Analog separation negatives from color positive film

Hi all,

Can anyone explain or point me in the right direction (books/internet resources) on how to make analog separation negatives from color positive film? What kind of equipment is needed and how the process is.

Cheers
Peter

Look for books/websites/videos on "pre-press graphic arts" - seperation negatives from slide film (Ektachrome, Kodachrome, etc.) was very common - all publications with color pictures used the technique until the early 2000's. The digital revolution hit the printing industry earlier than it photograhy. Wide spread scanning started in the late 80's. The Xerox star for desktop publishing was in use mid 80's. I've had rumor you can use a scanner at home...

You can do seperations the traditional way by loading up a slide duplicator with B&W film and taking images with color process filters - #25 red, #61 green and #47 blue.

Thanks - I figured Id do this the traditional way. However, as it is for sheet film I figured Id use an enlarger and a contact frame... but how do I figure out the exposure?

Have you checked this?

Thanks man. But its for 8x10 sheet film. I figured Id use T-Max 400 developed in Pyrocat HD for making 3 separation negatives from Ektachrome (I'll not be testing with 8x10 sheets though) for use in carbon printing. The problem is I have no idea how I should expose the separation negatives, or how to evaluate them.

pkr1979 said:

how I should expose the separation negatives

Click to expand...

That's going to be a matter of experimentation, I'm afraid. Test strips/sheets, just like when printing onto paper. Because exposure times will be very short, you will probably need to use an ND filter on your enlarger lens.

pkr1979 said:

how to evaluate them.

Click to expand...

Either through densitometry or by evaluating prints made from the separation negatives. Both will be a lot of work; the first option a little less so. What you're looking for is a straight (as far as possible) HD curve. You'll need to match the CI to your carbon tissue, sensitization and exposure parameters. Of course you could adjust those as well (mainly through pigment concentration, sensitizer, exposure) so to keep things somewhat manageable you'll have to fix some parameters so you're only left with one or two to adjust. Otherwise you'll go mad pretty quickly.

Lacking a film densitometer you might get somewhere using a scanner combined with a calibrated step wedge.

I've been considering a film-based color separation approach lately myself as well, also for carbon, but frankly I'm not sure if it's going to be a realistic plan. My idea is/was to separate color negatives (typically 35mm or 120, but occasionally 4x5) onto a more affordable panchromatic sheet film stock than TMY2 in 4x5" size yielding separation positives, and then enlarge to the desired size to create the actual separation negatives on something cheap & cheerful like single-sided xray film.

I haven't started with this; currently I'm doing it digitally and trying to figure that out. That's already so complex that I wouldn't dream of doing it on film before I figure out a decent workflow with digital negatives. Based on such a digital workflow I think it's slightly more feasible to embark on the film-based approach because you know what kind of issues you're going to run into. It also helps to have nailed other parameters like tissue formulation so that you're not sent off on a goose chase all the time hunting non-film related parameters.

Don't underestimate the complexity of this endeavor. I think there's a good reason why successful color carbon printers are using a digital workflow for the separation part.

If you start work on this, please do keep us updated; I'd be very interested indeed to hear about your experiences!

pkr1979 said:

Thanks man. But its for 8x10 sheet film. I figured Id use T-Max 400 developed in Pyrocat HD for making 3 separation negatives from Ektachrome (I'll not be testing with 8x10 sheets though) for use in carbon printing. The problem is I have no idea how I should expose the separation negatives, or how to evaluate them.

Click to expand...

Start here.

Forget about the staining developer mythos. Your material choices should be determined by the required density range of your separations, which you'll have to determine via test.

You'll need to mask for colour correction/ contrast correction - and if you're making a K sep, you'll need a #8 filter too. Most colour assembly processes have a fair bit of flexibility at assembly (at least those using dyes or inks). A means of accurate and repeatable registration will also help - pin registration being the obvious one, but there are others.

None of this is trivial, but not is it as unsurmountable as people on the internet need you to believe it is.

pkr1979 said:

Hi all,

Can anyone explain or point me in the right direction (books/internet resources) on how to make analog separation negatives from color positive film? What kind of equipment is needed and how the process is.

Cheers
Peter

Click to expand...

Peter
Have you seen this website?
Also, look up "trichotomy" on internet, it will give you filter names and numbers and you
just need to rephotograph your film. One site in French with Mr Gaud is very good, I don't remember the name.
Any panchromatic black white film will be good some may argue one film better than other. I have used 8 different films
they all work, just need to make sure you employ "panchromatic" film, not Ilford Ortho
Do not get buried in details, operation is not too hard.

Thanks fellas! Especially for that file Lachlan. Im using Pyrocat HD as it gives me the curve I want for black and white (never measured myself though), so I figured Id just use that same developing procedure and just adjust exposure.

Nicholas Lindan said:

All publications with color pictures used the technique until the early 2000's.

Click to expand...

Well, colour seperating drumscanners entered the graphic industry in the 60's.
Leaving colour separations by film to those who dids not want,or not yet could afford a drumscanner or those who could, but who had originals not fitting on a drumscanner and wanting to make separations directly, and not via modern, integrated colour-film.

Keep in mind that, at least over here, there were dedicated firms doing reprographic work for printers, and as such could run such scanners and added analogue computers all day.

So it was rather already the 80' that colour seperations by film became a niche.

AgX said:

Well, colour seperating drumscanners entered the graphic industry in the 60's.

Click to expand...

It's rather more complicated than that - and they only really took off after Hell and others realised you could do things to the file between scanner and film output - and that was some time later. Nevertheless, less than a decade before those first prototype scanners started to appear in the early 60s an awful lot of colour sep work was still being done using glass plates for making the seps (dimensionally stable in the era before ESTAR/ polyester base) and then retouching those glass plate seps for contrast / colour correction done largely by eye rather than masking (which was pretty high tech in comparison - along with densitometers!).

pkr1979 said:

Im using Pyrocat HD as it gives me the curve I want for black and white (never measured myself though), so I figured Id just use that same developing procedure and just adjust exposure.

Click to expand...

I think you may need to revise this line of thought, as it seems to suffer from some potentially fatal flaws.
Let me keep it at that you want as straight an HD curve as you can possibly get over the relevant contrast range. Whether or not you can do that with Pyrocat, I leave up to you to decide, but it's not the first candidate that comes to mind. In fact, I'd argue that Pyrocat is pretty far at the bottom of the list of likely candidates. Closet to the top we would find 'boring' stuff like D76 and D96 (see the pdf you were linked to), which unsurprisingly somewhat resemble each other.

Keep in mind that when doing color separations, you're not so much aiming for a 'pleasing' curve like you'd do with B&W. You might, but it won't relieve you from the necessity of closely matching the curves of the 3-4 (minimum) separations. Trying to coax pyrocat into that story will likely work, but would necessitate eradicating the very properties you might like it for (i.e. its tendency to create a pronounced shoulder).

Unless you have a kind of Lomo-look in mind with grossly distorted colors, there's not much artistic about making color separations. In fact, it's a fairly tedious, technically-inclined process that is quite at odds with some of the aesthetic notions that are common in B&W printing and that essentially revolve around all sorts of non-linearities.

Its been my understanding that TMax400 do get a straight curve in Pyrocat HD, and a useful CI for both silver and carbon printing: https://unblinkingeye.com/Articles/PCat/PCat4/pcat4.html

If thats the case shouldn't this work well for separations negatives as well? I dont have to use Pyrocat but that happens to be what Ive got accustomed to, and if it does provide that straight curve it seems easier to keep using that instead of getting used to a new developer... I am however open for tried, tested and working suggestions

pkr1979 said:

Its been my understanding that TMax400 do get a straight curve in Pyrocat HD, and a useful CI for both silver and carbon printing: https://unblinkingeye.com/Articles/PCat/PCat4/pcat4.html

Click to expand...

Depending on development. See e.g. figure 19 on that page. If you get something like that in a color separation, you're in trouble. Figure 24 does look OK. So in response I would ask why choose a developer that apparently is, let's say, temperamental when it comes to curve shape instead of a more straightforward (literally!) option.

Don't get me wrong, I quite like Pyrocat for B&W negative development. But for this application, it wouldn't be my tool of choice. You might want to ask @DREW WILEY which developer he prefers for making color masks. It's a similar task with similar requirements. I'm pretty sure he doesn't use Pyrocat for this, but as said, ask him if you want tried, tested and working.

The generalities are simple, but the exact details all depend on the specific printing process you have in mind, and require a great deal of time and effort to fine tune. And no, not just any panchromatic film works well. I strongly recommend TMX100, by far the most appropriate separation film available today (it was originally engineered for that specific application, among others), unless you have in mind UV printing, where TMY400 or Ilford FP4 would be preferable due to the slight amount of UV blockage of TMax 100 base. You want a traditional long scale developer like HC-110 (or DK-50 on a budget), not a staining pyro in this case. Note that the linked Unblinkingeye resource is NOT about color separation work at all. D76 etc would be a poor choice as well. Making good color separations is fussy work; but all the extra effort you put into it will save you even worse headaches later, when you try to print the result. I simply don't have the extra time anymore, perhaps the biggest necessary ingredient of all.

Old literature might help with respect to general concept, but not with the details, since film selection itself has completely changed. And as far as punch and register alignment gear goes, it depends on how big you want to contact print the separations. I use the same registered micro-pin contact frames and punches as for masking work, suitable for up to 8x10 film. Bigger than that, then you want to use graphics style larger punches and pin systems readily available under Olec-Stoesser or Ternes Burton brands (if seeking used, search the industrial section of EBay, not the photo one). Everything can indeed be done in a properly equipped darkroom, just like the old days; but it takes more patience to assemble the necessary tools and supplies.

AgX, Lachlan, and Koraks have just given some additional good advice. There are dedicated websites to things like color carbon and dye transfer printing. Think in terms of decades of practice, not months. I'd strongly recommend investing in a transmission densitometer if you don't already have one. Large sizes of the appropriate sheet film are getting quite expensive. So expect a Spanish Inquisition experience to your wallet too. Still, if you like a challenge, you will have a fair amount of company.

The two most common separation filters sets are 25 red / 58 green / 47 blue, or the denser and more precise 29 red / 61 green / 47B blue.
But what can really come around like a boomerang and hit you from behind is the fact that most films don't respond with equal gamma to blue, and in fact shift with respect to all of the filters due to unequal long exposure recip failure. One more reason for TMax 100, where it's possible to fine a sweet spot of exposure allowing all three sheets to be developed together for the same amount of time.

But most color processes requires various contrast, highlight, and color correction masks first, when working with color chromes - a complex subject in itself, also dependent on the exact output medium and its specific dyes or pigments. The statistical odds of coming up with something better than a modern inkjet print are low, but when it happens, a handmade print can really stand out from the crowd. Only a few outfits still have the ability, up-front investment, and stockpile of supplies to still do this on a commercial scale, and do it predictably. Just getting fun results is a different story - lots of people are in that category.

Thanks Drew for the correction on the D76.

TMAX100 would indeed be a no-go for carbon transfer, since it's a UV-based process. If memory serves TMAX100 attenuates UV by about 3 stops or thereabouts.

Substituting TMY 400 would give you the same long straight-line response as TMX 100 in an appropriate developer, and the same reasonably high gamma potential. But in its case, the green filter response is out of synch and needs longer dev, rather than blue as in the case of FP4 or now-obsolete Super-XX and Color Sep film of the olden days. A lot depends on your UV exposure source as well; some have plenty of extra muscle, some don't. When I make separations, it's in relation to optically enlarged prints, not UV contact exposure, so it's a different story.
But having to print through a yellow-brown or yellow-green overall pyro stain would seem to be a far greater Slow sign in the road. I can't imagine contending with that in the case of UV tricolor printing.

Then there's the added complication that most pigment sets require an additional "K" black pigment separation.

DREW WILEY said:

But having to print through a yellow-brown or yellow-green overall pyro stain would seem to be a far greater Slow sign in the road

Click to expand...

Not necessarily. The stain from pyrocat is largely image-wise and actually adds significantly to the density the carbon tissue "sees". Unless overly long development times are used, overall added stain density is low and not prohibitive. 510 pyro is different in this respect as it tends to add overall stain much more so than pyrocat.

UV exposure sources today are less problematic than they used to be due to the availability of high-power UV LEDs. It's quite easy and fairly cheap to make an incredibly fast UV light source by the standards of the NuArc/metal halide years. If so desired the wavelength van be tailored to the process, although in my experience the omnipresent and affordable 390-400nm devices work quite well for alt. processes, giving good contrast and speed.

DREW WILEY said:

Then there's the added complication that most pigment sets require an additional "K" black pigment separation

Click to expand...

This is certainly the case, at least in my still limited experience. With the pigments I've used so far, a very dense purple is feasible with just CMY, but a true black requires an additional key layer.

pkr1979 said:

Hi all,

Can anyone explain or point me in the right direction (books/internet resources) on how to make analog separation negatives from color positive film? What kind of equipment is needed and how the process is.

Cheers
Peter

Click to expand...

Peter.
I hope you come back and talk about your results. I moderate a cooking forum and people are able to complicate making a hard boiled egg too.

David Doubley's book on Dye Transfer Printing has a lot of information regarding making separations, masking and materials required and ideas for making home made equipment But as Drew pointed out the references to the films from the 80's will have to be updated. You can find a copy of his book at

Koraks, even a mild pyro stain with its native fbf is adds at least a full stop of density in relation to blue or UV light. It's not strictly proportional. On the other hand, the developers I use in relation to either speed of TMax in relation to lab applications per se, like color separations per se, have the clearest fbf of any films I've ever worked with - as low as .04 in the case of TMY !

Mr Tomato - you make it sound so easy. What's in your secret salad? Have you hybridized four distinct colors of tomato - pure magenta, pure yellow, pure cyan, and pure black. Well, I can understand pure black if you fry green tomatoes long enough, but the others have me baffled. And I strongly suspect holding a red magnifying atop a ripening tomato does not turn it cyan. And this topic is indeed thousands of times more complicated than making a hard boiled egg, at least if making decent prints via matched separations. But let us know when you start offering ketchup in true process colors.

Dye transfer printing is quite a different ballgame than pigment printing, and even the color separation requirements are different. The problem is that it's dependent upon a special kind of matrix film which is very expensive to make any fresh run of due to the high minimum batch cost.
You can't just order it anymore, anywhere. Some philanthropist or institution has to be willing to invest a huge payment up front.
But with most varieties of true pigment printing (inkjet is NOT a true pigment process), you can affordably batch up the needed ingredients yourself, and select pigments from various sources.

The 30 page booklet that came with the Marshall Studios densitometer explains clearly what you have to do. It suggests DK-50 1:1 (because stock would develop so fast that it would be hard to control gamma). It said don’t use a staining developer.

You are going to make time-gamma charts for the film with the different filters. That’s fifteen sheets of film the way I see it. But once you have all that you will be in business.

Drew Wiley knows this stuff. It’s going to take years of experience to be proficient. But the instructions themseves fit in a 30 page booklet. And you can get color prints of beginner craftsmanship quality in about the time it would take to complete a semester photo course at a community college.

tomatojoe said:

people are able to complicate making a hard boiled egg too.

Click to expand...

That's a silly comparison.
If you mean to say, do give it a try and don't be put off by the complexities, then sure. Suggesting those complexities are overrated by comparing making color separations to boiling an egg is just hilariously naive.

DREW WILEY said:

Koraks, even a mild pyro stain with its native fbf is adds at least a full stop of density in relation to blue or UV light.

Click to expand...

This is certainly not the case. With a poorly developed 510 pyro negative it may get out of hand like this, but with a decently processed negative it'll be in the range of 1/3 stop at most.
And the stain certainly is proportional. Perhaps one day I'll set up an experiment to show you although I frankly have more relevant and interesting things to do than demonstrate well-established facts. Such as those color separations!

Thanks again for all help For the figures on Unblinking eye Im supposed to get what you see on fig 13 (UV) and and fig 24 (blue). These are development times for CIs though. As I’ll be using a UV source I’ll use TMY, but Im happy to use DK-50 if that gives me the correct curve with TMY. I can also mix it myself which is convenient: https://www.digitaltruth.com/data/formula.php?FormulaID=137 What is a Balanced Alkali though?

Also Id like a transmission densitometer, and on ebay there seems to be several at different prices… I dont really know what to look for in a densitometer though - any advice?

On my timeline… Im probably spending the next 4-5 five years experimenting with this (will show results if I get anywhere). I got a box of 10 8x10 sheets Ektachrome in the freezer I’ll be using the next couple of years to finish, and its from these positives Id like to make color carbon prints. First of I figured Id be skipping true black to see if I can get anywhere I like without it, is it possible to make a black layer though… I assume its not that straight forward to filter it out… or in?

pkr1979 said:

What is a Balanced Alkali though?

Click to expand...

Kodak Balanced Alkali; a fancy name for sodium metaborate. You can use borax converted to borate using for instance sodium hydroxide. There are formulas for this online if you root around a bit.

pkr1979 said:

First of I figured Id be skipping true black to see if I can get anywhere I like without it, is it possible to make a black layer though

Click to expand...

I don't think that's feasible. The problem is that the pigments you'll end up using are not perfectly complimentary and not perfectly transparent. Both aspects combined results in a black resulting from mixing all three is never a true black. You could try, of course, but I can assure you it's going to be a disappointment. It's not just getting the pigments to mix in the right amounts, i.e. dialing it in - it's just the physical properties of the pigments themselves that precludes a true black to be formed from them. You can't dial out the laws of physics

Sorry, I'm of no help on the densitometer part! For that, it might be best to post a separate thread in the darkroom equipment section; I bet several people will jump on it and offer useful advice.