Yes, the backlight makes a HUGE difference. Diffused sunlight bouncing off a white card works well. Solux 4700K works well, if you can still find a source. A 3200K Quartz-Halogen source, fully diffused through milk Plexiglas, works, if you can dissipate the heat. Come to think about it, I need to try the light from my 36-year old Ektagraphic III AT slide projector! Bounced off a white card through milk Plexiglas ought to do it, if I build a proper rig. Photo grade incandescent (3200 to 3400K) lights have continuous spectrum output, although strong on red and weak on blue. Simple white balance with eyedropper tool should fix that, though.
Methinks you guys are overthinking this a bit. I think the diffusion is the most important thing. If I had any opaque perspex I would probably try a using a diffused flash further diffused through the perspex. Also moving at least 6 inches away from the light source works the best. You can always do a custom white balance straight from the diffuser.
Diffusion is important, but so is the quality of the light. Imagine a backlight composed of green LEDs. Youād probably not get any enjoyable color conversions from it 
1 Like
Maybe Iām just picky!
Aperture is important. Using an aperture 2-3 stops down from wide open maximizes lens performance on most macro lenses, avoids contrast reduction from diffraction, AND avoids deep depth of field that would render any pattern in the light source such as the surface texture (or fingerprints and dirt!) on a diffuser.
SPACE between the film and the diffuser is also important. The light diffuser should be several inches behind the film, to ensure that it is out of focus.
Light sources that do not cover the full, continuous spectrum of visible light are likely to result in color shifts that you do not want. Thatās why nearly all of the 78 optical color negative printers in the lab I worked for in the 1990s used quartz-halogen lamps, projected through translucent dichroic color correction filters and then very heavy Plexiglas diffusers (Brits call it Perspex). Tungsten filaments output a continuous spectrum ā but itās lop-sided towards the red. Thatās why film has an orange mask, to compensate for it. Remember, youāre working with opposite colorsā¦ To add blue to a digital capture or a print (i.e.; compensate for a LACK of blue in the light source), you take it away from the light going through the negative.
Electronic flash is a discontinuous spectrum light source. It does a REASONABLE job of rendering color, but what it leaves out is noticeable in side by side comparisons. There is another thread here in the forum that compares a lot of different light sources, with examples. Itās worth a read.
YES, the diffusion is very important. So is evenness of illumination. The light intensity has to be consistent from corner to corner, edge to edge. Absolutely no patterns or details should be visible in a capture of an unexposed, processed film chip, other than surface dust and film grain. If you see fall-off, or areas of the frame that are darker/brighter than others, make adjustments.
Blocking stray light from causing lens flare or reflecting off of film or film holders is also very important. Everything surrounding the film, including the film holder, should be black. Back in the 1980s, when I duplicated thousands of slides as an AV producer, I wore a black apron. I worked in a dimly-lit copy room with black walls, a black ceiling, and a black shroud mounted to the macro lens to block reflections from the camera body. We never had issues with flare, because we eliminated the sources.
In-camera custom white balancing off the light source is not necessary, because you should be recording only raw files for use with NLP, AND because youāre always photographing through a colored film base. The procedure for white balancing with the eye dropper tool through the film base between frames takes care of it most accurately. Every color film emulsion has a slightly different color mask, that varies with brand, film type, film speed, emulsion batch, age, negative storage conditions, and film process accuracy. So the eyedropper tool eliminates all those variables in one click.
While this may seem like over-thinking, I assure you, itās knowledge gained from over three decades of experience in the photo lab business, and 52 years of serious photography.
2 Likes
Burk, thatās a āput downā as a relative copying newbie I am happy to accept. Your years of experience shine through your answer.
Thank you for taking the time to write and post it.
Much appreciated.
Is there a risk of undermining a very high CRI index by using a piece of Plexiglas/Perspex that hasnāt been specifically created for this purpose?
Would any white/milk panel do, or is there something Iād need to consider when building a diffuser myself?
I am in fact attempting to do something using some 99 CRI LED Linear Modules.
Answering that is beyond my knowledge base. But I do use a Viltrox L-116t light panel underneath an Essential Film Holder, which includes a sheet of Perspex. It seems to work well. Is it perfect? I donāt know. Iām sure someone would say not.
Indeed. Of that, here, I am most certain! 
My thought is that you can spend your life in a world of diminishing marginal returns ā and never get much done or have any fun. Sometimes perfection is the enemy of good enough. (As an all-too-often perfectionist, Iām talking to myself here.)
If I ran a photo lab like I used to work in, I would be concerned about 99 vs. 95 CRI and the effects of various diffusion materials. But as a retired hobbyist and occasional project for hire guy, Iām getting the results that work for me, at minimal expense and hassle.
PerspexĀ® Spectrum Opal 1TL2 Acrylic Sheet is recommended as being preferable for its ability to pass the entire visible spectrum consistently, itās easily obtainable and relatively inexpensive. I have no idea whether any difference would be discernible but it is advertised as being used in one very popular and successful film holder. Similarly there is little real world evidence online as to the visible difference between using a 95 & 99 CRI light source so I suspect it is slight. Generally 95 CRI is lower in the red and I wonder if colour profiling was more accessible for camera slide copying whether that would in any case accommodate, or correct for, any differences.
The light source thread on here:
1 Like
āSometimes perfection is the enemy of good enough.ā
Boy, do I agree with you there. Iāve gone into analysis paralysis too many times, and perfectionism can suck the joy out of hobby. In fact, part of the fun for me is assembling Rube Goldbergesque contraptions from found or free materials.
It is too easy to ābuyā a solution, and I refuse to go that route, although it is certainly slowing down progress on my giant box of my Momās negativesā¦
Minimum expense is important to me; it usually requires maximum creativity.
I believe that is the Perspex diffuser used in the EFH or Essential Film Holder.
There are some excellent videos on YouTube by Hashem at Pushing Film where he compares several different light sources. One of his favorites is the Viltrox L-116t, which is also recommended (along with its sister product by Releno) by the guy who makes the EFH ā Andrew Clifforth.
Here is a Dropbox link to my white paper on how I use the EFH and the Viltrox light panel in a home-brew setup.
I agree, I didnāt think that the comparison with the Negative Supply 99 CRI light and the different brand 95 CRi was all that conclusive and he agreed in the comments that profiling would have probably brought them much closer, but profiling is not straightforward, or at least itās pretty pricey to do it right. He didnāt compare Negative Supply 99 & 95 CRi panels, and neither do they as far as I know, except by charts, but then they probably wouldnāt want him to in case the difference wasnāt that marked, you pay a lot more for 99 CRI.
Yes, thanks, your guide is good, Iāve read it before, a few times in fact. I like your adaptation of the EFH to address the fact that it doesnāt take short lengths, which is otherwise a potential disadvantage of that holder.
Yes, it was the EFH that I was referring to regarding the perspex.
Thanks. I had a chat with Andrew (EFI maker) and he assured me he will be addressing the need to handle single negatives and shorter lengths in the near future. Iād love to get rid of the negative carriers (sell them with my B22XL).
As for my copy stand, wellā¦ Iāll probably buy a commercial unit with finer, faster control over height and leveling. What I have works, but just. I had a huge Bencher unit with 48" column back in the 1980s. I miss that!