned Lee, they told me each of these filters has the same filter factor (6x) even though, to me, the 47B appears to let less light through.

I don't really understand what I need to achieve an 'ideal' set of negs, so I'd really appreciate some pointers.



11-Mar-2005, 15:17
To achieve an ideal set of separation negatives all three records should receive the correct exposure for each of the three filters, and be developed to the same contrast or density range. Regardless of what you read this will take some trial and error.

If either the density or contrast is off in one or more of the separations you will have color and contrast cross-overs that will prevent you from achieving correct color balance.

If you don't care about color balance, as as in an artistic application such as 3-color gum printing, you may be satisfied with less than perfect separations.

Joe Smigiel
11-Mar-2005, 16:37

I tried this once with a set of 29, 61 and 47B tricolors and found the filter factors were different. IIRC, the tungsten factors were 5x, 8x, 12x respectively. This may be an incorrect recollection as I did this probably 13 years ago, but it sounds like your set has a similarly denser blue so I might suggest using these factors as a starting point. Your filter factors will vary with the color temperture of the exposure source so perhaps Lee was quoting a daylight factor of 6x for each filter.

Sorry I can't help with any development recommendation although it seems to me the blue negative also had lower contrast than the other two and required more development.

Andy Eads
11-Mar-2005, 16:40
David, Sandy has described the pitfalls well. Doing the testing isn't too hard but it does take some systematic work. The trick is to photograph a gray scale with each of the three filters and plotting the characteristic curve for each filter's negative to see how the exposure and contrast vary. From the plots, you should be able to determine quickly if the slope (contrast) varies with the filter and how the film's spectral sensitivity responds to the filtered light.
Consistency is the key here. You must use the same lights, film emulsion, developer, filters, agitation technique, etc. The contrast index will be determined by your final application for the separations. I made separations for dye transfer printing many moons ago and the testing wasn't too hard to do.
Have you searched any of the sites that cater to alternative printing methods? I think there are some die-hard dye transfer printers out there that can help you along. Kodak published a guide called the "Three Aim Point Method" for the graphic arts (offset printing) field. The main ideas and testing methods were spelled out pretty well as I recall. You might be able to find a copy of it moldering in some old book store. Good luck!

Jorge Gasteazoro
11-Mar-2005, 16:57
David, do what I did to test filter factors. Take a series of exposures of a gray card and bracket around the filter factor written on the filter. Develop the sheets and the one who has a density for the gray card of 0.65 is the one with the appropriate filter factor. This acomplishes tow things, it tells you the "correct" filter factor for you and it gives you 3 separation negatives to test your procedure.

Of course, if you are doing a process that requires a denser negative like carbon or gum dichromate you might want to "move" the density of the gray card a notch up the curve, maybe use 0.75 or 0.8.

If you plan to do dye transfer then a "zone V" of 0.65 density should serve you well. Since JandC is selling dye transfer matrix film I have been very tempted to try it. Let us know how it works for you.

David F. Stein
11-Mar-2005, 20:43
I agree about the gray scale calibration. Why not also try with a set of 3 filters that is a bit off red, green and blue-just as painters will work with different palettes of primaries. See what happens. This is what is missing in our computerized, clip art world-going with accidental discovery; doing something because we don't know what might happen with the light. This is why Ansel Adams was much more an artist, albeit a rigid one, than a photographer. He was trying to take reality and translate it into a vision of reality. No skies are as black, no rocks quite like Ansel Adams photographed them. I still recommend that Library of Congress ite-that fellow had it worked out well many moons ago. One key was the sequential plate that just dropped into place. No putting 3 film holders in a row in the camera-which causes delay and registration difficulty. I bet an 1x14 format used verticlaly to give three 4+ x 11 images would work nicely. GOOD LUCK.

Paul Fitzgerald
11-Mar-2005, 23:23
Hi there,

(Sandy, check ebay # 7500196263, National One-Shot 5X7 kit, just listed. Good Luck)

David, the filter factors for tri-color are different than filter factors, they were based on the red exposure: and yes, they are very different by light source.

For SuperXX they were:

Tungsten R 1.0 G 2.0 B 2.0

Photoflood R 1.0 G 1.8 B 1.1

White Arc R 1.0 G 0.8 B 0.3

For Tri-X they were:

Tungsten R 1.0 G 2.9 B 2.0

Photofood R 1.0 G 2.1 B 1.1

White Arc R 1.0 G 1.3 B 0.3

Kodak had made targets with 6 patches on them to add into the photo. Red, Green, and Blue to identify the filter used on each neg: Black, .50 density gray and paper white to check exposure and development. They wanted exposure for .30 density for the black patch on the neg, development for 2.40 for the white and check the gray patch to match contrast across the set. It gets a lot more involved but that is the quickest way to a baseline.

It only took Kodak 100 years to figure out that using the tri-color filters to make the color correction masks doesn't work well. The filter set-ups for the masks are completely different for seps made from an original, reflection piece or from a transmission slide.

Hope it's a help