……..
……..
Last edited by Greg Davis; 30-Jul-2021 at 08:11.
I just watched this last night. Very interesting!
Sent from my SM-G960U using Tapatalk
A few tips: First, putting the diffusion up against the light is usually not very effective. Adding even more diffusion in that way leads to a loss of light, and it still won't be that even. Spacing the diffusion panel away from the LED panel works much better, and determining the optimum spacing takes some trial and error. Second, the light source needs to be quite a bit bigger, preferably 2 inches in each direction than the negative, as their will be falloff at the edges of the light source, and that's true even if one uses reflectors in the box built for the light source. So for 5x7, I'd prefer a 7x9 light source, and so on. My experience comes from making light sources for scanners. Also, edge-lit panels can be more even then the type of panel in the video. I use a Pixel P50 with my scanner.
Last edited by Peter De Smidt; 23-Jul-2019 at 08:54.
“You often feel tired, not because you've done too much, but because you've done too little of what sparks a light in you.”
― Alexander Den Heijer, Nothing You Don't Already Know
Nice to know that a CR96 light works like tungsten light regarding VC filters, cold cathode had strong shifts. Nice experiment.
... but it's way better to use an RGB panel to have a true VC head that does not require filters anymore.
A RGB panel has 4 wires, common and R, G, and B power.
We place a general switch in the common, we switch ON when we arrive, we switch OFF when we leave the darkroom.
R is always connected, we always have red light ON.
Both G and B wires have its own darkroom timer, one timer triggers green and the other triggers blue, that's all.
From that we set green and blue exposure times separately, each grade will have a balance. One may finish earlier than the other if we fire both at same time.
...and we have split grade:
In the general exposure, when we fire blue then we can dodge the shadows. When we fire Green then we may dodge in the highlights
Then we may fire more green to burn highlights, or more blue to burn shadows. A kid's game.
I have the green timer at enlarger's left and the blue timer at enlarger's right, no confusion.
For split grade we need no calibration, we make test strips with blue and with green so we directly have starting times.
If wanting to nail ilford grades then we have to find what B-G balance gives each grade. We may also use voltage regulators to balance B-G for matching grades with same exposure, but I find the 2 timers solution way more convenient, in special for split grade.
RGB panel is the way to go, IMHO. Also cheap.
Very good Greg.
I also experimented with various LED sources for years.
Some here will never accept experimentation.
Keep up the good work and videos.
Tin Can
for fun it is perfect, also it's very interesting to see that with CR96 illumination we match the same grades than with tungsten, there was some debate about that and you results are important to clarify that, but if recommending an illumination retrofit for BW usage, tungsten to LED, then the good recommendation is RGB, IMHO.Originally Posted by Greg Davis
For color prints it's not that clear, I'm not sure if RGB LEDs may have a shift in the mid tones, in that case perhaps a CR96 LED illumination in the color head would a safer choice than RGB, not sure. I'm learnig about that to enlarge MF slides to 8x10" CDU-II...
I sure wouldn't gamble 8x10 CDUII on the highly imperfect spectrum of LED. CRI doesn't begin to tell it all; and I'm very skeptical of some of those advertised CRI ratings, which is stating it mildly. One of the newer CMY subtractive heads would be OK; a narrow-band additive halogen system, even better, which I use; but it's all custom-built. I've printed several 8x10 interpositive-derived images this past week. With CDUII it's best to start with 5000K emitted from the colorhead, then after inspecting initial results, fine tune the settings. High-quality dupes require contrast masking the original, or else careful pre-flashing. Masking is a lot more accurate. For other folks, please note that making duplicate transparencies is a very specialized operation. Suitable films are no longer made; Pere got lucky finding some. There's nothing inexpensive about 8x10 color film in any event. It is, however, still possible to make very high quality internegatives from chrome films using Portra 160 if someone wishes to take the extra effort to mask the originals. Flashing is not quite as useful in that case, though it was routinely used when official interneg films were still being made.
I only have 50 sheets, so not much gambling is possible
In theory CDU-II requires no pre-flashing...
Drew, we discussed this earlier, IMHO with a CRI 80 spectrum can be very imperfect in the tungstem matching, but a CRI 96 has to match very well to deliver that rating, while a cold cathode has 1 full grade shift compared to tungsten but in test made by Greg the grade is nailed, still this is not enough to say that it will be good for color, but it's not bad news.
Those fun tests didn't even remotely address the issue of color printing, much less color duplication work. But don't confuse marketed CRI ratings with the real deal. You have to factor in a significant BS coefficient. Fooling the eye is one thing; fooling film isn't so easy. Duplicating film requires contrast reduction in advance. In your case, you might get a bit of contrast reduction automatically simply because you're enlarging MF to 8x10, but it might not be enough in some cases. Good luck with your project!
Thanks !
of course, anyway today it's way easier to make good LED sources, or at least good illuminations are easier to find. Spectral match has improved a lot in recent times, in general.
Posting Permissions
Reply With Quote
Bookmarks