Additive color printing brainstorming

[BetterSense]

This is a conceptual exercise, but I want input from experienced color printers.

Suppose that instead of building a color pack with CMY filters, or using a dichroic head, I want to print with 3 separate exposures through R G B filters. Suppose I make an automatic shutter mechanism to make the 3 exposure sequence through the different filters for me. Since I am good at electronics, I make myself an enlarger timer to drive my color-exposure-sequencer. I can make the electronics do whatever I want. Now it's a question of what would be the best and most convenient for the printer:

The scheme I am imagining, I would conceptualize the R G B exposures in terms of their relative ratio to one another, which I would use for color control. I would have my timer store this ratio of the 3 R G B exposures to each other, and then I could set the total exposure time be whatever I wanted. When I went to print, I would enter my desired exposure time...say, 10 seconds...and if I had set a color balance of 20%/50%/30%, I would automatically get 2s of R, 5s of green, and 3s of blue. If I changed the color balance, the total time would stay the same. Do you think this would be a smart system or would it be better to have the ability to individually increase the exposure time of one color while leaving the other colors' exposure times the same? With the scheme outlined above, increasing the amount of one color would automatically reduce the other colors' exposures to keep the total exposure time the same. Is that what you want when color printing?

If I wanted to do something like this, what filters would I use (what specific Lee filter gels or kodak/wratten numbers).

 

[holmburgers]

I like your question! This thread might have some hints... (including the Wratten #'s)

(there was a url link here which no longer exists)

 

[BetterSense]

Where can you buy these filters? I can't find them at Freestyle. I have a Lee/Roscoe gel swatchbook; do you think there is equivalent RGB filters in there I could use?

 

[Mr Bill]

With the scheme outlined above, increasing the amount of one color would automatically reduce the other colors' exposures to keep the total exposure time the same. Is that what you want when color printing?

Personally, I'd rather operate with the more traditional method, where the operator chooses plus or minus adjustments to CMY plus a "density" adjustment, all going plus or minus roughly in equivalent cc filter values. If you've never learned to work this way, though, the auto expos system might be better, I don't know. But since you'll be programming this into a system, I presume, why not just keep both options? It seems it would be a relatively minor programming change, given that you WILL need some exposure adjustment already. Certainly it will be easier than building in "slope control" adjustments, which I think you may also want. At least stay open to the possibility of building these in.

I think that the more difficult parts of the job will be the hardware, setting up the light source(s), filter mounts, etc. I don't know how much you know about this, but here's a little background: The traditional industry method has been to use subtractive systems, where a single white light source is set up. There would be a mechanical shutter in the system, so that the lamp can operate in a low power mode, whilst not exposing the paper. Preparing to start the exposure, the lamp is switched to full power, where it very quickly reaches a stable operating temperature. Next, the shutter opens to begin the printing exposure, with white light. Somewhere in the light path are what they call "filter paddles", equipped with sharp-cutting cyan, magenta, and yellow dichroic filters. The filter paddles are essentially arms, attached to rotary solenoids, which hold one of these filters at the business end. So that the filters can be swing into, or out of, the light path. So the exposing method is: start with white light, essentially red+green+blue combined together. When enough of one color, say 2 seconds of blue light, is achieved, then the yellow filter paddle swings into the light path - this immediately terminates any blue light exposure, while still allowing red and green to come through. As the next color reaches it's aim exposure, say 3 seconds of green light, the magenta paddle enters the light path. Although magenta filtration only blocks green, the yellow filter (already in the beam) continues to block blue. Thus, only red light is coming through. When enough red expousre is reached, a cyan filter swings into the light path. At this point, all light is totally blocked, at least withing the limits of what the filters can do, and the mechanical shutter finally closes.

If you consider how much light energy is being used during a full, color ajdusted exposure, you will probably see that the subtractive system is much, much more efficient at using light, and this is one main reason why it has been so much used in industry. Also, because it has been so commonly used, I'd expect that surplus parts - filter paddles, etc - might be found on the used market. Just something to keep in mind while you do your planning.

 

[Photo Engineer]

Well, I actually recommend the 98, 99 and 70 for purer colors due to the cleaner separation between the filters. However, the method in the OP is what was used in the original Kodak Color Printer series. It worked very well for years.

It is inefficient and takes much longer exposures. It is also much harder to adjust color balance to the untrained eye. A special sensing system did this job in Kodak printers.

PE

 

[holmburgers]

Check on eBay. You'll probably have to get gelatin filters, and if you wait long enough you'll be able to find a large lot for cheap. Buying them individually would be easier but is generally more expensive per filter. I got lucky and found a huge lot of 3"x3" gels with a lot of useful ones in it.

#61 Green is also an option as is #25 Red (though perhaps too broad). I have a Kodak filter handbook which has extensive information about all the filters and if you find a large collection you'll find this book invaluable.

 

[BetterSense]

I think that the more difficult parts of the job will be the hardware, setting up the light source(s), filter mounts, etc.

Na, I already have it designed and half-built. It uses magnets to attach to the enlarger so I can switch it from my 4x5 to my 6x6 enlarger.

start with white light, essentially red+green+blue combined together. When enough of one color, say 2 seconds of blue light, is achieved, then the yellow filter paddle swings into the light path - this immediately terminates any blue light exposure, while still allowing red and green to come through. As the next color reaches it's aim exposure, say 3 seconds of green light, the magenta paddle enters the light path. Although magenta filtration only blocks green, the yellow filter (already in the beam) continues to block blue. Thus, only red light is coming through. When enough red expousre is reached, a cyan filter swings into the light path. At this point, all light is totally blocked, at least withing the limits of what the filters can do, and the mechanical shutter finally closes.

That is pretty smart. It sounds like it would give faster exposures, and it wouldn't be that much harder to build and program. But I heard that modern color paper is very fast anyway, almost inconveniently fast for enlarging; is that true? Plus, I have also heard that using sharp cutting R G B filters is an actual advantage in image quality due to better separation.

It is also much harder to adjust color balance to the untrained eye.

Do you mean using RGB additive printing will be harder to learn that subtractive printing? I don't have any subtractive printing to 'unlearn' if that's an advantage...

 

[anikin]

I'm in the middle of building RGB LED-based enlarger to try exact same idea. I've got all electronics done, and in the
process of writing the firmware for the timer. The cool thing about RGB additive enlarger is that theoretically color
filtration can be found with just one test strip. Fix one color, and run perpendicular test strips for the other two.
Find intersection with the desired color, and you are done. I don't know yet how well this works in practice, and
I can't wait until my enlarger is done. Anyway, I'm very interested to hear about your progress.

Eugene.

 

[Mr Bill]

But I heard that modern color paper is very fast anyway, almost inconveniently fast for enlarging; is that true?

I can't really answer that - much of my experience has been with industrial-level finishing, where the fancy lamphouses built by folks such as Lucht Engineering (long gone, I think) could expose 8x10 paper (from 35mm negs) in, say 400 milliseconds (~ 1/2 second). If high output is your goal, this is arguably not sensitive enough. On the other hand, if you want time to burn and dodge, these times are way too short.

I have a suspicion that your lamphouse will not be too efficient at getting light through your lens, which would greatly extend exposure time. If times are still too short, given that the lamphouse is your own design, you could probably easily modify it to leak away light. In enlargers/printers, It's lots easier to give away light-collecting efficiency than it is to to go the other direction.

Do you mean using RGB additive printing will be harder to learn...

I think that, with computer-based controls, you can separate the two functions - adjustment input vs the actual hardware - so that this would not be an issue. I think it would be useful for your human interface to use the standard nomenclature of cc filters, just so you can talk to other printers, if nothing else. When someone says, for example, that a skin tone needs about 5 more units of yellow, this loosely means that the yellow-light-printing-exposure should be reduced by the equivalent of inserting, into a white-light beam, a filter with optical density of about 0.05, for yellow light. Alternatively, in an additive RGB system, you would want to reduce both the R(ed) and G(reen) light exposures to about 89% of the original (the math is 1/10^0.05 = 0.89). Anyway, in the lingo of the printers, this is simply boosting yellow in the print by 5 units.

 

[holmburgers]

Hahaha, no. It was many moons ago (at least several months).... bitch.

 

[BetterSense]

If times are still too short, given that the lamphouse is your own design, you could probably easily modify it to leak away light.

The lamphouse is not my own design. I haven't modified me enlargers at all, only added the filter-switching mechanism to the optical path. There is no light loss except that caused by the filters.

 

[Photo Engineer]

The additive + subtractive method mentioned above may nor work depending on design. Blue + Yellow = no light at all! Also, the subtractive filters are not matched well to the additive filters.

And yes, learning the additive matrix is not easy for most people. That is a major reason that all printing changed to subtractive printing methods.

PE

 

[Mr Bill]

The additive + subtractive method mentioned above...

PE, if this is in reference to my post, it is just a misunderstanding. I spoke of no such system, I referred separately to them.

 

[holmburgers]

Hmmm. I take it you didn't win?

 

[holmburgers]

That's lame. I had the best luck with a badly titled and described auction. He didn't say "gel" or "wratten" or any of the key search terms that most people are looking for, and none of the filters were described in the auction. I asked him and he told me what he had and I was "like whoa".

 

[Photo Engineer]

PE, if this is in reference to my post, it is just a misunderstanding. I spoke of no such system, I referred separately to them.

It was a reference to placing a yellow filter over a blue filter which will not work properly due to the cutoff differences of the two filters. The other combinations will not work either.

PE

 

[Mr Bill]

Hi PE. To be more clear, I don't know if you were in reference to my post, or not. Since you have just said "placing a yellow filter over a blue filter..." I'm guessing that you are in reference to my post #5. If so, perhaps you misread, possibly this part?

[excerpt from post #5] So the exposing method is: start with white light, essentially red+green+blue combined together. When enough of one color, say 2 seconds of blue light, is achieved, then the yellow filter paddle swings into the light path - this immediately terminates any blue light exposure, while still allowing red and green to come through

I believe that everything I have stated is correct, so if you are disputing anything, please give me the opportunity to defend it. If you were not refering to my post, then I apologize for raising the issue.

 

[Photo Engineer]

Bill;

I was referring to it. And, the method you suggest was usable but not very successful due to the broad cut of the filter compared to the blue light. The filter did not cut off all of the blue light. That is my point. The subtractive filters are broader in bandwidth than the additiive filters and thus "leak". You need a special, high density yellow, magenta and cyan filter set to get it to work so that when superposed, all 3 = BLACK and not gray! Test it yourself and you will see.

PE

 

[Photo Engineer]

I have a copy of Kodak Tech. Pub. E-81N that describes making separation negatives for use on Matrix Film. In the instructions, it recommends adjusting your exposures so that the times are as close to equal as possible across channels. To achieve this while using a dichroic color head it recommends dialing in filtration opposite the color of the separation filter used under the lens to act as a type of ND filter if an ND filter is unavailable and the lens is to be maintained at it's optimum aperture.

Again, this is possible with separation negatives where you are striving to make 3 B&W negatives or transparencies for DT or other reproduction methods. When used with color negative positive systems, this is difficult due to the speeds of the three layers.

If you take Cyan as = 0 speed, then magenta = +0.8 Log E and yellow = + 1.2 log E in speed. You have to balance this somehow. It can be done, but juggling this becomes a feat for someone who can do matrix algebra in their head!

See my previous post.

PE

 

[Photo Engineer]

I was more concerned that there are people out there that might misunderstand you, me and Bill.

That about sum it up.

PE

 

[hrst]

Suppose I make an automatic shutter mechanism to make the 3 exposure sequence through the different filters for me. Since I am good at electronics, I make myself an enlarger timer to drive my color-exposure-sequencer. I can make the electronics do whatever I want.

I would use RGB leds. It would be so much easier mechanically. However, you probably need to analyze the wavelengths a bit. If you buy the most usual, cheapest LEDs, I can say that the red is probably a bit too short and blue is a bit too long, and green is quite optimal. However there are longer reds available, some people use them for plant growing where 625 nm red is not optimum. Check for Ebay, Dealextreme and http://www.satisled.com/ for cheap leds.

This is based on the absorption maximums of color negative film dyes. They are designed to match with paper sensitivities.

But anyhow, I would GUESS that even with the little wavelength mismatch, the RGB leds will work very well, comparable to CMY filtered white light sources. I have yet to make a RGB LED color head so I don't have experience in practice, though.

I would start with 5 to 10 Watt R, G and B (something like this: http://www.satisled.com/30w-watt-rgb-full-color-high-power-led-10w-for-each-color_p288.html ; I haven't bought this so I cannot comment further!), a diffuser very close to them, a small gap (an inch maybe?), a second diffuser (or an 90 deg angle with white surface at 45 degrees, as used in some dichroic enlargers) and then either a condenser lens system or a third diffuser. Then I'd use either PWM or analog circuitry to control the color; or, as the current papers have very low HIRF, you can just use the leds full power and stop them one at a time when the time for that color is full.

With RGB LED system, you actually have CMY knobs that works like adjustable CMY filters! It's just in which way you write the labels to knobs. The red knob at full power is Cyan 0%, and red knob at zero power is Cyan 100%, and so on.

 

[BetterSense]

I would use RGB leds. It would be so much easier mechanically.

I disagree, Coming up with a RGB array that is bright enough to print 4x5 would be fun, but it would he hard to design and expensive in parts and in power electronics. By contrast, I already made my 3-filter servo-drive colorwheel, which I can use with any enlarger.

 

[ic-racer]

Another option (for simple light-source-independent conversion to a color enlarger) would be to get a Janpol.

 

[rjmeyer314]

Another approach is to get a used Beseler-Minolta 45A color head. They print with RGB.

 

[Wade D]

Does anyone remember the old Unicolor RGB printing system from the late 60's? I used it briefly back then but the subtractive system turned out to be much easier for me. The exposures were really long with the additive system. Also keeping everything from moving was a pain as the filters were rotated. Just a blast from the past.