What is anti-Newton ring glass exactly

[Pere Casals]

Yes... I propose testing on real RA-4 paper, printing on it an IT8 target

[Drew Wiley]

Well, good luck with that, and when you get results, post them.

[interneg]

...needing less a G type lens that's required if wanting performance at large apertures, f/11 also allows greater DOF to cast all "through the negative"quality on the print.

There's a long list of why a G-Componon or Rodagon-G is a good idea when you're working in their optimisation range & it's not all about the aperture where performance maxes out. For that matter, most longer f5.6 enlarging lenses of recent generations seem to hit optimal performance about a stop down. And if you're using unsharp masks etc, the wider an aperture you can use the better, again for a variety of reasons that are pretty obvious to those experienced in the art.

Next step will be checking if an special LED retrofit will work OK with color paper. I plan to project an IT8 transparent target (came with an EPSON) on RA-4 to see if I can obtain the same with LEDs than with Tungsten, scanning both prints (from the IT8) alongside and comparing.

You could do this test far more simply with an average negative or two & a set of separation filters. And you are going to run into quite a few problems with your use of an IT8 for a procedure it is not designed to do. You are going to get a very limited quantitative dataset with no qualitative understanding of what it is you are imaging. Your approach needs to start qualitatively (making the best possible prints from various negatives with both light sources), then if it looks like a worthwhile path to explore, quantitatively analysing the visually significant differences. That is where an IT8 print target will be useful.

[Pere Casals]

There's a long list of why a G-Componon or Rodagon-G is a good idea when you're working in their optimisation range

The G componon is nearly impossible to find, and Rodagon G is quite expensive and found in very long focals...

A friend has a G and he found it only makes sense for large apertures, you may expose one stop wider...

And if you're using unsharp masks etc, the wider an aperture you can use the better,

But a large aperture also has drawbacks... also known by those experienced in the art...

For that matter, most longer f5.6 enlarging lenses of recent generations seem to hit optimal performance about a stop down.

The lens may perform well enough wide open, but still we benefit from additional DOF if we stop two clicks, making focus/alignment a bit less critical. Single reason to not stop those two clicks is shortcomings in the illumination power leading to excessive exposure time, but a LED powered condenser doesn't have that shortcoming: Job is done without excesive reheating in the negative (and in the entire darkroom!).

You could do this test far more simply with an average negative or two & a set of separation filters. And you are going to run into quite a few problems with your use of an IT8 for a procedure it is not designed to do. You are going to get a very limited quantitative dataset with no qualitative understanding of what it is you are imaging. Your approach needs to start qualitatively (making the best possible prints from various negatives with both light sources), then if it looks like a worthwhile path to explore, quantitatively analysing the visually significant differences. That is where an IT8 print target will be useful.

Well, by simply comparing the RA-4 (negative) prints of the IT8 target, by naked eye, we may see if we can get or not a good enough match in LED vs Tungsten, by correctly scanning both prints alongside we may find quantitative nuances that perhaps cannot be perceived by the eye, but still it would be good to know what quantitative precision we have...

My guess is that I'll be able to retrofit the 138 head with RGB LEDs for perfect RA-4 jobs, what I don't know is if I'll have to place some LEDs to fill the spectral valleys to adjust channel crosstalk.

With LED, probably the 3 main colors are easy to balance, but a discrepance may be there in the intermediate tones, which would be balanced with additinal LEDs of 500nm and 580nm, controlling the crosstalk. This may require an analytic study...