Re-creating Disney's sodium vapor process [video]
youtube.com
youtube.com
There have been a few art installations over the years that use these lights:
https://www.youtube.com/watch?v=hd077pa-5CI
For a second I thought that the video was from this channel, which also does experimental cinematography stuff:
It's so weird, as those people say, so see things coloured... yet monochrome.
At least by the time they got round to doing my street they used a lower temperature white than the ultra-harsh 'prison yard white' colour used in the earlier installations.
I am continuously amazed at the level of innovation done at the Disney studios back when Walt Disney was still alive. Rewatching Snow White with my daughter a few months ago, I was flabbergasted by the animation quality, even when compared to animation movies made decades later. Just look at these water scenes:
https://youtu.be/54QeNL5ih6A?si=dOdrRrISu8f4I96G&t=95
https://youtu.be/khmrr7-W6BA?si=whn5mvuL9x95JP1l&t=70
This was in 1937! Compare these scenes to this Paramount animation clip (also featuring Snow White), which was finished just 4 years earlier: https://www.youtube.com/watch?v=cKOSJ5AAwfc
The multiplane camera [0] is also ingenious and is imho the single biggest reason why Disney could produce animation movies that looked "cinematic". I read somewhere that they re-used it for "The Little Mermaid" in the 80ies, and nobody really remembered how it worked anymore.
Disney also developed Xerox animation [1], which is cheaper than traditional animation, but which gave 101 Dalmatians a very nice "sketchy", artsy look .
[0] https://en.wikipedia.org/wiki/Multiplane_camera
[1] https://www.smithsonianmag.com/innovation/how-one-hundred-an...
Disney was known to recycle some of their animations in older movies [1], so this could be because they have a source live action recording somewhere that they rotoscoped over more than once for different productions. Also note the different backgrounds with different characters that move - excellent example of the multiplane camera in action. In the foreground they have a "recycled" animation, in the background they can then put something else in, and record it all in one go with their special camera.
> Very few of the animators at the Disney studio had had artistic training (most had been newspaper cartoonists); among these few was Grim Natwick, who had trained in Europe. The animator's success in designing and animating Betty Boop for Fleischer Studios showed an understanding of human female anatomy and, when Walt Disney hired Natwick, he was given female characters to animate almost exclusively. Attempts to animate Persephone, the female lead of The Goddess of Spring, had proved largely unsuccessful; Natwick's animation of the heroine in Cookie Carnival showed greater promise, and the animator was eventually given the task of animating Snow White herself. Though live action footage of Snow White, the Prince and the Queen was shot as reference for the animators, the artists' animators disapproved of rotoscoping, considering it to hinder the production of effective caricature. Nevertheless, all of the above-mentioned characters were fully rotoscoped and utilized by their respective artists, some more, some less.[98] Despite Graham and Natwick's objections, however, some scenes of Snow White and the Prince were directly traced from the live-action footage.
https://en.wikipedia.org/wiki/Snow_White_and_the_Seven_Dwarf...
Cool! Training for what? You can't just drop a juicy comment like this and then leave us hanging ;-)
I'm guessing the answer is "they do, it's just uncommon and you haven't heard about it"
Not with a non-visible spectrum, but along the same lines. I can't seem to directly find anything with NIR though;
It seems possible to produce a four-channel camera sensor specifically for this use case, if it was popular enough. Given how common green screen is in Hollywood I'm surprised I haven't heard of anyone doing it. Maybe Hollywood just doesn't care because they can do regular green screen and just hire someone to fix it in post.
I am guessing that that is dangerous to human health because infrared fails to elicit the flinch reaction that too-strong visible light does.
That's one of the things that makes it suitable for thermal nano cancer therapies. You
(1) Create a nanoparticle with a gold core of the right size that it excites when exposed to near-infrared radiation (the far end of "near" is ideal),
(2) Coat it with something that's hopefully less toxic than the cancer because the gold nanoparticles are pretty reactive (silica was popular last I checked, though I'm not convinced that's actually safe, and there hasn't been much testing on it),
(3) Coat that in something that binds to the right antigens,
(4) Inject that into the patient so that you'll eventually have a tumor rich in these nanoparticles while the rest of the body has a very low concentration,
(5) Shine an intense near-infrared beam at the nanoparticles. It safely penetrates body and causes minimal heating, instead depositing all its energy at the tumor, selectively cooking just the bits of tissue you don't want anymore.
It's yet another way to make mice immortal, but human testing is a long ways off. The hard parts are (2) and (3), along with the (4a) I didn't mention where most of these things don't want to stay "nano" in most chemical environments, and the solution they're suspended in is usually also not ideally suited to being injected into living mammals who would like to retain that "living" property.
Glowing-hot steel produces enough IR that the workers that work with it have 7 times the rate of cataracts as the general population.
NIR is much safer than the visible spectrum at the same power levels. The threshold for a few minutes of high-intensity NIR increasing your chance of eventual cataracts is staring at an NIR source 10x brighter (w/cm^-1) than the sun integrated across all its wavelengths. The threshold for sun-colored wavelengths causing eventual cataracts above the baseline from a few minutes of staring is under 1x.
Also worth noting, very very near IR (basically red, though we can't see it) doesn't quite enjoy those same properties, and during their time working with hot metal you'd expect a lot of energy in that band.
Also worth noting, many steel workers are exposed to dangerous amounts of visible light too. I absolutely believe that they can get enough NIR to cause problems, but if I wanted to try to prove that NIR specifically causes their problems to somebody else then I'd want to try to account for that fact (and for incidental welder exposure, ...).
Another alternative I was thinking about is just to use one of those monochrome astrophotography cameras, something along the lines of this[3]. The monochrome version doesn't have the IR-cut filter and sensitivity is pretty ok down into near-IR.
edit: The IMX492 sensor in that one has a 40% response at 850nm, and 850nm IR diodes are plentiful and shouldn't emit[4] much at all below the 700nm of the IR-cut filter the color camera should have.
Not sure how mixing sensor sizes and such would affect things, or if it's better/easier to just run two of the same camera.
[1]: https://s1-dl.theimagingsource.com/api/2.5/packages/publicat...
[2]: https://www.lifepixel.com/
[3]: https://www.zwoastro.com/product/asi294mm-mc/
[4]: https://lumileds.com/wp-content/uploads/files/DS191-luxeon-i...
Building 2 nigh identical cameras, but one with a sensor outside of the frequency range of the other, is probably a very expensive custom job.
It seems like the cube with mounts is the most expensive part, and there's no way it can cost more than one of your cameras, let alone the lens, right?
Now that I think about it, Disney must have done something pretty special for the lighting in the foreground in their version of the process. They didn't have single wavelength LEDs back then. I wonder how they did it?
They did that to get a 'perfect' mask.
If you use infrared, which is generated by random environmental objects, how would you prevent random spill?
Practically (my understanding) is if you got random spill on the surface, you'd get random alpha in the mask.
The point of using the sodium light is that it's a single wavelength that you really really would not get naturally, so you can totally control the emission sources of it.
Could you do that with a single frequency infrared? I guess it's possible, but I'm not sure how you'd do it technically, and it's unclear what the benefit of using a different frequency would be?
It may be generated by lights but could be filtered out. It is generated by the Sun, but so is the sodium wavelength. The technique would work indoors just as well as the sodium one, but with important advantages. I think it could even work pretty well outdoors if you used a wavelength that is mostly filtered out of sunlight by the atmosphere, or maybe two different wavelengths in two channels simultaneously.
> but with important advantages
What's the benefit over using any other single frequency visible light?
Even if the atmosphere absorbs most of a particular band, if any of it gets through, your mask is going to be messy, which kind of ruins the point; I don't know anything can really fix that for outdoor purposes?
I mean, broadly speaking, all you need is a way to emit a sharp frequency band and avoid having any environmental pollution in the same band.
If there's a good way to do that in near-IR, it would work; but I feel like the same could be said about pretty much any frequency.
I'm not aware of anything specific about near-IR that makes it particularly attractive.
(but, clearly, I'm no expert in near-IR)
> It would allow using full spectrum lighting, and would be easier to selectively filter.
Making a broad spectrum IR filter is far easier than a notch filter for a single visible wavelength. In general it's harder to make more selective filters and an IR filter can be much less selective because the IR band is so large and far away from any visible light.
Also the infrared version would allow using any color background including black, and the on set lighting would appear to the eye exactly as it does to the camera. With the sodium vapor process the bright yellow background would make everything on set look different in person because the eye wouldn't be able to subtract it the way the camera does.
Yes, you would need a filter on any incandescent light. But this is also true of the sodium vapor process. Actually I'm not sure anyone makes a sodium vapor wavelength notch filter suitable for putting in front of an incandescent light. Consequently, as seen in the video, they restricted themselves to special LED lights that emit essentially only three wavelengths. These are very much not full spectrum and will not produce natural colors for some materials. And they had to completely block all windows, whereas for IR you could just install some IR blocking film on the windows and still use natural sunlight.
You're basically proposing something completely different here.
So... while I agree you could do something as you describe, and I'd be interested to see the results of it, I'm not sure what the results would be like.
> The advantages I stated were
>> It would allow using full spectrum lighting, and would be easier to selectively filter.
All I can say is the CC results in that video were pretty great, despite the obvious caveats, compared to green-screen.
If you (or anyone) can build on it to do better, people will be interested in it; but there's obviously a difference between us idly speculating on how/if it might work, and this, where they've actually done it and shown that it does work.
> "they do, it's just uncommon and you haven't heard about it"
I've never heard of it being done.
If anyone has, please post a link or something.
Yes, Corridor's results were great. Fewer limitations would make the process even better!
Alternatively you could use a solid state laser or narrow band LEDs if you need the band to isolate to be in the visible spectrum too.
One advantage might be to pulse the IR and detect it via that, if the detectors are sufficiently fast. Maybe have a filter that's pulsed also (Pockels cell?)
My guess though is that using a visible wavelength makes it easier to set up the lighting and ensure it's even and so on (remember, this was before electronic technologies that would make it possible to live-preview the result). Also, sodium lights and sodium band filters had other uses in industry so they might have been more available than NIR/NUV equivalents.
I know in IR still photography you're supposed to add a focus offset marked on your lens to adjust for that. A regular achromatic lens quickly fails outside of the wavelength range it was designed for[1].
[1] https://en.wikipedia.org/wiki/Chromatic_aberration#Minimizat...
EDIT: actually, since it's outside of the visible spectrum, typical lens design correcting longitudinal chromatic aberration and doing the spherical to rectlinear projection simply might be off, resulting in soft edged and general artifacts unwanted in a clean mask
However, over time these LPS streetlights started to be replaced by Induction lighting. And as of 2016, the new standard is LED based, which is supposedly even better for light pollution [3].
In areas where sea turtles nest, LPS lighting is still preferred, as it is very different from the white light from the moon they use to navigate [4].
One nice consequence for residents: as a result of the strict light pollution regulation, and the consistently clear skies, most nights one can see a surprisingly large number of stars in the sky for an urban city environment. Nothing close to a clear night far north or out in the desert, but still more stars than you might expect.
---
You can read more about it here:
[1] https://www.sandiegocounty.gov/content/dam/sdc/pds/docs/pds2...
[2] https://www.sandiegocounty.gov/pds/docs/LightPollutionCode.p...
[3] https://www.sandiego.gov/sites/default/files/12-14-16-adopti...
[4] https://conserveturtles.org/information-sea-turtles-threats-...
I'm curious if this means anything in terms of good vs perfect technology or what tech wins out in practice. This process seems just plain better than green-screen, so why did it lose out? Is is just that green screen is good enough? Was it too hard to do this sodium vapor process with random cameras back in the day? Are sodium vapor lights bad enough in some way that practically it lost out?
But it seems so much better that that shouldn't have been a blocker. Like optics isn't magic, and knowing that someone did it before should be enough incentive? Not like movies are a small or poor industry.
"Sorry boss, that's industry best practice. Nothing we can do to make sit better."
Keep in mind stuff gets lost all the time, even really important things like nuclear weapon components like https://en.wikipedia.org/wiki/Fogbank
The movie industry had cheaper options, so they went with those.
Fair on the number of films/cameras.
Is normal keying actually more tolerant to spill or is it just worse so you're going to have to manually deal with spill anyway so you just live with it? I don't even have good eyes and even I sometimes see spill in actual movies, or especially in TV. Edit: I guess that's _different_ spill though.
I mean you do famously need a special screen :) And you don't need _special_ lamps per se but you do have to light your greenscreen fairly well to have real success.
It looks that the problem was manufacturing the prism, as simple as that.
Today you can use unusual wavelength LEDs without phosphorous to do the same thing probably way cheaper.
LEDs are 10~50nm, untuned laser diodes 1~10nm. It is my understanding that it's easy to make a filter for the camera that very selectively targets this line pair, though it may entail mild optical complications as a sub-nm pass/reject bandwidth detunes from sideways angles, so lenses may need to collimate the light first before feeding the filter, and de-collimate again after.
The sodium vapor light technique allowed for a very easily generated matte with a high level of detail. It however required specific lights and a lot of control over spillage from the background lighting into the foreground. It was also difficult to film from multiple angles since the lighting would need to be adjusted when the camera was repositioned. With those limitations in mind it was much cheaper and less time consuming than making mattes with multiple exposures and such. Cheaper effect shots meant more effect shots for the same size budget. With a movie like Mary Poppins that wanted a ton of effects shots the traditional method of making mattes would have been cost prohibitive.
Chroma keys came about because they were much cheaper than the multiple exposure method of doing mattes. The back and foreground could be lit together and the matte could be knocked out relatively easily by applying a chroma filter when making the work print. The mattes weren't as exact and had their own limitations but they were relatively inexpensive. For video (as opposed to film) chroma keys could be done electrically by filtering and mixing the video signal.
Digital mattes have a lot of flexibility. As shown in the video digital editing software has lots of toggles to help make chroma mattes look good. It's a very manual process and hard to really get right. Also as shown in the video there's some things that just do not work really well with chroma keying like translucent materials and anything with high index of refraction. They're typically a good balance of cheap and effective with a lot of known techniques to make look good.
The video suggests that it was difficult to replicate and extremely expensive.
For example, Niko doesn't seem to be aware of the IBK keyer (based on his comments in the VFX subreddit), which is the keyer used on almost every shot in high-budget movies. While it's not a single-click solution, it's the main tool to tackle thousands of shots with defocused edges, motion blur, and semitransparency every year.
Niko deserves credit for bringing Paul's incredible experiment, but Corridor is at its core a group of entertainers, and most of them have never had experience in a large production/VFX environment. Take everything they present with a grain of salt.
I think there are a lot of analogies here to find related to the software industry, such as the relationships between closed source and open source software.
I think this is the reason they're using the Z Cam (presumably E2) cameras visible at 2:48, 3:31, 4:05, & 11:05 because that's one of the most inexpensive genlockable cameras available. This would rule out most mirrorless and prosumer video cameras unless the subject isn't moving fast or a half-frame misalignment is acceptable. Heck, a non-genlocked version of this might be preferable to weird motion blur on existing chroma keyed backgrounds. Very exciting tech
Same guy that made sponsorblock; crowdsourced titles and thumbnails. Tends to have good coverage on anything reasonably popular. Also has features to reduce the clickbaitiness of videos that haven't had someone submit better titles/thumbnails.
It's just a dichroic beamsplitter, which is a standard optical element available from online sellers. Variants of these are commonly sued in scientific experiments, optical devices, etc...
Here's one for under $300: https://www.meetoptics.com/beamsplitters/plate/longpass-dich...
Just search for: "589 nm beamsplitter".
However, I'd guess these are the components the guys are using between the prism and the cameras.
I am wondering if part of the reason that finding these at the frequency of sodium vapour lamps is partly due to those lights no longer being common. Or perhaps that those just aren't frequencies common to laser optics, which seems like the biggest market for these kinds of devices anyway.
Though, I see no reason why you couldn't use a laser for this process anyway, with matching notch filter.
https://www.edmundoptics.com/p/589nm-cwl-125mm-dia-hard-coat...
They had to do a test and then process the film for each adjustment. A extremely tedious process and that doesn't include the work that was required to produce that prism.
Or even magenta ones with different filters.