I tested the tail light LED assembly today with pre-flashed Arista Edu #2 glossy paper, which I believe is Foma Fomaspeed "Normal" contrast paper.
The Foma site shows the following curve for spectral sensitivity of the paper:
The safelight recommendations are:
...and from another version:
The spectral curve of the LED diffused globe bulb used in the Yankee housing from Superbrightleds is:
The large blue arrow shows the longest wavelight to which the paper is advertised as being sensitive. The recommended safelight illumination is indicated by the black arrow.
I will assume that this is a fairly representative curve for "normal red" (not "photo red" or "deep red" LEDs), as I have no data on my 12 volt tail light.
Here is the tail light assembly, from Lowes:
https://www.lowes.com/pd/Reese-6-in-Oval-Sealed-Light-with-Chrome-Rim/50437648
It can be seen in the graph that there are small portions of the red LED output in the green region and below the recommended safelight wavelength. I can easily see these by viewing the off-axis diffracted reflection of the LED in the playable side of a CD.
Some tilting and adjustment of the CD is necessary to expand the spectrum to maximum size in the reflection, but the green is definitely there, indicating the need for additional filtering.
I can get up to 6 minutes exposure from the high setting of the tail-light with the Rubylith filter without fogging of pre-flashed Fomaspeed paper. That is double the recommended maximum exposure. The light is extremely bright and easily enough to read by. It is indirectly reflected from the white ceiling at a distance of 18 inches (light to ceiling) and 1.5 meter from ceiling to trays,.
At the low setting, representing the "running light" (non-brake light) brightness of the assembly, I can get 30 minutes of exposure without any trace of fog. The light is sufficiently bright to read by and a more "reasonable" level of illumination.
These results (with the Rubylith filtering) are several times better than obtained without the filter.
I have been unable to find any specs on the filter characteristics of Rubylith, but it clearly has a beneficial effect.
I also see NO signs of green with the CD test using Rubylith.
I hope this is of some use.
Regards,
Don