Light Sources for Film Scanning
Vložit
- čas přidán 3. 08. 2024
- A video discussing the differences between light panels for digitizing film with a mirrorless or DSLR camera. The importance of CRI, diffusion, distance from the film, and more are discussed.
Thank you to Negative Supply for providing their light source for testing purposes, check it out at the following link!
►NS 99 CRI Light Source Basic - www.negative.supply/shop-all/...
🖼️ PRINTS - hashem.darkroom.com/
☕ Buy me a Coffee - www.buymeacoffee.com/pushingfilm
👕 Merch Shop - pushingfilm.myspreadshop.com
📘 My Zine - www.exaframe.com/product/sund...
►Dev & Scan gear purchase links:
The Essential Film Holder - bit.ly/47nuFOg
L116T LED light - amzn.to/3GMzKUE
Logan 4x5 Light - bhpho.to/478Jvax
Raleno Rechargeable slim video light - amzn.to/31wMLOt
Rocket air blower - amzn.to/2G0pZpN
Negative sleeves - bhpho.to/47gAXyw
Negative binder - bhpho.to/3v93RmD
Paterson developing tank - amzn.to/3TwrFek
Rocket air blower - amzn.to/2G0pZpN
Negative sleeves - amzn.to/32rEuub
Canon 5D MK IV - bhpho.to/3vaVzLa
Canon EF 100mm Macro - bhpho.to/3NDO05X
Using the above affiliate links to make any purchases allows us to earn from qualifying purchases; which is a much-appreciated way to support the channel without costing you any more!
►WEBSITE
🏠 www.exaframe.com
I use Pixpa for my website - an affordable and effective website solution.
If you would like to try Pixpa you can get a free trial at bit.ly/3UTof2e and get 50% OFF on annual plans when you subscribe with the promo code PXREF50
Video Chapter Markers:
00:00 - Intro
01:17 - What is CRI
01:39 - Format of video
02:18 - Scan comparisons
05:26 - Above vs below 95 CRI, initial findings
07:03 - Scan comparisons
08:59 - 95 vs 99 CRI
11:30 - Is a 99 CRI light worth it?
11:56 - Slide film comparisons
14:51 - Key takeaways
17:48 - Your experience? iPad results?
19:04 - Conclusion
#filmscanning #negativesupply
Stoked you did this. I'm endlessly frustrated with what I've been using. Cheers man!
No problem! Happy to help
btw, glad to see another dedicated film scanning person existed on CZcams. I had owned Epson V700, a Howtek D4000 drum scanner and now I am doing scanning with MACRO lens. btw, I would recommend tuning your camera to UniWB, it will help you achieve maximum bits range when you do pushing to the right exposure, which IMHO is kinda necessary for slides due to their very high dynamic ranges. The steps are quite easy as you only need to adjust your red, blue multiplier to around 1 using a pallete APP on your phone or tablet.
Thank you very much for your review! very helpful as always.
Thank you so so much for testing this out! :)
No problem 😊
Thanks for the comparison. Was looking for a video like this
Very useful video with tons of good data. Many thanks.
Glad it was helpful!
Great info and well presented thanks for doing this. I use a durst colour copy 350 to digitise my negs. It’s an old slide copier set up that are pretty much worthless these days. You mount your camera to the stand with its inbuilt bellows and you use it with enlarging lenses. I never had issues until the halogen bulb needed a replacement and I thought I’d use a light pad, wow so many issues due to low CRI and false positives for light leaks due to the light pad not being directly below the essential film holder. Replaced the halogen lamp and back to rock solid 100CRI. I was surprised at how finicky the led light pad was, granted it was the wife’s spotlight bought one with no advertised CRI.
I’m definitely upgrading from my Logan!
What a fantastic comparison, thank you for putting the time and effort in on this! I've been using a home made setup of a massive diffuser box made from the frame of an old computer desk, lined with reflective panels on the inside and a solid piece of (masked) frosted glass on top with a down firing Godox 400 strobe. I can scan (using essential film holder) at around 1/100 sec at f11, the strobe is set to about 1/4 power and wirelessly triggered. Scans are great, colours are pretty much bang on lab scans and no sign of any casts or blotches. Only problem is the whole setup takes ages to set up and pack away and really grates on me if i cant get my scanning done in one session. I've just ordered a NS CDI 99 light from your review results, hoping this may streamline my workflow a bit. Congrats on an awesome channel!
Sounds like quite the set up! How do you go about previewing and aligning things? just using the camera's "exposure preview" type setting which cranks up the gain for live view? Thanks for the kind feedback, I'm glad the video helped you pull the trigger on the NS light :-)
@@pushingfilm Hey thanks for replying, i forgot to mention the preview setup. I have a cheap flouro work light that sits on the floor that i use with a dodgy foot switch. Switch on to align and switch off before capture so i dont get wierd casts. The flouro illuminates the whole diffused glass top so the table doubles as an excellent negative-in-sleeve preview (light table) with a loupe. I use a canon RP and 100mm macro on a tripod set to manual focus, tethered to my PC so dont really need to mess with the camera once its dialled in (unless i knock it) i also use a bit of gaffer tape to hold the Essential film holder in place while i load and pull thr film through. It really is a bit of a frankensteins monster that takes a bit to setup and pack down but once its dialled in you can knock out a heap of rolls in one sitting. I cant wait to get the NS light pad for simplicity! There are some other comments here about using strobes and speedlights, seriously give it a go if you have the time as all issues with CRI / Brightness etc are moot. Just dont expect a streamlined setup:) Cheers mate, and keep up the great work on your channel.
Great comparison, thank you!
Glad you liked it!
Wow this really makes a difference. I've been using a cheap led panel that I have no idea its CRI. But I do have the Falcon Eyes F7 and never even thought to be using that for scanning. Now i'll at least switch to that. Thanks for this. That Negative Supply light really made a difference too..
Yeah it can make a difference! If you do a test of the same frame with both lights and see how they vary, let me know 🙂
Did you get any results with Falcon Eyes F7? I'm also considering using mine for scanning.
This is great
What a boss! Love it 🤩🤩🤩
Thanks man! 😁
Really useful. Thanks.
Glad to hear that!
Just a suggestion for some testing that you may want to consider...
I think it is viable to use the light source to calibrate your digital camera for better color performance. Basically, take a color target and your light source and set it up to provide good uniform light on the color target. Use the camera and lens you intend to use for the copy work. Take reasonably tight photos of the color target and use the settings that are as close as possible to what you will use in the scanning (but probably with a different exposure time).
Take that file and run it through your calibration software as normal but this will be a special profile just for scanning purposes.
If you do this, I suspect your 95CRI light source may become very close to your 99CRI source and you won't need to do heroic color correction to get approximately the same color from both.
I doubt this will be able to resolve the shortcomings from lower CRI sources because the colors will be a bit too discontinuous to correct, but it would likely improve all of the light sources somewhat by taming the odd colors a bit and bringing up the defficient colors somewhat.
I have done a lot of scanning of chromes (35mm) for my father (50+ years of slides to archive for him) and I used a high CRI compact fluorescent light source to do it with an olympus slide duplicator and macro bellows setup (this stuff is pretty inexpensive on Ebay these days) that I adapted to a Sony camera and later to a Fuji GFX camera. I did a lot of testing and the results are quite excellent.
I am now putting together a much more sophisticated copy setup for 4x5 and MF film. I got one of the larger Negative Supply light sources (95 CRI) and am going to begin testing the setup very shortly (I adapted an old Durst enlarger for the copy stand arrrangeent). I have four lenses here that may be viable for the 4x5 and MF copying tasks, and I'll be doing lots of testing to figure out what approach is producing the highest quality results, and hopefully, the quality will be high enough for my needs (I'm pretty particular on these kinds of details).
However, I will be testing out the camera calibration approach that I mentioned above. I have done this in the past with the fluorescent source and it did work well. I'm note really sure it will be needed for high CRI sources, but I think it will probably benefit for doing the calibration.
@@mjmdiver1137 let us know your results on calibration, sir!
Building out my DSLR scanning setup.
Bought a EFH to use with my D750 + 60mm 2.8 macro.
then I was rooting around my gear and remembered that I bought a light panel back in 2019 just for funzies to use to shoot dancing
Open my cabinet and find that it’s a Viltrox L116T
WHAT.
I bless my past self for researching and buying the right panel
$400 is not worth it to me rn for 99 CRI
Nice! You'll do perfectly well with that combo
Thanks for sharing this. I appreciate your taking the time. One thing I'd love to see you add is tests using a strobe or flash alongside the LEDs. I'm putting together a system now using a flash with a modeling light as the source. In theory, the broad spectrum light should be superior to mixed LEDs but I won't have an easy way to test it against LED panels.
Thanks for watching! This is something I've been meaning to try for a long time (I tried it ages ago around 2013, but my setup was terrible and I didn't have the conversion tools or knowledge yet). I have heard that there would be such advantages, so I might try and rig it up with the EFH sitting on it's side, with some kind of makeshift snoot on my 430EX speedlight!
Yes to this. I have a strobe setup that uses a diffused source into a white box with the EFH on top. It eliminates any shutter speed issues but very curious to know how color rendering is affected.
@@morefiction me too! A small rolling stand with a white diffuser box on the lower level and a hole cut into the top surface. The EFH is mounted on top and I have an adapter that mounts a macro lens directly to the EFH. Now all I need is to finish a large job scanning old stills so I can turn to the negatives...
Yes to this as well! I was hoping strobes and/or enlarger bulbs would be included in this comparison. I am currently awaiting my EFH and I will likely do a comparison between my Kaiser Slimlite Plano Lightbox and a strobe of some kind. Would be great to see how it compared to all those other options though.
Would be interesting to a comparison between the Negative Supply panel and a Kaiser Slimlight Plano
I should totally buy this. Then I would have a hodpodge mix of different holders to scan with. I would have essential film holder, valoi and then NS in a single setup :D
Thanks!! (ps: incandescent and halogen light sources have a CRI of 100. But not easy to use for scanning)
Fun to look at these while being mild colorblind. I need an excellent light, so that I do not need to correct, as that would be futile
Hi and I had just did some experiments with your recommendations and lifted off my film slide about 3~4mm above the light table, however the blooming/bleeding at high contrast edges do not disappear completely, still very apparent, and now my film curvature definitely hurts the clarity(before it was against light table in complete flatness). Before when I had film directly taped against light table, aperture 5.6 can be achieved for full clarity (Sigma 105mm Macro Art), but now due to curvature, I have to lower it to F8 and the resolution definitely hurts.
I then did some other experiments, and found that lowering the light table (Raleno) brightness intensity definitely helps with blooming issue at high contrast edge, and lower to about 50% of original brightness can achieve very good blooming/bleeding light balance vs film detail clarity. now the blooming is curbed about half as bad as it was before, and I don't have to lift the film above which will result in curvature. Anything lower than 50% brightness, the dark portion of film begins to loose the clarity but the blooming issue definitely improves further. It seems to be an art of balance to achieve the best result but then again I would probably blame Velvia for being way too high in contrast, which is an very hard film to scan.
Useful info, well researched and explained. Side by side comparisons tell the story. Good points about film elevation over light source, and ambient light. “CRI” new one to me - learned something.👍
Re iPAD, screen protector definitely suspect. What method did you use to have the iPad display a white field? I’ve taken a “pure white” jpeg image (white balance checked in LR) and then displayed full screen in Photos - seems to work OK.
Calibration. If you do lots of scans it might be worth the effort to calibrate color by shooting a color chart, scanning via favorite light source and then correcting in LR; build/apply correction pre-set. No?
Some folks scan w/digital camera placing film emulsion side up - more accurate focus; then reverse in post. What’s your preference?
Suggestion for future vid (unless I missed one you already posted): “bulk up/roll your own”. How-to and value proposition of loading your own film canisters from bulk in 2022. Cheers!
Hey, thanks for the comment! I think someone mentioned the display mode being suspect, if its on true-tone or something similar. I need to check those settings and do another test. I have been meaning to get a colour checker for a few things like this, but the good ones seem rather overpriced! Since it's for just occasional use, I might go for one of the cheaper Amazon ones. I definitely prefer to capture with the emulsion side up, then flip them all in Lightroom. I find it reduces the chance of any minor stray reflections sneaking into the capture, and it gives a sharper rendition of the grain in some tests I've done. Thanks for the suggestions also! I've mentioned bulk loading in passing in some past videos, but yet to do a dedicated video on it. Could be a good one to add to the list :-) I've currently been bulk loading Kodak 500T.
Interesting video, thank you for sharing. One of thing I'm really curious about... How do you go about setting the optimum exposure in camera for capturing negatives via DSLR?
Thanks! I try and expose so that the information is evenly spread across the histogram, which tends to be around +2/3 stop on my 5D's matrix metering (but that can vary from frame to frame) I go over this in a little more detail on my video on "Batch scanning film with a digital camera"
Thanks for the detailed review and especially for including slide film, where there is often considered no “interpretation” of the picture unlike inverting color negative film.
Do you think that given an IT8 target and a 95 CRI light source, the minor shortcomings compared to a 99 CRI light source could be profiled out?
No problem! I do think so, as long the light source has a decent output such as the 95 CRI examples used
Great video. Good investigation. Good information for the community. I feel as if I can’t be bothered to scan my film with a DSLR and would almost rather just get a Noritsu, although it would be 35mm only. Last time I was scanning with NLP, I felt that I was spending way too much time getting the colors right. Do you find that you’ve become quicker with the process once you got a proper setup?
Thanks! I have become quicker, but its definitely not as quick as using a Lab scanner like a Frontier or Noritsu. If you only want to do 35mm, you can look into the LS-600... but they've become extremely expensive... and if course hard to service or repair.
I am thinking of going this route. What you describe here wrt colours, is the white balance of the light source vs the white balance of the process you apply to the camera image. It will be interesting to see how much the poor performers improve if you change your wb setting in post?
I just ordered the negative supply light due to noticing the cinestill cs-lite just isn’t good enough when scanning portra 400. I get a purple in my shadows. When I scan Kodak gold or even Colorplus they come out good but I believe that’s because those film stocks aren’t as thick as Portra so the light goes through ok. Pumped to get a better light for scanning !
Interesting... So you think it's a brightness issue? or the spectrum/quality of the light? I've yet to try the CS-light.
I thinks it’s a little bit of both. It might be ok to get by with certain film stocks but not for all which is the opposite of having a dedicated light for scanning film. I think two having different modes on the light could play to inconsistency overall. It is rated for 95+cri but I saw on Reddit someone measured the wavelengths of the light and the “cool” setting for color negative was 89cri with a 70r.
@@pushingfilm Just wanted to give an update. I just got the NS 35mm cri 97 lite and put it next to the cinestill cs-lite and there’s a magenta hue coming from the cinestill light and the NS light is definitely brighter. I’m gonna scan tonight but I feel that’s definitely why I’m getting magenta/purple in my shadows.
@@viewfinderthis I see! I would since it gets inverted, that magenta would translate to green... But maybe there's more to it than that. It'll be interesting to see how you go with a side by side test using synced exposure and conversion settings.
never seen a so deep study on the subject tahnks for doing it. Is there any other option to the Essential Film Holder ?
What about OLED screens and some kind of masking app. When colors are off then you can compensate white balance using light source. Also using oled you can apply color temperatures so you may get more accurate colors and loose less info. You can compensate tint when you scan borders and make adjustments.
I really want the iPad option to work, but I scanned some Pro400H using my iPad Pro 2021 11inch and I got oversaturated scans (and reddish cast). I raised the holder but I'll ty again with additional diffusion to see if I get different results. Great video, btw!
Thank you! Yeah I think I'm gonna do more investigation with the ipad display settings and try it again
@@pushingfilm Download an App called "Color Screen" that allows you to override all the nasty settings and give you the brightest, purest light you can get on your IOS device.
Suggest you remove the screen protect if you want that final 1%. Latest retina iPad since 2018 are OK, but they are not so bright (= longer exposure times).
Really thorough. How did you mount or stabilize the Raleno 116? It doesn't look like it will lie flat on my copy stand.
It lies flat on the battery, and I stabilize it using a method shown in my video on "batch scanning film"
I just bought a kodak light box i hope it's good lol
The current March sale that NS is having puts this at a great price. Just picked one up to compare against my Skier Copy Box II
Yep, it's a pretty good time to get stuff from there especially since there's free shipping atm (usually the deal breaker for us here in Australia buying from the USA)
I'd be keen to hear your findings too, once you do some comparisons with the Skier Box
@@pushingfilm I could also bring it over to you if you wanted to do a video with it
When I say "it", I mean the Skier Copy Box if you're interested in testing it. I'll need to wait for my NS Light Source to arrive first and then I can drop the Skier off to you.
@@donovanchin sweet! Let's wait till then and see if I can work into something
The thumbnail shows a distinct difference, but does the difference come from the CRI values or from the color temperatures of the lights?
For a better comparison of different CRI values, I would let my camera establish a custom white balance for each light source before putting a film over the light.
Yeah it would be good to run a tighter study in the future and account for more factors like that.
On the iPad issues, is it possible you had “True Tone” and/or “Night Shift” turned on? Both of these could effect the colour of the light quite significantly.
I was going to say the same thing. It’s likely the True Tone affecting the actual colour temp
I think you're onto something there! Haha I should have known to check, that's what I get for being an Android user?.. I'll have a look at the display settings, thanks
Wait... no more 5-10s of stillness before the video start! that was my moment is smooth zen before diving in! :P
I want a video of just him being silent and still for 30min strait 😂 To be fair I think he just does that when he streams but still.. 😅 "Lofi Pushing Film beats to stay silent to" would be an awesome meta play on his part.
@@AlexOnStreets I'd hit the notification bell, subscribe, comment, and hit the like button for quality content like that.
@@AlexOnStreets lol you do it first 😁
@@didiMao "Dead Air" video coming soon 😂 Appreciate the love!
I see the reason why you would want to liftup the film away from the light source a little bit, the diffuser in light tables would actual cast stray lights should you put the film directly against the table, and at high contrast edges, the scanning result would have this light bleeding/blooming issue because of the stray light goes sideways... am I right?
Yep! It's generally better to lift the film away from the light source for the reasons you mention. Also less chance of picking up dust or other imperfections on the surface of the light source.
@@pushingfilm Thanks, it bewildered me for months why all my Fuji Velvia 100f (high contract slides) all has this weird blooming/bleeding issue around high contrast edges. I should have lifted them away slightly from my Raleno light table in the 1st place .... bummer..
There’s a Kodak LED light box that has a CRI of 97-98%. I wonder if the results would be better than the Viltrox.
Excellent comparisons. I know this video is a year old but was wondering if you have had a chance to use the CS-Lite Camera Scanning Light Source? I recently set up my DSLR scanning station and am using that light source with their CS-LiteBrite+ 130% Light Enhancement Sheets. The light source is 95 CFI and the sheets claim to brighten the light. I'm also using Valoi difuser with my Valoi 360 Advancer. Would like to hear your thoughts on these products.
Thanks! I've never tried that light myself.
I need to scan a large batch of black and white negs. Is light source as critical a factor for this task?
Hello, nice video. I have a Gepe slim lite 5000 (G2001) which has no CRI number. It is rated at 5000K on the box, is there any of working out what the CRI may be?
Thanks John. This looks like an exact clone of the Logan branded panel I show in the video, in which case I can not be sure what CRI it is but would guess it's in the 80-90 range.
What about brightness/exposure, and heat output?
What whitebalance are you using in the camera? AWB or some special color temperature?
I use daylight white balance in camera
@@pushingfilm Aha, am using AWB. Maybe thats wrong.
I wonder what it would look like with a mini-led iPad pro.
I have the Raleno CRI>95 light panel with built-in 4 layer diffuser, do you think it's good idea to put film directly against the surface for scanning? Why do you need lift-offs when the panel has build-in diffusers? I have found that this way I can keep my film very flat against the light table..... Do you mind answer my concerns? thanks a lot
I haven't used the same light, but am sure it's fine for scanning. As for laying it directly see my reply to your comment on my other video.
One thing that never seems to be mentioned in guides like these is what color temperature to set the light to? Some people say that 5600k is the best and i’ve also seen 5000k. I have tried both and not see a noticeable difference between the two, so I generally stick to 5000k, does anyone have any insight into this ?
It depends on the specific light used, and the actual color temperature it outputs. This can be measured using a color spectrometer like a Sekonic C-800. Without actually measuring the light, if you don't notice any difference then I wouldn't worry about it. We're only talking about a subtle warm-cool shift between 5000-5600K.
Hey, I find it doesn't matter too much since you'll be balancing it back to make the film base neutral pre-conversion. I generally just leave it all on daylight 5600K . I've done some tests with 3200K which I'm still experimenting with for slide film (since projectors use a warm light and slides were designed with this in mind? Purely experimental though) but so far, I'm not seeing much difference.
For negative film, the color temperature of the light is less important (since the colors are interpreted in the inversion process), but for slide film, Fuji recommends a 5000K light source for viewing (Velvia 50 datasheet). Kodak doesn't make any recommendations, but in the E100 datasheet, they use 5000K for the dye density curves.
@@pulpufictione that's interesting about 5000K. The old slide copiers use tungsten lighting bulbs and they are all around 2700 - 3200K I use Osram halogen in my slide copier and they are bang on 3000K. Don't know for sure but didn't Kodachrome have a blue cast on purpose to work with the yellow cast of tungsten lighting in slide projectors? Does that mean newer slide emulsions are made for led lightning?
@@poniatowski3547 not necessarily LED lighting, but a light source that conforms to ISO 3664:2000 D50 (which could be fluorescent or halogen, but I suspect that most new systems migrated to LED for cost and efficiency)
Would you know if the Neewer NL-116AI is a good light source?
I would presume so! It looks like another variant of the Viltrox branded one I use, along with the Raleno version and a few others out there.
@@pushingfilm Thank you!
@@pushingfilm I've received the light pad today! Can't wait to try it. Should I use the highest brightness at 5600K temperature?
My question, and this is for anyone really, is it worth the investment if you are using a Epsom flatbed? I just find it slow and tedious with the flatbed at this point. I hear it’s much faster once you are set up? Any input would be appreciated!
Hey, it's different for everyone I'm sure, but I personally did find it worthwhile since I already owned a macro lens and DSLR. Therefore I sold my flatbed and just use this method now. However, if you were going to invest in a whole camera setup mainly to scan film, that might be a different story... depending on the volume you plan to go through.
I have a (quite old now, but still probably superior to the Epson offerings) Microtek M1 scanner that I have used a good bit. My experience with digital camera scanning is that if you use qualty lenses, camera, light source and technique, you will produce a scan of 35mm film that is clearly superior to the Epson scanners. Probably similarly so for MF film. Somewhere in there as the size of the film gets larger, your digital camera won't match the resolution that the scanner can deliver because the scans can get larger as you go to larger film, but the camera doesn't really offer that ability. Of course, you could do mutiple shots and stitch them...
However, whether the difference is worth the investment depends on waht you will be doing with the scans... simple screen viewing or internet display will probably not make the effort worth it, IMO.
How do you use an ipad as a lightsource?
What are your thoughts on using a Canon EXII flash as a light source?
I've yet to try a flash, but want to!
@@pushingfilm I’ll be giving it a go in the next couple of weeks so I’ll let you know of any learnings.
Out of the box thinking about light. Really enjoyed your demo of the Essential Film Holder, will be getting one of them soon. Getting back to the tight source. Lights designed for copying negatives tend to be expensive so I did some research into alternatives and came up with a S.A.D. lamp. One lamp in particular has a dimmable 35,000 lux and 3,000 to 6,500 variable colour temperature control. In the UK a typical cost is around £20 for a 6 inch by 4 inch lamp. Is this sort of lamp a viable option? I will be interested to read your comments . Greg
Hey, I'm sure it would work and that you'd get enough brightness, but I can't say much about evenness or colour accuracy since I don't have any experience with SAD lamps sorry!
What camera and lens did you use for the scans?
Canon 5DIV and 100mm 2.8 L macro
Does the Vitrox lay flat?
It lies flat on the battery. I address this and how I keep it stable if necessary, in my video "batch scanning film with a digital camera"
how did you get the viltron to sit flush with the batter. that's what I don't get
It just sits on the battery. I show/explain it in more detail in my video guide to scanning :-) czcams.com/video/B4Ex83cYAes/video.html
@17:33 - Masking that Viltrox 116- that doesn't look like normal gaffer tape. Is that some kind of Black Cine-Foil?
It's Nashua brand gaffer tape
@@pushingfilm Back for one more thing: Is it possible that the shutter speed you get with the Viltrox 116T with diffusion layer is 1/30 ? For an f/8 + iso 100 setting?
Damn, that's real low shutter speed if you wanna get sharp scans, isn't it?
We're here scans of slides with the slide film in the frame? Or did you remove the slide film?
These were unmounted positives
Both the source film and the digital sensor have 3 colour channels. Provided the light source is uniform, there is light for each channel and you expose to the right to get enough resolution, a simple white-balance correction or calibration will equalize the effect of all light sources. Comparing 95 vs 99 CRI is as silly as getting gold-plated connectors for your Hi-Fi. It's just marketing hype.
So for people who can't and won't watch a whole 20 minute video with sound.
What won and what was the main points?
Sorry but these tests are a total waste of time. You can only compare the fidelity of a light source and it's spectral response when you digitise ISO standard colour reference TRANSPARENCY film frames (slides). There are too many infra red film mask conversion variables when you try to digitise digitise colour negative film. Also when you digitise a colour transparency you can easily compare the digitised imaged to the original film frame (on a calibrated monitor). Just buy the highest CRI rated light box, mask it down to the framing device and calibrate everything. Then you are half way there - just batch adjust in LightRoom. P.S. I own a $400,000 Kodak Pro Photo CD scanning system, supply high end custom scans to Pro Labs in Australia and do all the Lucas Film still scans for Star Wars. However I built fast 36MP camera based digitisation systems like this 1n 2004. I use TTL flash systems on extension cords with diffusion boxes for the main exposure illumination along with small previewing lamps for set-up and frame positioning.
I use vintage chassis of repronar 805, 36mpx camera + best in class Rodagon Dx2 lens and flash. No any ugly LED lights. Results are OK, but not as good as real drum scanner
I hate to be a buzzkill but CRI does not matter to "accurately" represent a color negative, if the manufacturers are familiar with encoding color negatives then they themselves know this and are intentionally using it as a marketing gimmick, but in all fairness they may not be color scientists. First it's worth noting that the only quantitative profiling for color negative is how it looks when printed on paper optimised for that stock by Kodak or Fuji. The paper's spectral response is not akin to a high CRI broadband source, it's more akin to specific narrowband light in the 460nm, 550nm, and 690nm range. It's not arbitrary either, it's a very specific spectral response which takes into account both the dichroic lamphouse spectral irradiance and print spectral sensitivity. The masking couplers in the negative emulsion exist to guide color timing under those print responsivities. Now I don't bring this up to make the point that one should try to achieve print spectral responsivities with their scanner, that's very difficult especially with bayer filtration of a DSLR, my point is that high CRI does not get you any closer to the intended trichromatic response of that negative printed. High CRI is useful for slides as the goal for those are just to get close to a projection from a tungsten illuminant, but for negative it's completely irrelevant.
What's actually more important is maximising the bit depth of image dye density above the minimum density composed of the masking couplers. This means neutralising the orange mask with light, not white balancing. Because white balancing is an adjustment of analog gain, post capture. If you scan color negative with white light, more electrons will fill the well depth of the red and green filtered photosites from the orange masking than the blue filtered photosite, this leaves less dynamic range for the blue channel density, when you white balance the orange mask, this adjusts the gain of those channels, it doesn't improve dynamic range. When I say the orange masking should be neutralised with light, I mean that the number of electrons which fill each bayer filtered photosite should be equal when scanning minimum density. Dynamic range matters because for high densities (highlights) they will be stored to very low code values, where differences in density introduce more and more quantisation noise. Given this, cooler bluer light will produce a cleaner scan of color negative compared to high CRI white light.
So what cooler light can you recommend for color negative? I am thinking about using a solux 4700k halogen bulb..this is the lamp inside the great Kodak Pakon F235 scanner. But still not sure how to diffuse it and use it in a practical way for camera scanning. What do you think?...is there a way to replicate the lighting from that scanner?
@@Jason_Red I would recommend that source only if your scanner sensor has a similar response to the CCD used for the F235, because the final trichromatic response is always the product of both your sensors spectral sensitivity multiplied by the backlights spectral power distribution, a significant change in the spectra of either component will not produce the desired response. Most high end scanner manufactures today such as Arri and DFT employ narrowband LEDs for calibration.
The reason why I don't recommend any broad spectrum source is because the orange masking in color negative is stock specific, every stock has a slightly different concentration of masking couplers, so thus require a different timing for correction. If you're really looking to upgrade your scanning setup I would recommend ordering a custom narrowband LED panel, where you're able to select the peak wavelengths of each RGB LED. Where the peak wavelengths you choose correspond to the wavelengths which maximise the density of one dye layer while minimising the density of the others, that achieves the best color separation. You can determine these wavelengths by observing the spectral dye density curves published by Kodak or Fuji for the stock you wish to scan. There are some common trends, such as 460nm for the yellow dye layer, 550nm for the magenta, and 690nm for the cyan.There should be plenty of companies which offer custom solutions, alternatively if you have experience soldering you could order some SMD LEDs from mouser electronics and order a template PCB designed to support pulse width modulation for dimming.
As for commercial options currently available which approach negative scanning with this philosophy of cooler emission, cinestill's CS-LITE backlight does shift to a cooler color temperature for scanning color negatives, this won't offer as much control as a tuneable RGB LED source but it may get you in the right ball park for a budget solution.
@@Jason_Red As for diffusion, opal glass has been the standard for scanners and densitometers for decades. You should be able to purchase some from Edmund Optics.
@@SHDEdits Thanks for your detailed information and sharing your knowledge. What you say about the different color channels feels a bit like color filtration in the darkroom. Makes sense that Cri values are for marketing. It's the same with megapixels. People think that you get higher quality photos with more megapixels.
We should focus more on the color of light and not Cri. The only thing I don't understand why the scans for example from a Hasselblad X5 or Noritsu scanner are so beautiful but don't use different color temperatures?
I am going to try the CS lite. It seems the easiest and cheapest solution for now. I am located in the Netherlands, so ordering from the US is not ideal. Many thanks!
@@Jason_Red Happy to help. As for your question regarding those different scanners, hasselblad essentially offers factory calibration via 3D LUTs per stock from their fixed scanner response employing broad band illumination. This would be similar to the calibration practices of continuous motion picture scanners like telecines as those machines can not intermittently capture RGB flash exposures, so must compromise with broad spectrum illumination and custom 3D calibration LUTs for every stock. As to why they went this route I’m not sure, as they don’t face the same limitations as a telecine, so on paper they’d have to deal with potential limitations in bit depth however they may have chosen a specific broadband source alongside their scanner sensor similar to that Kodak scanner for minimising noise. This unfortunately is left to speculation.
As for noritsu and frontier, it’s a completely different story, they both employ narrowband RGB LEDs with peak wavelengths ideal for maximising color separation which would be dimmable for neutralising the orange masking of any negative stock in real time as described above. Both noritsu and frontier also employ IR LEDs in their boards to obtain a pass of the film where only the density of dust is present as image and masking dyes do not absorb infrared light. The noritsu employs a line array for their LEDs and frontier an area array. The frontier SP-3000 has peak emission in the 440nm to 460nm range for blue, 535nm to 545nm for green and 640nm to 660nm for red. Some scanner manufacturers choose shorter wavelengths than 690nm when employing a Bayer filter sensor as sensitivity in that range is typically lower.
You haven’t ‘scanned’ anything. A camera uses an instant capture sensor not a scanning array. You are ‘DIGITISING’ film frames!
protip: great photo will be great when scanned with the cheapest chinese "phone scanner". shit snapshit will be shit when drum scanned in a vacuum using arcane aztec mirrors bouncing sunlight from antarctica.
if your photos are worthwhile you can get away with canoscan 7700 for what? $100?
dont let this pixelpeeping bullshit take away fun from the hobby, dude is overanalysing sky color when the photo itself isnt even worth taking, let alone scanning. too much magenta? there is this wonderful program called lightroom.
fuck, the colors will change depending on the camera brand, home scanning is the worst option for the film "purist", get a RA4 processor, some chemistry, set up a darkroom and THEN you can fiddle with the colors. he's using NLP ffs, which by itself fucks with color balance to simulate lab scanners.
to anyone reading this interested in film photography- dont listen to crap like this. you cant buy your way into being good, as this video shows that perfectly.
Nice guy! 😁
@@pushingfilm tell me im wrong. look at people taking great pictures lately using iphone 3gs.
chasing that last one CRI is just silly when the photos present nothing of value. i understand that we've all gone through the phase of "i'm not shit, the gear must be shit" but really, is that photo at 9:00 really going to be better when all the colors are *just* right? and what is "right" anyway? some color cast? do a test shot of a white wall, correct the cast, paste onto all other scans, there, i've saved you hundreds of dollars. when scanning we loose the "analog" quality, its now a digital picture and pretending that you cant correct the casting because it's not "pure" is just silly. Nate changes the whole engine of NLP between versions and the results are drastically different. Are you going to unconvert all the scans and invert them again?