LED RGB Additive Enlarger Head - Building One?

[L Gebhardt]

I've built a variable contrast LED head for my Durst 138S. It uses green and royal blue LEDs for contrast control. It dawned on me that I'm 66% of the way to an additive color head; all I need is a red channel. This would let me decommission the color enlarger and reclaim some space in the darkroom. But I don't know much about additive enlarging. So I have a few questions:

1) Is this even feasible? I know the Chromira uses LEDs to exposure paper, and the Lightjets use RGB lasers to expose papers so it seems like it should work, but neither go through a negative first.

2) What would the ideal frequency ranges be for the three colors? How much tolerance is there in being off. For example I can get a couple of different frequencies in blue and red LEDs, but very little choice in the green. It seems they would need to be matched to the color absorption parameters of the paper.

Thanks,
Larry

 

[DREW WILEY]

Larry, I'm extremely skeptical that the technology is quite there yet. But I'd go ahead and do some tentative thinking about it, since this kind
of lighting is evolving rapidly. Don't believe everything you read from marketing people. Getting a good spectrum is impossible at this point
in time. I've tried to keep myself updated frequently, and given the sources of supply and the shady reputation of certain distribution channels, take things with a grain of salt. More legit advances are being made in true architectural lighting at dramatically higher price levels. Some of this
has to be computer controlled, which means potentially obsolete software as well as hardware if something goes wrong or bankrupt. The hypothetical advantage would be low heat. The downside would be trying to get complicated circuits to get along with one another. I know how and currently use two additive halogen enlargers of my own design (employing all kinds of cannibalized components too). But if the
spectral characteristics are proper, and you can design some way to control the channels and the diffusion, the color quality will be at least
equal to what you can get with a laser printer or Chromira at a fraction of the operating expense. I will admit that ZBE helped me out with
some of the special devices that helped me psychoanalyze my own printers, similar to what the NSA uses in their additive 9x9 enlargers.
But they're out of business. The electronics side of all this can get pretty complicated unless you're an electrical engineer yourself. The way
Durst themselves did it (yet never marketed the results) resulted in a very very hot colorhead, so you'd end up having to replace the filters
every six months of so, at terrible expense. But I know who will sell you the guts of the system, if you don't mind dropping tens of thousands
of bucks. That's why I started from scratch and built a relatively cool system which is far more economical to operate. Old school rheostatic
controls or dimmers won't allow you to have enough muscle for serious color printing. I won't go into xenon flashtube systems. The Minolta one was weak, to put it mildly. And a few old Durst sequential additive systems might still exist in storage somewhere from back in pre-scan
graphics days - but they will almost melt your original negs or trans, ideally need water cooling, and won't allow you to dodge and burn...
but like I say, keep an open mind, because the picture might change pretty soon. This control aspect will remain tricky unless you go to
computer control on the cheap.... I'd look for some used Mac setup from a rock concert lighting crew, that can handle at least twenty circuits simultaneously, plus some backup components. Otherwise, that alone will hit you around six grand new.

 

[Bob Carnie]

Matching the paper sensitivity is the key and is way beyond my scope.
I do know that Agfa Classic works in a Lambda but Ilford Warmtone does not..

I believe the Adox Classic paper would work as well.

Very fine line you are reaching for, way beyond my knowledge level , good luck would be interested on how you do.

But you are using the LED for exposing through a negative right? You are not contemplating designing a system to work form digital files?

 

[L Gebhardt]

Drew, thanks for the input. I think I have control aspect down well. With the two color LEDs I'm using in the variable contrast head I control them with an Arduino (small programmable microprocessor). Combined with a constant current power supply that dims with PWM (pulse width modulation) I can control the brightness of the LEDs from off to full power fairly precisely. One concern is it's only 8bit control, so only 255 possible between on and off. That may not be precise enough, but with a newer Arduino I could get 12bit precision on two of the channels.

From what I can determine the CREE LEDs are produced to a high standard and closely match the documentation. They are also available "binned" where they are tested for color. So I could be assured of getting ones with the characteristics that I request. Still they have a distribution of frequencies around a strong peak. Between the blue and green absorption curves for both Endura and Crystal archive the green layer could still be sensitive to a small bit of light from the blue LED, but less so to the Royal Blue (but the blue is also less sensitive to the Royal Blue). I think the Red end is OK. I can't get LEDs that match the absorption peak of the paper perfectly, but they won't cross expose the green.

The diffusion aspect is also working well. Currently the individual color LEDs are mounted a bit farther apart than I would like, but they still produce an even light given a few inches of space between the LEDs and the diffusion screen. I have since found I can get three LEDs mounted on a single "star". This would allow me more freedom to position them for evenness, and would get all three LEDs as close as practically possible. I think this might be required for color, whereas in black and white you wouldn't notice the slight contrast unevenness (if it's there - I can't see or measure it).

So how do you print with an additive head? I'm used to a subtractive head where if I want more red I need to add cyan filtration. I assume with an additive system I would need to subtract red?

 

[DREW WILEY]

You can study the spectral characteristics of various paper and film profiles. Each of the three color of light need to pretty much hit the nail on
the head in terms of peak sensitivity. Otherwise you need some distinct overlap on each side, which can be further restricted with dichroic
"trimmer" filtration. This get pretty involved since it is operating temperature affected. Remember, you're aiming for a much high color quality
than ordinary subtractive colorheads, so the rules get a lot stricter. It also takes much more light intensity when exposing additive versus
subtractive. If supplementary filters are involved, you need to mechanically incorporate them at the correct angles of incidence. You can't
just partially overlap them like in subtractive work. So things get real complicated real fast.... or you could just wait things out awhile and
see if someone does come up with appropriate RGB narrow-band LED bulbs. The parameters need to be far stricter than what will work with
VC paper if you want quality results. It might require a very different head design too. But don't let me discourage you! I could be a fun
and worthwhile project if you're patient with it.

 

[L Gebhardt]

Matching the paper sensitivity is the key and is way beyond my scope.
I do know that Agfa Classic works in a Lambda but Ilford Warmtone does not..

I believe the Adox Classic paper would work as well.

Very fine line you are reaching for, way beyond my knowledge level , good luck would be interested on how you do.

But you are using the LED for exposing through a negative right? You are not contemplating designing a system to work form digital files?

That's interesting that you have issues with different papers in the Lambda. I am surprised you have issues since I thought all papers were approximately the same in there spectral response (but not their sensitivity). I've tested the VC head on Ilford MG IV, Adorama RC (no idea who makes it), and Forte Polywarmtone FB. In all cases it worked as it should. I'll need to test the Ilford Warmtone.

Yes, I'm looking to expose through a negative. Currently I use a ZBE Starlight (I must have the only good pair of them in existence from what I've read). But this can only do 4x5, and it would be cool to be able to enlarge 5x7 color. I would also like to keep only one enlarger with the ability to quickly switch heads.

 

[DREW WILEY]

We posted at the same time, so here's some more tidbit. The color theory is basic, so in printing color neg material like Crystal Archive, your
primary additive sources affect the direct complement of CMY dyes in the paper. I have the added luxury of controls programmed to CMY,
even though the actual lights are RGB. There are ways to cannibalize feedback controls from other pro enlargers like Durst to make this simpler, but you could certainly print without either feedback loops or subtractive math. The less the number of light sources the better. Once you get into more than three or so, EMI becomes a very serious headache, and can potentially wreak havoc on anything solid state within the vicinity. So one more super top-secret trick to getting even diffusion: something called a linear-array fresnel (not to be confused with fresnel lenses in the conventional sense). These can be used in tandem with conventional ground diffusers, and placed in mirror boxes or whatever - but oddly, I think I'm the first person who every used these in enlargers - and what a difference!

 

[Prof_Pixel]

Keep in mind that printers like the Lambda are designed to print from RGB digital files and they don't have to worry about film dye curves; just the paper spectral sensitivity curves. Your task is to thread the needle of both the color negative dye curves and the color paper spectral sensitivity curves. Good luck.

 

[L Gebhardt]

So I guess the question is what's the ideal frequency (and how wide can the band be) for each channel that works well with the film dyes and the paper absorption? If I can solve that I think I can build an even head. I can also probably use the same Arduino based controller to drive it, but worst case I control the brightness of each channel with a very fine potentiometer (manual dial setup).

Also, assuming I get this working perfectly, is there an advantage to an additive head vs a subtractive one?

 

[Bob Carnie]

Larry do you still make negs with your LVT recorder?

So I guess the question is what's the ideal frequency (and how wide can the band be) for each channel that works well with the film dyes and the paper absorption? If I can solve that I think I can build an even head. I can also probably use the same Arduino based controller to drive it, but worst case I control the brightness of each channel with a very fine potentiometer (manual dial setup).

Also, assuming I get this working perfectly, is there an advantage to an additive head vs a subtractive one?

 

[EdSawyer]

Larry, sounds like an interesting project. Theoretically an additive head should be better than subtractive, less color crossover, etc. Despite Drew's not liking it, I use the Minolta Beseler 45A head and find it quite good. I haven't used it for larger than 16x20 prints yet but eventually may. It's weakest point is that flash tubes for it are somewhat hard to find, and the output is not huge, so doing big prints takes quite a series of exposures of the flashes to expose the paper. Other than that, it's great. Built-in color analyzer, sophisticated controls (8-bit resolution which seems like plenty.) They are cheap on Ebay, usually $100-ish. I have at least 5 of them, mostly since they are cheap and it's nice to have spares. Tubes, though not all that available, seem to last quite a while, I have done hundreds of 8x10s and 11x14s and have yet to need to change any of the tubes. I am in NH too, btw.

-Ed

 

[DREW WILEY]

The problem is identical whether using a digital printer or an additive head. The logistical difference is that with the expensive laser units they
have done some of the homework for you and can supply you with appropriate starters setting for your chosen film/paper combo. In fact, both
Kodak and Fuji give you this tech data up front in spec sheets. Otherwise, calibrating you paper to an additive colorhead is just as easy as doing it with a conventional subtractive head - you might opt to spend a little more time doing this simply because of the final reward of greater hue purity, but still just involves matching a master neg to the paper batch. Real world negs have to be tweaked for esthetic and less than ideal exposure parameters anyway, so one works with test strips regardless. Pretty damn easy.

 

[Photo Engineer]

Use of R/G/B printing with an enlarger is a real pain! There are two major problems.

1. Color correction is difficult to hit dead on with a small quantity of paper. It is a laborious trial and error process. Once you get things spot on, it is still going to take tweaks for the next negative, and the methodology is not intuitive.

2. The enlarger must be very firmly mounted. Any vibration causes misregistration and color fringing. The slightest movement causes problems. So, an enlarger for simultaneous two color printing may not be stable enough for 3 color separations.

PE

 

[DREW WILEY]

I'm not contradicting you Ed, but do wish to point out the limitation when I refer to the Beseler/Minolta flashtube system as weak. You are
making small prints, so that might be perfectly suitable. But once you start going larger, or hypothetically choose to fine-tune printing controls
using silver masking and its added density, that system just doesn't have the necessary muscle. Now add the capability of 8x10 film and you're
into entirely new territory. The same limitation applies once one attempts rheostatic control or dimmers. Those flashtubes are not exactly
narrow-band either. But as one does get nitpicky in that respect, things can get more complicated. Filters which are too narrow will shift
their sensitivity with both temperature and incidence, and potentially miss the peak, or pass too little there and end up with prolonged
exposure and recip issues. And I don't know enough of the real facts about LED's to speak to the issue. If I thought a particular set of bulbs
would work, I'd sample them and run a true spectrogram myself. Reliability would be another issue. I might ask a buddy of mine who specializes in this kind of lighting. But having to hit a home run on the first pitch, without a prototype, can be a bit spooky. I built a 5x7
system first and got accustomed to it, before building the big 8x10 unit. I get an occasional foible with the electronics of the 8x10, but
can hardly believe what it does with the color negs. Hope to print another 30x40 tonite.

 

[DREW WILEY]

Ron - I'm speaking of simultaneous additive, not successive. In this case the headaches are up front in the initial engineering. Daily operation is just as easy and fast as any subtractive colorhead.

 

[Bob Carnie]

I used a Nord Additive printer while attending school , and since have only used subtractive on enlargers.. I found no difference in usage but that was a long time ago for the Nord

I use a RGB laser unit now but I do all my corrections subtractive as I would on a durst enlarger.

My unit has endpoint aim points supplied by Fuji and Kodak as well Ilford for the BW paper.
The unit calibrates a 21 step wedge to the manufactures steps and until its calibrated I never work... Worst case I have seen is three step tablets of the onboard densitiometer and calibration unit to get me working.

I also have used a big ass HK unit for rear projection and hot lights on front for flat art to mix and match on transparency film.

I think the key is to have a good grasp of photographic colour theory and one then can get use to any type of unit.

If Larry is using his LVT , he can adjust contrast and tweak colours even before he gets on his LED unit.

 

[Photo Engineer]

Ron - I'm speaking of simultaneous additive, not successive. In this case the headaches are up front in the initial engineering. Daily operation is just as easy and fast as any subtractive colorhead.

Drew;

Even so, at the very start of your work, getting the correct color balance is going to be a pain and each change will be a pain in some other part of your anatomy! And sequential R/G/B exposures give the best color rendition. Simultaneous exposures will probably do the same if the LEDs are sharp cutting enough.

CC correction is best for rapid and accurate color balance changes.

PE

 

[L Gebhardt]

Larry do you still make negs with your LVT recorder?

Not in a while. I was having issues with the evenness. It could be the laser was going, or simply dust (cat hair is my suspicion) in the optical path. So now it's just taking up space. I really should take it apart and see if I can get it working well again.

 

[Bob Carnie]

Salgado is using this for all his digital files to go on a enlarger..
I think you could be making some income from this device if you could get out the cat bugs.

Seems to be a bit of a renaissance for this type of technology.

Not in a while. I was having issues with the evenness. It could be the laser was going, or simply dust (cat hair is my suspicion) in the optical path. So now it's just taking up space. I really should take it apart and see if I can get it working well again.

 

[L Gebhardt]

Use of R/G/B printing with an enlarger is a real pain! There are two major problems.

1. Color correction is difficult to hit dead on with a small quantity of paper. It is a laborious trial and error process. Once you get things spot on, it is still going to take tweaks for the next negative, and the methodology is not intuitive.

That's been my experience with the subtractive head until I got the ZBE unit. Once you dial in a paper you can get a good starting negative easily. The fact that it uses a sensor on the baseboard to calibrate the intensity helps a lot. I'm having trouble wrapping my head around what changes would be needed for color changes with an additive head, so it certainly isn't intuitive to me at this point.

2. The enlarger must be very firmly mounted. Any vibration causes misregistration and color fringing. The slightest movement causes problems. So, an enlarger for simultaneous two color printing may not be stable enough for 3 color separations.

PE

I hope to do all this in one pass, so no issues there. However I haven't seen any issues with registration when using spit grade printing with the VC head either (I don't touch the enlarger between exposures, which I think is key).

 

[L Gebhardt]

Salgado is using this for all his digital files to go on a enlarger..
I think you could be making some income from this device if you could get out the cat bugs.

Seems to be a bit of a renaissance for this type of technology.

Interesting. I've been assuming that interest in film recorders was almost dead. It's on my short list to see if this is fixable. If not, I can hopefully sell it off for parts and reclaim some floor space. It I get it working I guess it could be a side business (not that I'm not short on time already).

 

[Photo Engineer]

Vibration may or may not be a problem. I have none with my enlarger. However, if you have to make any changes to settings on the head, the touch of a hand does cause problems in some cases.

PE

 

[L Gebhardt]

Vibration may or may not be a problem. I have none with my enlarger. However, if you have to make any changes to settings on the head, the touch of a hand does cause problems in some cases.

PE

I had seen that with my old Omega system when I tried using the Chromega head for split grade control. That's one reason my current system has all the adjustments through a controller box that sits next to the enlarger. The Durst is very stable, but I still try not to touch it during an exposure.

 

[L Gebhardt]

Larry, sounds like an interesting project. Theoretically an additive head should be better than subtractive, less color crossover, etc. Despite Drew's not liking it, I use the Minolta Beseler 45A head and find it quite good. I haven't used it for larger than 16x20 prints yet but eventually may. It's weakest point is that flash tubes for it are somewhat hard to find, and the output is not huge, so doing big prints takes quite a series of exposures of the flashes to expose the paper. Other than that, it's great. Built-in color analyzer, sophisticated controls (8-bit resolution which seems like plenty.) They are cheap on Ebay, usually $100-ish. I have at least 5 of them, mostly since they are cheap and it's nice to have spares. Tubes, though not all that available, seem to last quite a while, I have done hundreds of 8x10s and 11x14s and have yet to need to change any of the tubes. I am in NH too, btw.

-Ed

Ed, where in NH are you located? I've heard mixed reviews of pulsed heads. It sounds like it would give me a head ache in short order, but maybe they don't behave like strobes. So how easy do you find color printing with this system? Does it take any mental contortions (more than normal) to dial in the color?

 

[Chan Tran]

It's something I wanted to do in a long time. Good luck to you Larry! I am thinking although there is pro and con I would want to do a variable timing for the RGB instead of trying to dim the LED. You would need 3 separate timers but that's easy thing to do.