Let's see what spectral peaks we'd like to be able to switch on in our most desirable enlarger light source. If you could switch on any in particular which ones would they be ? Of course, if all switches were 'on' we'd like a white light with a flat response curve.
Perhaps an indication of which ones are the most important would be another thing to think about ...
Ideally matching those of separation filters - something like 630nm, 520nm & 440nm for starters (ie #29, #61 & #47 filters) - or if you need the ultimate set, then use #70, #98, #99 (650nm, 545nm & 430nm), which are reckoned by Kodak to be more accurate. Mixed correctly, should deliver white light.Originally Posted by swmcl
A perfect retrofit for the illumination would be a LED with continuous spectrum of the CRI 98 kind, of about 3200K to 3400K.
Problem happens when we want to use RGB LEDs for a VC or a Color head, substituting filtration by electronic control: VC in BW simply requires a calibration, while IMHO color printing it's more complicated, my guess is that 5 peaks would be necessary, instead 3, to adjust channel crosstalk.
As much as anything, this thread serves a purpose to outline the important frequencies.
Interneg, your 6 plus another for white gives 7 ! That would be difficult to manufacture for sure.
Maybe the best solution is to just have a really high power white source and use filters manually ?
I assume for B/W or color too???
B/W is easier as G/B are the sensitivity of MG paper... Something to note is to remember that most filter systems subtract other colors, but also reduce the main color... Previous systems were calibrated based on this loss, but if you ever exposed paper to pure G or B light, they tend to go bonkers as they build density quickly, so exposure can be very short and hard to control...
Using white LED'S can be difficult as the spectra has much more of the blue end of the spectrum, so materials can go bonkers again... And they don't match the 3200K of the old color head bulb...
Color heads would be another animal in terms of calibration, but I'm sure possible with a lot of difficult calibration, but I'll stick with a halogen and calibrated filter stack for now to get some work done, without my head exploding... ;-)
Steve K
No, read it again - it's a choice of one set of 3. The 70, 98, 99 gel filters have much lower transmission than the 29, 47, 61 but with much narrower spectral peaks - the power of LEDs would potentially negate this issue. You will need a good control system to deal with balancing it all out to get even outputs across the different colours (realistically, probably MOSFET or similar system) but when you mix RGB together correctly, you will get white light, that's the point of additive systems... And it should do 00-5 with few problems, essentially a question of mixing G&B durations in specific amounts in the same exposure period via high speed switching.
Getting this to work well is why Heiland's light systems cost so much.
Agree with above; using white LEDs is bad. Thus picking filters is unnecessary. One needs to choose from the available LED colors.
Royal Blue
Green
These two target the paper's sensitivity fine.
No need for RED for black and white. Focus with green.
Speaking as someone who made similar equipment, Heiland systems cost what they do because they must function in all markets (US, EU, etc), require quality parts, lots of engineering, and the market is small and fickle (on a good day).
Kodak also reckons that those particular filters (70, 98,99) as Wratten gels aren't partickl'r resistant to fading either. And I've already seen the kinds of problems that come with extremely narrow peaks with dichroic filters in color printing. I routinely do additive color printing, so have an pretty good idea of what is practical and what isn't. And if it were me, I'd simply wait awhile until RGB LED's significantly improve in spectral characteristics. CRI using mixed units to obtain a relatively balanced white light doesn't begin to tell the whole story, because in additive color printing you have to dissect that into three, and exactly three, precise components of equal energy.
So Pere,
You would say that a design needs to be say ROYGBV or some such to get a real balance overall and an even spectrum for the white ? (Yes I included another colour to make 6) We photographers would like the RGB part to be a specific set of frequencies and to be able to turn each of the RGB on separately leaving the rest (OYV in this example) to be switched on separately. This would mean 4 different switches and power supplies and perhaps 5 or 6 different LEDs in the mix.
The best example would be if it were manufactured on a square COB design using the latest nanometer fabrication technology for higher efficiency, no ?
I'm assuming the number of differently specified LEDs would depend on the spectral output of the important RGB LEDs. If they were really narrow band then there would be a need to have more LEDs that are broadband set in between to compensate. Either that or make up the white from a completely different set of LEDs.
For BW it's easy, it's about calibration...
For color, beyond rgb, perhaps adding some Orange and Cyan it would allow to have a better control.
This would be to fill r-g and g-r valleys of the LEDs, emulating original tungsten spectrum, just my guess...
Posting Permissions
Reply With Quote
Bookmarks