Color out of B&W in PS at the end

greenspun.com : LUSENET : Large format photography : One Thread

This is a somewhat naïve and worse case scenario question but here goes. Let’s say in 4-5 years color LF film has disappeared. However, I stash several thousand sheets of B&W film in the freezer. Would there be a way to simulate color by:

1) Developing a standard of calibration in PhotoShop so that the density of the primary colors could be associated with B&W neg densities. 2) Shoot a three sheets of a scene, each filter with one of the primary colors (or is it the compliment?) 3) Scan the three negs. 4) Marry the three negs up in photoshop, and then “apply” the color based on the neg densities. 5) Make a color light jet print.

I suppose this would be a bit like dye transfer, at least what I know if it, which isn’t much. Sound crazy?

THX

HY

-- Hyperfocal Yokel (hyperfocal@attbi.com), January 11, 2002

Answers

Sounds like it should work in theory. The problem being that you would have to have the subject and the camera absolutly still. Any movement in either would cause on the the negs to be out of alignment with the others. Thus you would have a strange color seperation. As for the photoshop side. I would say i'm an expert but I have used PS for seven years so here is my advise. While I don't think color neg fims will be discontinued, I'm pretty optimistic. In PS it wouldn't be too are to simply as each image as a color layer and add or subtract color from the channel. Your colors will look pretty weird but after some adjustment should look slightly normal. However, you will most likely have to do this adjustment for each image you take, so quite a bit of digital work will be involved. Mainly to solve something like the color bias of each scene and produce an output to your vision. I sounds like it might work, I would say you should try out the idea now instead of waiting until its too late. It worked for magazines and publications before the digital age, it should work now. Heck, by the time they discontinue color neg sheet film, digital quality might have excelled enough to be a replacement for the film. Well, try your idea and don't let anyone tell you otherwise. Just realise it will take you a long time to get the colors right.

-GD

-- Greg De Stefano (mink@minervamink.zzn.com), January 11, 2002.


Err, I mean I'm not an expert at Photoshop, just a very experienced Amateur. Gotta stop posting when I have a headache.

-GD

-- Greg De Stefano (mink@minervamink.zzn.com), January 11, 2002.


I'll bet there are some ways to deal with the potential registration problem. In my work (molecular biology) we can sort out proteins using two criteria (size and charge) to get a two dimensional map of the proteins. The end result is a bunch of dots in different positions. If we do this with two different samples (cancer versus normal tissue) we can identify proteins present in one sample but not the other.

The problem is that when we sort out the proteins, the final results aren't directly superimposable. There are computer programs that can take two different samples, find "landmark" points that are going to be the same in any protein sample, and from those, warp the image so the two are superimposable.

I've seen demos of it, and it works well enough for what we do. Whether it would work well enough for photos is another question, but the basic technology is around.

DW

-- Dave Willis (willisd@medicine.wustl.edu), January 12, 2002.


I suppose you could measure the gray values in PS, convert to RGB, then use , and to get the color values you want, measure those color values and make some sort of log to keep track of both, but that's a tedious process.

I'm unaware of any PS facility that would associate grayscale/color values, (other than measuring them seperately yourself) but someone else may ... there are a number of PS forums, visit Adobes' home site for the links.

-- Michael Mahoney (mike.mahoney@nf.sympatico.ca), January 12, 2002.


something happened to my answer above, after " then use" - there should be "color balance, hue/saturation"

-- Michael Mahoney (mike.mahoney@nf.sympatico.ca), January 12, 2002.


So, isn't this one way in which color got started? They had a special camera that used image splitters (partial mirrors) and filters to disect the image into it's three component colors on three sheets of black and white film. Because of the image splitters, all three sheets were exposed at the same time, which took care of the movement problem. Then, they brought these three images together using presses with the three corresponding colors of ink. They took photos of movie starts this way.

Except that, the three negatives would be scanned into Photoshop. I'll bet that, through dodging, exposure, development time, ink selection, etc., they had some of the same flexibility in reproducing the image that we currently have in Photoshop.

Isn't it interesting how things can come full circle.

-- neil poulsen (neil.fg@att.net), January 12, 2002.


Is there anyone that thinks color 4x5 will be unavailalbe in say.

2-3 years? 3-5 years? 5-10 years? 10+ years?

-- Jonathan Bujndick (jebdlb86@worldnet.att.net), January 12, 2002.


How quickly we forget!

Just a few years ago almost all color photography in print was made with separation (b&w) negatives as a necessary pre- press operation using these same principles, but using printing press inks instead of inkjet printers.

Just instead of photographing the colors of the original scene using the primary filters they copied a transparency with three exposures to make separation negs through filters to result in the CMYK plates that go on the press.

-- david haynes (studioblsp@mindspring.com), January 12, 2002.


Neil's right. Think "autochrome." Or look up gum bichromate printing.

-- Chad Jarvis (cjarvis@nas.edu), January 13, 2002.

Well actually, the autochrome was made by a quite different process, involving no filters and only one black and white exposure. The color was created by a layer of colored starch grains either under or over the black and white emulsion. I've never been able to quite wrap my mind around how it could work that way, but that's my understanding of the mechanism.

But as for Hyperfocal's question, most respondents are making this harder than it is. If the three exposures were taken without movement of the camera or the subject, the three resulting black and white negatives or slides should align perfectly and when aligned, should produce a perfectly colored picture, if the right filters were used. All you'd need to do in Photoshop would be to layer them in the same document and create an RGB file. That's it.

-- Katharine Thayer (kthayer@pacifier.com), January 14, 2002.



Correction: I used the word "layer" when I meant "channel;" you don't use the three scans as layers but as channels in creating the RGB file.

-- Katharine Thayer (kthayer@pacifier.com), January 15, 2002.

Years ago I read of Land demonstrating color images with two black and white transparencies projected simutaneously on the screen, each transparency taken from the same camera in the same location, no chanmges made other than switching filters, one projector with the same color filter mounted on the lens as was used when taking the original transparency placed in that projector, and the other projector without the color filter used to take that image. So much time has gone by that pehaps I am mistaken exactly how he did it. As I recall, it did not matter which two color filters were used, so long as they were of distinct colors.................................... Perhaps he or his company Polaroid patented

-- Quien No Sabe (caldw@aol.com), January 19, 2002.

Moderation questions? read the FAQ