I feel sure that you will be going to get those 5000ºK LEDs, having got this far I may well have done so also. There are a few things that you’ve said that I’m not entirely sure about though so excuse me for highlighting them in case they matter.
Firstly, you did underline a quote from Nate from his Guide " It’s generally recommended to use a daylight balanced light source with a temperature of 5000°K or 5500°K - you certainly don’t want your light to be any warmer than this!" That is an argument not to use light sources with a lower colour temperature (for colour negatives!), not a slightly higher one like 6500K.
You don’t say what software you wil be using, you don’t actually say if you are shooting RAW but I imagine that you are. With RAW you are free to set the WB to anything you like, within reason of course. It wouldn’t be a question of setting it for each exposure, as I said in Lightroom you can use a preset to set it across all your images as you upload them.
Alternatively I had a look at your Canon R5 manual and you can set a custom value between 2000ºK to 10,000ºK, alternatively ‘Daylight’ is 5200ºK and ‘Shade’ is 7000ºK but with your custom light source there is no need to choose a default. The manual also says if you measure the colour temperature you may need “to take some test shots in advance and adjust the setting as needed to compensate for any difference between the color temperature meter and
the camera.”. Whereas that advice is meant for ambient lighting the same must be true for this application.
I don’t know why you are finding that you are needing to adjust the WB for certain images, is it simply correcting for the colour balance in the original picture? In any case I don’t see that changing from a 6500ºK light source to a 5000ºK light source would have any effect on that.
Since you are using your own custom light source and are likely to stick with your Canon R5 and your Canon 100mm Macro this would seem to be the perfect reason to get one of the Wolf Faust targets. Even if you don’t go ahead and make a colour profile straight away you may decide to do so in the future and you will always have this as a reference. Similarly if you change any part of your system, LEDs, lens or camera, you can do another profile and expect consistency with those images copied before your changes.
Colour temperature set in camera is only relevant for ooc jpegs and the raw’s thumbnails and previews. You can set whatever you like and the raw file will be the same with a few exif tags that have other values which Lr will use, but NLP will not use.
I mostly set UniWB for a histogram that relates to raw data more than, say, 5200K does. Previews take a green hulklike appearance, but converted images are fine.
BTW, UniWB helps to set exposure. The negatives I scan are about 50 years old and exposure varies greatly due to guessing the exposure when I took the photos back then.
Aware of that but the OP is very concerned about the WB of the light source and how it is reflected in the resulting images, not entirely sure why in fact because this is set in post with RAW files. As you say in any case it does not matter what you set with NLP and NLP does seem to cope very well with different colour temperature light sources, however Nate does say that generally a 5000ºK to 5500ºK light source should be used, as in the quote. However the OP is seemingly not copying colour negative anyway.
The crucial question as far as the OP is concerned is the difference between these 99 CRI Waveform 6500ºK LEDs currently in his panel, bought in error, and the 97 CRI 5000ºK LEDs which will cost $400 or so and a day’s work.
I think the post would get better traction, and wider range of advice, if was posted on its own rather than in this thread in fact. Might be worth doing.
I’d be happy with the 6500K DIY Lightbox (except for its thermal design) and start scanning with well thought-out test plan about settings, exposure etc. Some additional power in blue helps with the orange mask too.
Using the dichroic head, I set it to compensate the mask … but found no noteworthy differences in converted images to make that compensation worthwhile.
Yes, Nate’s quote was on the opposite end of the temperature spectrum of Peter’s quote. Given both, I had two concerns:
Am I losing color information by using the wrong temperature light? This was Peter’s concern is his book (and videos on the included DVD). In subsequent testing with the IT8 target(s) I have it was not apparent this was a problem, and this concern is alleviated with color slides (I have not tested with negs as I do have a negative target if there is such a thing).
Having to adjust the WB in Lightroom manually. For a number of reasons I am using Canon EOS Utility for remote shooting, and capturing in RAW. This process requires that I choose to a. Set the WB in EOS utility (guessing at it, or just pinning at 6500K), b. Change it in LR manually (or a preset), or c. Ignore it altogether. As I prefer the utility and am using Lightroom 6 LS11 which does not recognize my camera (or CR3), I can’t adjust as I go.
The WB is off per the target, however the adjusted WB does not work for every image. This may be due to the film type or the WB in the image. Given this, the revised ligthbox may not resolve it. As I am shooting RAW, I think I will deal with it later, but build another lightbox with 5000K LEDs (which I have ordered ) as a hedge.
I am not sure a profile would help here, unless I make one for each film type.
UniWB is interesting, but having looked at it prior, I thought it would likely be cumbersome in my workflow.
Given the DIY nature of my lightbox, there were just a few options for thermal management:
a. Reduce the number of LEDs and find a base material to mount the LEDs on to insulate the bottom of the metal box. Or use some other material for the box. The main concern here is over heating the diffuser material and/or the scanner base/film holder/film.
b. Reduce the number of LEDs and mount some heatsinks around the sides of the box for a passive solution. This likely would not reduce the heat on the base of the box, which in turn may damage the copystand.
c. Keep the large number of LEDs for the very bright solution, and actively cool it. As I build a second box I will reduce the LEDs by 10%, lower the fan speed to 300 RPM, and reduce the number of holes drilled in the box.
If someone has a better suggestion for a passive solution, please do let me know.
There’s quite an exhaustive thread on the practice of colour profiling, perhaps you’ve seen it. Many different opnions expressed there.
I was wondering, given that the heat is a price worth paying for you, what shutter speed are you able to use?
I use a Cinestill CS-Lite and for 35mm slides I’m using around 1/160 at f5.6 at 200 iso. It doesn’t get hot at all, the Ipad wall adapter does get a little warm. I’m in no way suggesting that this compares with the quality of your light source, just wondering why yours needs to be quite so bright (and hot).
1/250 @ f/4 ISO 100 for an IT8 target +2/3 exp comp. I do have some dark 6x7 slides that will need more light, but I don’t need 70+ LEDs, so I will reduce them, but am concerned as they 2" away from the diffuser (the box is taller than most). I can use shorter strips, fewer strips or both.
If I shorten the strips by one LED on each end I am concerned about light fall off on the ends of a 6x7 image, and can’t go back and add them. I can however, test his with the current setup by covering them and testing. If this works I can reduce by 9 or 18 LEDs.
If I use fewer strips I would likely leave one off each edge, otherwise I would be concerned about even illumination. This could reduce another 16, leaving 38 in the box, or about 10 less than a Negative Supply 4x5 pro, which may be too much of a reduction, and without an integrated heat sink solution, may not be enough for a passive solution considering how thick this guitar pedal box is.
I think I will just reduce strip length for 54 LEDs total and hope I don’t need the fan, and the holes will be enough for passive cooling. I am thinking about adding a 1/4 gatorboard under the LEDs to insulate the bottom of the box.
If someone knows of a better box to use, please let me know.
I suppose there is also the slightly left-field option of Iridient C-Transformer, the resulting dng files could then be imported into your version of Lightroom. You can trial it.
My workflow is to capture with EOS Utility, convert with Adobe DNG Converter, Stuff metadata with Exiftool, then edit in LR. Is there an advantage to Iridient C-Transformer over Adobe DNG Converter?
Not sure because I use Fuji, and my Fuji X-T2 files are handled by my Lightoom 6.14 anyway. The Fuji version, X-Transformer, is certainly better than my version of Lightroom for Fuji .raf files but it is only noticeable with certain types of images, generally landscapes with distant foliage, so I use it only rarely. Fuji X-Trans files have their own processing issues because of the unusual 6x6 grid of photosites compared to the 2x2 Bayer grid.
I’m well out of my depth here so this might be considered silly, but how about mounting COB LEDs on a large heatsink of the area required with the proper adhesive. The heatsink could be fan cooled underneath if necessary. Soldering up the terminals would be a problem? The COB LEDs could be arranged in any pattern and number.
Here’s a COB LED I have been investigating available in 98CRI 6500K, 5600K, 3200K or 2700K.
I’ve just found the thread that I was thinking of here:
This shows the construction of the 99CRI Basic, this one was faulty. The Basic has side illuminated LEDs but they are fed through what was described as a ‘prism plate’ which turned out to have a manufacturing fault. Given that I think that my CS-Lite uses a similar contruction and I’m getting a similar light output to you, albeit from a much smaller panel, perhaps this is a more efficent way of harnessing the light output. As I said I can leave the CS-Lite on continuously and it doesn’t get hot.
Mind you I have a feeling that the NS ‘Pro’ version uses direct LED illumination but as you say that doesn’t have a fan either.
That could give you very “spotty” results. With small led’s you can get them tightly packed next to each other what means less diffusion material = more light. I use Aputures AL-MX for example for 35mm setup.
Any particular reason why you aim for 1/250? To avoid vibrations?
I run often below 1/100’th to get F8 and ISO 100 for my setup - as often it’s 4h+ run, I keep lights at about 50% of nominal output to avoid long term heat built-up.
I have no idea how to make a sandwich like that myself. Regardless, I thought the basic was a bit weak, which is why when mine had the same issue as in the post (it looked like a part of the LED strip came detached) I returned it instead of replaced it.
After the cost of paying for this twice now, with enough LEDs to make 6 boxes equivalent to a Pro box, I might have been better off sticking with the Basic…
I tested the fan at 370 RPM and there was very little heat buildup, and the fan is silent. On the new 5000K box I will mount 54 LEDs to 1/4" Gatorboard and test in the box that without a fan or holes (other than power in), and only add a fan if it needs it. Concerns with this are 1. can the Gatorboard take the heat, and 2. will pushing the LEDs up 1/4" impact diffusion negatively.
If I were to pick my darkest image I might get 1/125 or less with this light. I am okay with a slower shutter and would use electronic shutter if it was 12bit, but on the R5 it is not. But half the light (like a Negative Supply Basic model) is not going to cut it for me.
As discussed above I will reduce the light and add insulation in an attempt to run fanless, and if that fails use the fan at 370 RPM which is silent. This leaves the only concern being dust build up, which I will just do a regular cleaning.
Update! Built a 5000K passive version with 9 strips of 6 LED’s each (instead of 9x9). Using a 1/4" Gatorboard as a base for insulation, while it does get warm, it is not getting hot or heat soaked. Also, having the LEDs closer to the diffuser (due to the Gatorboard) did not cause hot spots. This solution should work for 6x7, and I will use the 9x9 6500K when I come across the rare 4x5 negative.
See attached photos.
As promised in my post in “Let’s see your DSLR film scanning setup!”…
Here is the first Collimated vs Diffuse backlight source comparison using my partner Gabrielle’s photo taken with Nikon FM on Ilford HP5 Plus at The Vatican Christmas Day 1993.
I’ve tried adjusting the diffuse lit image to match the collimated. I can get it close, but for black and white I will stick with collimated backlight.
Collimated vs Diffuse comparison using my photo taken with Nikon FM on Ektachrome E 100 in 1987. I’ve made no adjustments. Usually I bring up the shadows which works quite well! PS. I forgot to flip the image if anyone noticed?
Sony a7C - Sigma 105mm f2.8 DG DN Macro Art
Collimated: Durst Otocon 130 condenser to film and 180-1 away, with Philips Photocrescenta 100 CRI and Wratten 80 Gelatin Filter.
Diffuse: Same with opal acrylic diffuser.
LED: Negative Supply Light Source Mini 97 CRI.
Interesting to see this comparison, I also have a 6x9 condenser enlarger that I was thinking of trying instead of an LED panel (CS-Lite) but of course it does require quite a lot of commitment to make it work, not to mention desk space. For me with your Taj Mahal example the jury is out, I can’t tell. From what you say the difference is more marked with B&W silver halide film though it’s a little difficult to discern from the full frame examples, for me anyway. Thanks for uploading.
) as a hedge.
