Lighting choices for Durst L138

[interneg]

I'm a bit skeptical about getting clean color prints from any kind of current RGB LED system.

As long as they hit the spectral transmission peaks of #70, #98 & #99 filters with good hard cut-offs (something LED's are very, very good at), you should get very accurate colour if the rest of the system is sensibly designed. That said, it might also show up deficiencies elsewhere in the imaging chain - for example, one of things that's popping up as a result of the emergence of laser projection of digital cinema is that it's super demanding of good colour calibration throughout the grading process - if you're off by 2nm from where you should be, it'll show, badly.

I also suspect that 475nm LEDs might not be a bad place to start if looking for a BW only system for Multigrade filters.

Anyway today RA-4 has a drawback. Papers are very saturated, ...while portraiture papers for optical printing are a bit hidden, at least to me.

Actually, they're pretty controllable - have you seen any recent top-notch optical RA-4 prints? Contrast & saturation can be controlled by flashing, masking etc very well indeed. Always easier to reduce contrast & saturation than add it.

[Drew Wiley]

Pere - RA4 contrast and hue punch can be tweaked slightly via specific paper choice, and more significantly by relatively simple masking techniques. ... And swmcl - LED light panels claiming to have CRI 95 have gotten pretty good, but are probably not truly in that range of quality; and even if they were, that means that five out of every hundred representative hues is going to be a mismatch. No toy video light like that Alibaba thing is going to do a decent job color printing. And the specs are unquestionably misleading because they are based on the hypothetical rated color temperature of certain LED's without factoring in the offset of diffusion plastics etc. Same can be said for camera store light boxes. It's not just about lumens or even averaged out color temperature. You need the ability to get fairly precise control on specific color dye spikes, which modern colorheads are designed to do, generally using halogen bulbs and subtractive CMY dichroic filters. Yes, you have a fighting chance. Every boxer who enters the ring and wakes up in the middle of the first round with a busted lip, black eye, and dizzy head learns something useful. ... Now for Interneg - Do you have any idea of how very little light actually gets through 70, 98, and 99 filters? - or how very few were ever used for actual color printing? That's not how it's done. For my additive colorheads I use a trimmer
system, which achieves narrow-band cutoff not via absurd density, but by providing a dichroic sandwich the light can pass between, trimming off what lies on either side. That's not feasible with banks of tiny little LED's. I'm not discounting the eventual possibilities. LED technology is still a bit adolescent for precise color work. VC black and white printing is more realistic. I'm now using LED panels for copystand work, including color.

[swmcl]

Thanks Drew,

I've always appreciated your thoughts.

I'm not going to be colour printing anytime soon. A 'fighting chance' is worth the cost of the journey. I appreciate that if manufacturers of LEDs were to listen to the concerns and wishes of the colour and B&W film photographic community they could make products that would very much make our lives better. We are a very small market segment and so this is not going to happen. What this panel is trying to do is approach the Arri panels in performance - we all know the cost of Arri panels at B&H. The Chinese want a slice of that market (I've seen adverts boasting 'Arri-like performance' but they spelt Arri wrong !!). My hope is that an offshoot of this direction is a suitable-enough light source that can do a reasonable and cost-effective job in B&W printing. It is silly to think we could keep finding bulbs and other parts for enlargers that are already 50 years old. I hope this kind of thing might get us out of a bind. It won't be perfect but it could do a 90 - 95 % job (to the photographer) and if the end viewer (common man) can't tell the difference, it is effectively a 100% job.

[Drew Wiley]

You're not risking much money, and if it doesn't work out, the light panel will still be good for something else. And after a few good brawls in the ring, the technology will have improved, and you'll have a much better idea how to make it win. Arri is pro stuff. I like their focussing fresnel lighting. Have fun with your project! Durst enlargers are wonderful once you've got them all tuned up.

[interneg]

Do you have any idea of how very little light actually gets through 70, 98, and 99 filters? - or how very few were ever used for actual color printing?

Yes, I am very much aware of both the low transmission & limited circulation outside of R&D labs, however it is the gold-standard of colour separation - which is why I pointed out that LED technology can deliver the same narrow bandpasses with much more light output relative to power.

That's not how it's done. For my additive colorheads I use a trimmer
system, which achieves narrow-band cutoff not via absurd density, but by providing a dichroic sandwich the light can pass between, trimming off what lies on either side. That's not feasible with banks of tiny little LED's. I'm not discounting the eventual possibilities. LED technology is still a bit adolescent for precise color work.

Again, I'm aware of these systems - are you adding dichroics with cut-offs at 390 & 520nm on blue & green respectively? Are you using MOSFETs to control the bulbs?

I think the biggest problem with additive LED heads has often been poor system design, rather than the competence of the components when appropriately chosen - we're probably rapidly approaching the point where it's likely to be at best a dead heat between the technologies.

[Drew Wiley]

Each of the three colors involves a PAIR of dichroic filters, cutting each off both directions, rather than a narrow single filter. Too narrow an nm width is counterproductive. The spectral sensitivity of dichroic filters shifts if they overheat, not to mention the complication of reciprocity failure getting longer and longer. I designed all this relative to make big prints from heavily-masked Ciba printing. Can't get into too much detail here, except that it's simultaneous additive, feedback-based upon pulsed halogens, which means a lot of headache triacs susceptible to cross-talk interference issues. The whole machine is exceptionally heavy, about 14 ft tall, and built like a tank (earthquake country!). The color output and panel control is superb. I designed it to run dramatically cooler than the Durst commercial additive option, which never made it to public market but did get employed by the NSA for aerial recon photos - those colorheads run extremely hot and need the filters replaced every six months ($$$). Since then sine-wave lighting control has become an option, but one dependent upon software, which is always a risk for becoming prematurely obsolete, so I bypassed it. I'd have the same worry about current LED's - the mere fact the technology is evolving so fast means that specific replacement components might become unavailable quickly too, while a drawer full of halogen bulbs might keep one going another fifty years. I built my colorheads using special aerospace gaskets and sealants, so they should last quite a few more years than me! But I also have a Durst L184 with a subtractive colorhead set up if the bigger fancier machine happens to throw a triac tantrum. My biggest problem at the moment is just finding the time to print color. I've been heavily involved in black and white printing lately.

[Pere Casals]

Actually, they're pretty controllable - have you seen any recent top-notch optical RA-4 prints? Contrast & saturation can be controlled by flashing, masking etc very well indeed. Always easier to reduce contrast & saturation than add it.

My ignorance is deep, I've never printed RA-4, but I think that I will...

I guess a way would be preflashig with a mask. A general pre-flash would lower a lot saturation in the highlights, but with little effect in the shadows, so the desaturation would not be uniform...

Pere - RA4 contrast and hue punch can be tweaked slightly via specific paper choice, and more significantly by relatively simple masking techniques. ... And swmcl - LED light panels claiming to have CRI 95 have gotten pretty good, but are probably not truly in that range of quality; and even if they were, that means that five out of every hundred representative hues is going to be a mismatch. No toy video light like that Alibaba thing is going to do a decent job color printing. And the specs are unquestionably misleading because they are based on the hypothetical rated color temperature of certain LED's without factoring in the offset of diffusion plastics etc. Same can be said for camera store light boxes. It's not just about lumens or even averaged out color temperature. You need the ability to get fairly precise control on specific color dye spikes, which modern colorheads are designed to do, generally using halogen bulbs and subtractive CMY dichroic filters.

Drew, I see two possibilities to retrofit a color head, one is using a white LED sporting CRI 98, this would match a tugnsten, but as Larry noted we shoud the see if we have an after-glow because of the phosphor.

The other way is an R-G-B LED, my guess is that by adjusting the spectrum with also Cyan and a Orange LEDs this would deliver enough control, just a guess, it should be tested in practice...

[Drew Wiley]

I'm sure LED will be the solution of the future, Pere, though use of enlargers for color printing will no doubt continue to diminish. But LED's just aren't really there yet, when it comes to taking a step to seriously improve color printing per se. Let me briefly state why. Assuming those who do continue to gravitate toward optical printing will do so to achieve a level of print quality difficult otherwise, that should take into account a strong enough light source to punch through supplementary mask density as well as the original neg. And even though RA4 papers themselves print fairly fast, achieving precise composition in big enlargements as seen through a dense orange dye mask, on color neg film, itself requires quite a bit of extra light. For example, I compose with adjustable masking blades on a big precision easel, so my white borders will act of the final frame. Even with a huge fairly high-aperture lens (a 360/5.6 El Nikkor) and 1700 watts of halogen, it's a bit of a headache when punching large prints. But don't be fooled by advertised CRI ratings. It's not that simple, nor can you always trust marketing hype. The color accuracy has to correspond to three specific wavelengths, and do so narrowly, in order to be ideal. But when that day arrives, separately controllable RGB would be preferable.

[Pere Casals]

Even with a huge fairly high-aperture lens (a 360/5.6 El Nikkor) and 1700 watts of halogen, it's a bit of a headache when punching large prints.

The equivalent LED power would be 170W...

But we have another source for the efficiency: Condenser !

Color condenser heads were also there: http://www.khbphotografix.com/omega/...er/D2Type2.htm

Condensers for 8x10 are king size, but now we have new resources that were not as available in the past.

Let me guess a 8x10 color solution:

> Condenser type.

> Condensers made from solar energy frenels

> R-G-B LEDs with also Orange and Cyan illuminators to control the channel crosstalk, my guess is that this would solve any color need.

> Perhaps some (more or less) frosted glass in the path to lower collimation.

With only 200W of LEDs a condenser (or "semi" condenser) configuration would deliver an impressive yield, even for mural prints.

[Pere Casals]

How many lumens of white light is required at the baseboard ?

In the baseboard you need Lux, not Lumens, Lumens are for the emiting source.

Basicly you may substitute the specified tungsten light source by a LED source sporting 10% of the power, because LEDs deliver a much higher yield.

As LEDs do reheat less you may overpower the head, to some x3 the equivalent power. In such a case you may want to have an electronic dimmer to decrease power for small prints, if not you will have to stop too much the lens to have long enough exposures.

I've a 18W LED bulb in the (condenser) 138S, this ends in some 8 Lux on the easel with lens wide open at f/5.6 for 40cm prints. This ends in around 20 seconds exposures at f/11, which is a suitable aperture for LF enlarging lenses.

A diffuser head requieres more power, but your starting point it is 10% to 15% of the original tungsten power. Use warm color bulbs similar to tungsten in Kelvin, or RGB bulbs for direct split grade.