Tricolor digital print

[brennig]

Hello people,

Please scrutinise my plan: Take three shots (of the same subject/perspective) with B&W film on a 4x5 camera, one unfiltered, and two with a magenta and yellow filter (changing the shutter to create uniform exposure). Develop and scan each negative. "Subtract" the two filtered negatives from the unfiltered to create a Cyan channel.

Now I have four separate images/colours/layers producing CMYK. Print each CMY layer onto clear acetate (like a not-as-good dye transfer, or technicolor print) and the Key layer onto white paper. Slightly space out the layers, so as you move about the print (not looking at it straight on) you get slight chromatic aberrations. Obviously this will require a lot of trial and error, getting the levels right and all, but I am working with a professional print shop specialising in transparencies, I'm sure they can help me out.

WILL THIS WORK? Do I need the key channel? Will subtracting M & Y from the unfiltered shot create C? I know and love Prokudin-Gorksii's split colour images, designed to project RGB images. But as I'm printing them then I am better off splitting with Cyan, Magenta and Yellow filters?

Before investing in filters, it would be good to know whether I should get RGB (25, 58, 47 in Wratten numbers) or CMY (44A, 32, 12) coloured filters.

THANK YOU! Be kind to me, first post.

[Rich14]

You need to re-visit color theory. You are trying to recreate the path that an image takes in the E6 process and misunderstanding how to extract the information to produce the Cyan, Magenta and Yellow dye layers that are presented to the eye in a subtractive manner. That CMY subtractive sandwich in the resulting "transparency" looks like the original scene.

To obtain the Cyan information requires exposing B&W film through a Red filter. To obtain the Magenta information requires exposing B&W film through a Green filter. To obtain the Yellow information requires exposing B&W film through a Blue filter.

The resulting B&W negatives are referred to as the Cyan, Magenta and Yellow "printers." They are the true "color separations" originally referred to in the printing industry when this mechanical process was in common use.

Those negatives are then used to expose three separate sheets of photosensitive material that when suitably processed chemically, result in, respectively, Cyan, Magenta and Yellow positives that are sandwiched together for the final, composite image.

You cannot obtain Cyan information by subtracting Magenta and Yellow from an unfiltered B&W image.

Rich

[barnacle]

You can, however, extract green from an unfiltered B&W image, provided you have the red-filtered and blue-filtered images.

This is something you might find easier to practice with in photoshop or the gimp. One helpful tip is to label your negatives when you make the image (perhaps a small label in a corner of the scene). It's very easy to lose track of which negative is which.

Another tip is to recall that the colour images do not add together in equal amounts: for fully exposed red, green, and blue (strictly, Y-R, Y-G, and Y-B as negatives) images you will want close to 30% of the red channel, 59% of the green channel, and 11% of the blue.

Neil

[sanking]

You need the R (25), G (58) and B (47) filters for making separations from life on B&W panchromatic film. You may get similar results with specialized work flow with other alternatives but the 25-58-47 set is the one used in most traditional three-color work.

There have been several threads on this topic on the LF forum, including this one.

http://www.largeformatphotography.info/forum/showthread.php?48583-Color-photography-with-black-and-white-film

You can find full instructions for the hybrid work flow, including assembly in PS, in The Carbon Print. https://johnlockhart.net/carbonbook/

Sandy

[brennig]

Sorry for the brain fart. I figured it out in my head that the CMY would be the invert of the RGB negatives. Easy.

At the moment I'm shooting FP4+ (I'm based in the UK), but looking at the technical sheets it seems HP5+ has a "better" spectral sensitivity, that it covers RGB wavelengths somewhat more equally (as is done with colour films, the FP4+ seems to drop off on the lower end of the spectrum). Is it right to say HP5 is more panchromatic then? Given the filters will require slower shutter speeds, the increased ISO is a plus too.

I'll do testing with medium format first of all, especially to get the balances right for printing on acetate. I imagine the background Key requires some tinkering to get the required contrast. Sourcing the filters (at a reasonable price) is the tricky part. Fortunately the Formatt Hitech factory is near to me and I've asked them if they have any tricolor separation filters in a dusty corner of their warehouse (they've discontinued it). They'll get back to on Monday.

Thanks for the help anyhow!

[ventdesable]

Hello,

You really should read Henri Gaud's work on "trichromie". It will give you all the information you need. And, don't know if he still can do it, but he used to give filters to people asking for them. Those filters are gelatins from LEE Filter. Lot cheaper than the ones you are talking about.

Here is the link to PDFs he has done : http://trichromie.free.fr/trichromie/index.php?post/2015/11/23/PDF2015. Text is in French, but, pictures, drawings and curves are... international.

You can contact him directly on this mail address : contact@henrigaud.com. He speaks English.

Have Fun.

Jérôme

[brennig]

That website doesn't seem to work on my computer, it's pretty frustrating as I've been trying before.

Thanks for his email though!

[Pere Casals]

Adding to what Sandy said about R,G,B filters, you can do preliminary tests by using a DSLR, then do the image stitching with Photoshop, use fill a layer (http://www.kennethleegallery.com/html/tech/bronze.php) to paint each layer with each primary color, and then add the layers directly.

A reason to slightly vary the particular RGB filters is to obtain nicer skin tones. Skin tonality depends a lot on that, for example Canon skin tones are different than nikon, and this is because the particular RGB dyes used on on the sensor, Photoshop like adjustments may it make similar, but not completely the same.

Here, a tri photo I have in my Flickr favourites gallery (plats moved a bit from one shot to the other):

https://c2.staticflickr.com/6/5608/29797289360_bae45d5a72_b.jpg

[RedGreenBlue]

That website doesn't seem to work on my computer, it's pretty frustrating as I've been trying before.

Thanks for his email though!

Perhaps this description of how to assemble scanned separations is Photoshop will be helpful: http://www.vintagephoto.tv/photoshop_assembly.shtml

Scott

[jnantz]

just scan the negatives ( RGB ) as sandy king stated
then drop the images into a new file's color channels
its a lot of fun
good luck !

[Drew Wiley]

If you want to do it with film, TMax is by far the best current choice. But because a 47 or 47B blue filter is so slow, one strategy is to use TMY400 for the blue exposure, and TMX100 for red and green. I do know of one person doing it with FP4, but the blue separation is a headache requiring overdevelopment.

[ventdesable]

That website doesn't seem to work on my computer, it's pretty frustrating as I've been trying before.

Thanks for his email though!

here is a direct link to the pdf : http://trichromie.free.fr/aipc/public/SupportcoursPdF/1516-Trichromie.pdf

In it you will find answers for all your questions. How to deal with the blue (keeping the same film), how to past them in PS and so on. 750 pages...

J

[brennig]

here is a direct link to the pdf : http://trichromie.free.fr/aipc/public/SupportcoursPdF/1516-Trichromie.pdf

In it you will find answers for all your questions. How to deal with the blue (keeping the same film), how to past them in PS and so on. 750 pages...

J

Thanks for all the answers people!

That link still doesn't work for me, perhaps I should email Henri and have it sent as an attachment.

I have all the filters ready to go! Making some modifications to my camera, in particular to the back so that changing the film doesn't move the camera as much. Also made my own primitive filter holder (out of a clothes hanger) that won't shake the lens when changing filters. I'm going to test everything out on my Rolleicord this weekend, getting the exposure times, scanning, photoshopping in order before moving onto 4x5.

As I'm printing these photos I will need a Key (monochrome) layer, I was initially planing on getting this from combining CMY layers, but then that layer will include the aberrations. Instead I'm planning on shooting one shot without a filter, then two with red and green filters. I will create the Blue (and thus Y) by subtracting R and G from the unfiltered image. That should also help with exposure times. My R and G filters are resin (Formatt Hitech), but my B is an old Kodak gel, therefore this method might be more consistent too.

[ventdesable]

Hello,

Here is a link to Dropbox where you'll get the PDF :
https://www.dropbox.com/s/vbk6no2mm5qgumj/1516-Trichromie.pdf?dl=0

Have Fun

J