Frame & Reference Podcast - 224: "Knives Out 3" Cinematographer Steve Yedlin, ASC

Episode Date: December 25, 2025

MERRY CHRISTMAS! Joining me for the SEASON 5 FINALE of Frame & Reference is none other than Steve Yedlin, ASC to talk about his work on the new film Wake Up Dead Man: A Knives Out Mystery, as well as all the #NerdyFilmTechStuff your heart desires.We'll be back in January to kick off Season 6 of F&R with another incredible episode, so until then: enjoy the holidays, enjoy the New Year, and we'll talk soon!► ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠F&R Online ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠► ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Support F&R⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠► ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Watch on YouTube⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Produced by Kenny McMillan► ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Website⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ► ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Instagram⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Hello, and welcome to this, the season five finale of frame and reference. We've got a great one today with Steve Yedlin, ASC, the DP of the Knives Out series most recently, Knives Out, three wake-up Dead Man. And I just wanted to quickly say, thank you so much for spending your time with me over the past year. It's been so much fun. I love how many amazing guests we've been able to have like Steve. And it only keeps going when people like you, you know,

listen and share it with your friends and all that kind of fun stuff. So thanks again for listening. We will be back in a couple weeks, two, three weeks to start season six. we've already got five episodes in the chamber so yeah the train isn't stopping now but just wanted to share that thanks with you all and I will see you or you will listen to me or however this works in 2026 enjoy you know what actually I was just hanging out with Alex Forsythe a couple nights ago he was like oh tell him I say hi that's great well hello I imagine you guys all I guess he did mention the

Hollywood Beer Association I don't know if you go to those I have not right I'm not frequented it but I know about it yes yeah yeah he's a he's a fun one to sit there and pick the brain on because you know at least at least your stuff I'm I'm able to go back and revisit when he's telling it to me I'm like bro here's a sheet of paper I need like study materials yeah yeah it can be overwhelming it can't but it's fun to learn you know especially like you spend so much time focused on at least at my level like trying to get better and better and then at a certain point you're like well I'm getting hired the filming enough people like my work and then you're like oh there's a whole new level of stuff you can learn I'm like oh that's

that's nice you know I hate feeling like I know quote unquote everything you know oh yeah no yeah if you feel like you know everything something's wrong because there's always a lot to learn yeah the uh they actually they sent me to it's nice sometimes i'll get sent to screener but they actually sent me to the netflix um office to watch it in there oh great oh good i'm glad you got to see it on see it on a proper projector i imagine theirs is probably calibrated to the uh whatever the correct spec is you know for you know it's no dim bulbs there but I did I found it I really enjoyed it there were a few

sections that I was looking at that I I guess this is just the only place to start that makes sense but it's just well I guess first of all like obviously when did the first one come out 2019 2020 I think 2019 we shot it in 2018 so it must have come out in 2019 I think right and you then you've had obviously a lot of time to learn more stuff yourself or whatever.

Was there, not that they look terrifically different, but are there any like methodologies or processes that you kind of added to your toolkit to help with this one's from the first one? I mean,

on the one hand, absolutely yes, but on the other hand, I wouldn't attribute the look to those. It's more of a, you know, tools to get the look that were after. which we would have been after the different look even without the different walls. But yes, absolutely. And, you know, the biggest thing that's changed since then is, you know,

in the realm of light control because, you know, I've been really interested in, you know, there's sort of been a promise of LEDs that hasn't really been. been delivered. And I'm trying to, you know, work on tools that sort of deliver the promise that was always supposed to be there. Because, you know, it's a little bit like what happened when we first started moving from film to digital image acquisition, where people got so excited that they can do anything. Like, there's all these knobs and different things you can do. And you're like, yeah, but the fact that you can do anything doesn't know how you, doesn't mean that you know how to do even the one thing that you're trying to do, let alone anything else. Like, there used to be

this in-place system, you know, there were only a few films, you know, camera stocks or only a few print stocks. You know, you could process normal or you could push or pull and that's it. And it was this kind of system that worked and made an artful color rendering, not photometrically or colormetrically correct, but an artful one. And it did it repeatedly and reliably. You know, And then that locked system didn't just get a little more flexible. It basically went in the trash can and we now have a completely unlocked system. So people felt like that. I think there was a sense, a general sense of that's powerful.

But the problem is when you're trying to make a movie and not reinvent the wheel of the technical stuff while you're doing it every single shot or every day or whatever, it's good to have a system that works. So to me, the idea, you want the best of both worlds. You don't want a baby with a bathwater situation where everything that was good no longer exists. You know, like the idea that you can have a, with that, the idea, you know, with that one, the idea that you could have a fixed system that's awesome, but you can also change it however you want. Rather than in order to not have a fixed system, it also means you don't even have a starting place. You know, so, you know, we've kind of, depending on. what kind of color pipeline, you know, because there's so many different ways people do it.

So depending on how you do it, you know, that part has been short up now, and it's possible if you want to, not everybody does it, but it's possible if you want to, to do it like the past where there's this completely fixed, predictable, simple, reliable workflow, but you can also vary it however you want. And then I think when we came in with LED lighting, it was sort of the same thing where people got so excited that you could make the light any color that they forgot that they don't even know how to make it one color. Like how do you make this, how do you make this light the same one color that incandescent lights always work? And, and, you know, so there's this kind of like, you know, like all the variables coming untethered, whereas what you actually

want is you want all of the reliability that you had in the past plus the ability to change it. You don't want, you know, you don't, you don't want, again, the same thing. You don't want to this baby with the bathwater situation. So, you know, when we did knives out, I was already working on that, and we were kind of in the, in the very early stages of it where, you know, we were using, you know, at that point I had developed a little color calculator where, you know, if we're using, so like, let's say, so let's say we've got an incandescent light, a professional LED movie light, and then a off-the-shelf, you know, just cheap LED ribbon that doesn't have all of the colorometric stuff that the movie light has. Well, we can match the movie light to the incandescent

light by, you know, taking a spectrometer, read the incandescent light, and then put those chromaticity coordinates into the movie light. That's already a little bit more advanced. than most people do, because sometimes people assume that because we offhandedly call incandescent lights 3,200, they think that means you can put it in color temperature mode and set it to 3200 and have it look like an incandescent light, which most of the lights don't, because the actual, there's a color science definition, there's a hard definition of what chromaticity, because we know how to measure hard chromaticities, and the chromaticity that's associated with 3,200 Kelvin is actually not exactly the same.

same as incandescent lights. That's an offhanded, you know, that's an approximate offhanded description. And also it's describing, it's actually more describing what the film is sensitive to more than what the light is, whether the light is making the same color that the film is sensitive to. So it's already a little bit more advanced to meter the light and then set the thing to that chromaticity rather than using sort of colloquial versions of color temperature. because there are actual color science versions of color temperature, but there's also the sort of colloquial non-specific ones, like just calling an incandescent light 3200

or calling daylight 5,600, which is really crazy because daylight's a lot of different. It's a lot of different colors. Yeah, it's the film that's 5600, not the daylight itself that's 5600. So it's already a little bit more advanced to do it that way with the smart movie light, but then how do you, do it with the dumb light that doesn't have anything, the dumb light doesn't let you type chromaticity cord and fits into it all you can do. It's got five emitters on it, red, red, green, blue,

cool white and warm white, and you can just turn them all up and down. So we, you know, I had developed a calculator where if you take spectral, you know, spectrometry readings on the five separate illuminants, it'll make whatever exact color you want. And it'll also do it with the brightest, smoothest spectral distribution possible, where the whites are as high as they can be. So it's using the broader, brighter illuminants to make it as much as it can. If you start making a saturated color,

it's got to turn those down. So on Knives Out, we were already kind of going down that road. But now, like all of that stuff, because we have our custom software that we do this stuff with, and it's much more advanced now, to where rather than typing it into a calculator, and then having to manually set it. Like the calculator says,

I need to set the five illuminants on the LED ribbon to be this, this, this, and this. And you're like typing them in. Now the calculator still does it, but then it also actually implements it in the background. So all you're typing in software is the chromaticity coordinate and just in the background, it's just sending the right blend to the light.

And we're making dumb lights into smart lights. And now it doesn't matter. We can match every, you know, we can have, we can have not only different brands of movie lights mixed together, but we can also have, you know, dumb, dumb lights that don't have all of the, all of the, you know, the high-end smart stuff that the movie lights have. And so are you driving, like how, what is the interface between, for instance, just some off-the-shelf LED ribbon? Is it like a raspberry pie kind of situation?

or like, how are you getting that information to and from the light strip, for instance? Right. So, yeah, so we have our sort of custom software version that I've been using that's basically instead of a dimmer board, it's not a dimmer board. It's like instead of dimmer board software,

we're basically controlling all of the lights, the whole stage, everything from this system. We're actually working on, a partner and I are working on. This is, eventually this is going to be put into a beautiful shiny box and it's going to be a product that people can use publicly. But right now,

that's still in progress. This is like a Frankenstein, you know, sort of prototypey version of it that I use. Which producers would love to hear, by the way. Yeah, I built this in my shed. They're like, yeah.

Well, I mean, the reality is we've been using it for a bunch of movies, now and and and you know everybody that's seen at work sees that it's it's saving us time it's making things better you know it's it's everything's better about everything from either you're saving time or you're using the time much better where you're always finessing you're instead of you know instead of troubleshooting to even get to the place you need to get to you get there right away and then using the time to make it even better um and uh so so So even though it's the prototype version, I mean, we've been using it for a while.

And so it's just instead of the, it's basically instead of the, the, the dimmer board console software. So like the lighting programmer might be, you know, they might pre-rig with their system and then I'll unplug their system and I'll plug their system and I'll plug my system in. And then just in terms of driving those LED lights at the end, it's really just that they have controllers that take DMX. the same way that the smart lights do, but the only thing that you can do with the DMX controller is turn the five illuminants up and down. You can't, you know, it doesn't have any chromaticity information in it. Like, it's not going to take a signal. You're just literally driving the lights. It's not going to interpret a signal where you're saying, hey, give me something like this, and then it figures out how to do it. You have to figure out how to do it. But now it's

automated where, so it's still on the control side, as opposed to the, the receiving side is where the calculation from chromaticity to physical blend of emitter happens. So that's now happening on the control side, but you don't have to do it manually as a human, the thing's just doing it under the, under the hood. And so when you say off the shelf, you mean like light ribbon, not Walmart thing on the spool? I mean, the thing is it actually could be, right? Because the, I mean, we're not necessarily using that, but it absolutely could be. We'd have to get a.

spectrometry data set on that specific ribbon you know if you use the data set from another ribbon it's going to be approximate at best um but once you have it you can do it with them because the the the ribbon itself is just has the i mean it just has wires on it right so you could plug the wires into one of the DMX driver you know and yeah so so yeah you could you could do it with you could conceivably do it with any anything like even if something's not meant to you could strip off whatever it's connected to and just push put the wires on to the TMX truck. If it's the right voltage, don't plug out, don't, people at home don't plug the 24 volt

one into the 12 volt. Don't put the 12 volt one on the 24 volt power supply, but, you know. Yeah. Yeah, don't try this at home. You actually turned me out years ago when I was reading stuff you'd put online, you turned me on to just the, like I had to interpret it, obviously, because at the time I was a little dumber, but I invested in getting a, you know, the spectrometer and, boy, like, just being able, because I do a lot of, I do a lot of more documentary stuff now, but at the time

it was a lot of corporate as well. Same, same idea, you know, looking at people, just interviews. And being able to just plop someone next to a window and put, I have like the keynote flow and just X, Y it in there and boom, it all just, I'm like, oh, it's the greatest thing in the world. I love to. I wrote so many articles. I write for this website called Pro Video Coalition. And the second I got it, I went and metered every bulb in my house.

You know, I was like, here's, and I wrote an article. I was like, you're going to be using these for, you know, film use because we can't get tungsten here and count. Yeah. Here's the good ones and the bad ones. They're all bad. But GE Sunfield was actually spectrally pretty good. But yeah, now, now I can't like.

Because exactly what you're saying, the big thing that pissed me off was like, you typed 5,600 in the back of the, that's not what comes out the front, ever. So you got to do it and then meter it and go, all right, what, 100 more and then you move it 100 and it went up 300 and you're like, all right, somewhere in the middle. Well, well, well, worse than that, though, I mean, even if it did, that's still not the color you actually want. You know, like the fact that some, even if something is 5,600 and, you know, zero, you know, in hard chromaticity, there's no such thing as tint, but there's delta UV, which is how far off of the plonk in locust you are. So if you, if something is actually in color science terms, not in dumb cinema dude terms, if it's 5,600 and zero, that's not really going to be the color of anything. You know, like it might be vaguely close to, you know, like if it's a, if it's kind of a sunny day and you're combining the sun plus the sky, not one or the other, and you don't have a lot of buildings. Yeah, but I'm saying that one thing it might be close to is if you just have sun and sky,

so it's not overcast making it super cool, it's not just the sky, it's not just the sun. If you have sun and sky and you don't have green bounce from trees and stuff, that might kind of, sort of be a little close to 5600 is still going to be greener probably but nothing that you would ever be matching unless it's a weird coincidence is going to be these you know these 3,200 or

5600 numbers and it's almost never going to be right on the Planckian locust most things most things that are semi either natural or you know sort of analog tend to be green they're you know on the green

side of the Planckian locust and you know something like an old LED might be, you know, before they started making proper color ones might be on the magenta side or whatever. But yeah, yeah, almost nothing's ever going to be those, those. Yeah, so it's like a series of problems with the Kelvin thing, like that we use it colloquially. It does have a mathematical meaning, but that's not what we mean. Even when you do use the mathematical meaning correctly, that's not usually the colors you're after. It can't properly describe anything that's not on that line you know we have right there's the there's the

whole color gamut and you can only describe things on one line in the whole in the whole gamut but the whole observer like the standard observer um not only can you not go above or be below the line green magenta you also can't go past it you can't go bluer than you know the where the line ends on the left um so you know there there's all there's like all of those problems with it. And then on top of that, it's a spectacularly non-linear scale compared to both perception and to like physical blending of illuminants to where, you know, with the low numbers, 100 Kelvin is a lot. Like you can totally see it. It's not just a blip, right? But if you're at 10,000 Kelvin, 100 is literally nothing. Like if you're 10,000, you could go to 20,000 and you're not going to see a change

because there's a point where it's piling into a point and not moving anymore. But, you know, like I know multiple real examples of like a gaffer will have their lighting programmer make presets at 100 Kelvin increments. Right. And like, this is, this is, you're making way too many. You don't need anywhere near that many,

and yet it's also too coarse. Like you have the opposite problems because you've made you made probably a hundred times or whatever more than you need near the high numbers and you don't even have enough on the low numbers to be even sort of perceptually even steps yeah yeah well you know that that lighting guy probably got so many dumb notes where they're like a little more a little more so the difference and their brain just went you know what 100 increments every time yeah whatever you want bro you know yeah yeah the the whole like At what point do you think we were able to actually, when you talk about throwing away the pipeline and just creating a chaotic, you know, do anything you want, how recently were we able to actually take authorship of that and create, as you tend to say, like your own thing, not an ascribed thing, but you're like this way, I like it this way.

Because when the fools were shittier before, you were kind of at the behest of, you know, especially when we went from like tape to digital, there was like, like four buttons on those things, you know. No slick. And no choices. But was that kind of a recent development that you've been like having to fight your way through? Or are we at a place now where the average person doesn't have to, you know, sit and nuke and figure it out? That's an interesting question. Wait, are you talking about the LED stuff or the image pipeline stuff?

But specifically just the imaging pipeline, you know, from kind of start to finish. Yeah, I think, well, I mean, I mean, nobody ever. had like had to sit in nuke like i did i mean that's a nerdy way to to build your own stuff because because you don't have to your tax isn't nuke right you're yeah exactly exactly yeah um uh that that is actually a line that i say that you could do because it's it is it is it is it is bit bucket when i teach a seminar at a fi and i actually say one of the reasons we're using nuke is because it's agnostic it's not doing a bunch of automated stuff that you don't know what it is and, you know, I literally say you could do your taxes

and would not be a good way to do it, but you actually could. But, yeah, I think it's kind of a, I think it's kind of complicated because I think it's, yeah, I mean, the one thing to remember is you never have to do it by yourself, right? Like you could do a locked system or somebody's helping you, right? like it's not um it's not one or the other where like um like let's just say a camera let's just say a camera is really good quality in terms of latitude noise you know resolution

all the stuff you want but you just don't like the manufacturer's lead it looks too video maybe it looks more filmy than some other lets but it's too videoy looking and you want a more cinematic, artful, less colorometrically accurate and more artful look. Yeah, exactly. If you want something like that, that doesn't mean you have to invent it yourself. I mean, I do because I'm a nerd on that stuff. But that doesn't mean, but if you want a more fixed down system, you know, because, you know, like we were talking about and like you were just saying again, that it used to be fixed and worked, like, when it was just film print system, then it got completely untethered where it's not like,

okay, there's a fixed system that works and you're allowed to go off of it, there just is no system. So if you're going to have a fixed system that works, you have to decide what that is for, you know, kind of project-based. And I think it's absolutely possible to do it. And it actually has been, you know, if you don't want to make the thing yourself, if you want to partner with the color scientist at your post house, or, you know, there's plus. plugins that do some of this stuff or, you know, whatever. You made one. I was involved.

If you're talking about Genesis, I actually didn't make that. I actually didn't make that. Colin Kelly made that, but I licensed some of my stuff to them. But, you know, whatever it is, yeah. And, I mean, they did all the color stuff. That's not none of my color transformation stuff and that. It's great. I mean, it's good stuff.

I'm just doing it out. It looks great. Yeah. But yeah, so like there's all kinds of things that you can do. And I think part of it is actually the philosophy more than the technology. Like you have to actually want to do it, not just say that you do, because there's a lot of these things where people say they want something. It's not really what they want, you know. So you have to actually want to do it and be committed to it. And then you really have to bet it because, you know, one of the things that happens, and, you know, And the thing is that this was happening sort of when the Alexa was new. So I would have imagined it would be not happening anymore by now, but I know stories where it is still happening, where, you know, if you want to use, if you do like the manufacturers, let, that's great.

There's not, you know, the different tastes, whatever, but there's the problem where people aren't really vetting it exactly right. So they're using it to make all their onset decisions about lighting ratios and stuff. And then it's not really what they want. So then when they get into the color grade, you know, one of the hallmarks of, and this is stereotypical, but it's a very good, broad thing is it looks more video-y, you know, like video-e as opposed to cinematic. It looks more video-e if, you know, okay, something that's black is black, but then as soon as there's any light on it, it rises up really fast, right? So it's, so the middle is very flat. Like as soon as there is light on something, it's very, it's very, visible whereas there's a tradition in cinema of there's all of this stuff that you can see it's totally visible but it's dark and hanging low in the frame and that way like when you light a person they stand out from the background and there's you know there's texture to the if they have a bright

side and a dark side of their face that that actually feels textural as opposed to um you know just feeling kind of flat and you know if if when you're you're you know when you're shooting you're not designing the whole pipeline. That's the whole thing I'm saying is you can't be reinventing wheel. It's like this is the only time you've got the camera, the actors, the location. You know, you have to be all in on the actual lighting and setting up the camera. You can't be like inventing Lutz on set every day or whatever. And, you know, so if you sort of just didn't think about that and didn't notice it, you were thinking about the stuff you should be thinking about on set, but then you were lighting to this Lut that wasn't vetted in advance, you know, you may have now

exposed it where you know you were sort of trying to get that richness because if you have good taste and you wanted it to you know not you wanted it to feel more cinematic and everything was looking bright and flats you actually just like you were putting negative fill on and exposing darker to try to get some texture in there but then you realize oh wait a minute it was the the lot now I exposed it so dark and I've put the you know that there's there aren't I don't have all of those tones because in that traditional cinematic look you do have all this stuff that's that's dark but visible it's not going black so if so when you try to bring it back to that you just never shot the information because now because you were trying to get some texture into the thing but

but doing all of your evaluation for that through a lot that's you know contrasty at the edges and flat in the middle compared to what you actually want you know you you've now put these tones so close together where, like, for example, the very darkest thing in the shot is very close in tone in the actual data to like something that you want to be medium or just below, you know, like middle gray. So now you, you know, now you have a choice you can, the whole image can kind of look milked or it can look torn apart because you're trying to take two tones that are very close together and rip them apart where you're going to see the quantization, you know, the image is actually going to fall apart. You can do one of the one or the other, but you can't do both. You can't get it to be where, you know, black is black. The stuff that you want to see, you can still see it and other things hang low and have nuance to them. It's like you can, you can rip it apart or you can milk it out, but you can't, you can't get darkness and the richness and the contrast and everything all at the same time. And by no means am I saying this is something that's ubiquitously happening. But the fact that, you know, like I said, when, This stuff was new and people, you know, you kind of like, okay, well, everybody's learning it. It's too bad that they don't understand it now, but they will later. And, but we're still seeing some movies that suffer from these problems.

And then, you know, in post-production, you know, a lot of your time color grading becomes damage control instead of finesse. You know, what you want is this starts out great and making it better rather than this is a disaster. How do we get it to something that's okay? Right. Well, and I actually had Alex, like, sit me down and explain to me, like, what a LUT is, like, what it is. And I guess to your point for, to simplify it, basically, if I correct me if I'm wrong, you're saying, better to do your tests, establish your look pipeline, whatever it is. Then you save a Lut. That's all that math there.

The Lut is the math solutions, say, if you monitor with that, you light with that and all that. so that when you go back into post, it looks more or less like what you were shooting because the underlying math is there and not the solution that the camera is like remapping stuff too. I mean, overall, the yes, except the one part of that that I would give the caveat to is the math is always, when you're applying it,

you're talking about it being concatenated where you don't see the components anymore, right? Like you're making a huge computation stack that's all of the math. That's how you decided to create the let, which is it's the transformation from this is uninterpreted camera data that's not even meant to be viewed. It's just information about the scene. And let's prepare it with a photographic look.

And you're taking all, like you could do something dumb and it's simple, but if you do do something very complicated, that's a huge computation stack, and then you flatten it into a lot. But the thing is, you're always doing that. Whether you do it a good way or a bad way, it's always flattened. So the sort of fact that it's flattened part

is not the crux of what the issue is. Does that make sense? Yeah, I just, I was just, it was more of a, you hate that in the QA. It was more of a statement than a question, but it was just having to realize that like, oh, like if I did the, the let itself, he was like,

you can theoretically, reverse engineer a lot, but I don't know anyone who's got the years that would take. When you say reverse engineer, do you mean invert or do you mean No, I mean like physically look at like the lookup, the numbers in the lookup table

and go, all right, what was the math that made these numbers? Oh, oh, the math, I see. Yeah, not, so yeah, yeah, reverse engineer the computation stack from the lot. Yeah, yeah. The reason I bring all this up is because there's, I need to get off the internet. But unfortunately, I keep going on. And recently, like in the in the past week, there's been a lot of conversations about, oh, why do films from the 90s and 2000s look so warm and whatever?

And now they don't. And every single comment said, they shot on film and now digital. And they shot on film. And I was like, I promise that's not the answer. And everyone, I got downvoted to hell. Yeah. And, you know, you come out with movies that I think look, you know, nicer. And then this gets into the whole thing of like, what version of a movie are you watching?

how was the transfer? Did they regrade all that crap? But I think your work has done a good job of proving that if you want it to, there are methods for which to take that chaotic pipeline and speak of and make it yours. I mean, yours is obviously a taste-based one. It's not empirically like this is the correct way to do it. Yeah, just how I like it. Yeah, but I like I also like it. Yeah. Well, but I, yeah, oh, go ahead. No, no, I was just, I was just going to say, you know, I think that's, you know, people cherry pick movies that they loved from the survivors, you know, yeah. And it's like, well, yeah, but what about all the other movies that you didn't like at that time? They were also shot on film. Like literally every single movie was shot on film, not just the ones you actually like.

I mean, you know, there's a thing where always seems like a coincidence which whichever magical, unachievable thing happened to be made at the time that the person has the opinion was like becoming an adult, you know? For me, it's the fight club in the matrix. I'm always like, how can we keep doing that again? Yeah. Yeah, so, yeah, and it's just taste, you know, like they,

people were doing what, you know, there was the taste of the filmmakers who made those movies versus the ones they didn't like from that time and then their own taste of which ones they like and they're trying to you know you said you know you started this

conversation with too much time on the internet and there's so much stuff on the internet that's the we want everything to be one weird trick you know this kind of like elapsing it down to some one thing and you're talking about like an entire artistic endeavor

like how is the movie lit like there's people making really pointed decisions using a lot of like expertise and artistry every day all day to make every single one of the shots in a movie

and the idea that it's this one weird trick of which camera model they used shooting in that's like that's a big one yeah cinematographers don't want you to know contrast yeah yeah anyway Sarah what are you going to say

I cut you off oh no no no no like I said this is a ranty podcast but it is a good transition because there was a couple scenes in the film that I don't think would stand out like in a sense of like people going wow that was beautiful but for me I was like what ha and the first one was that scene in the kitchen with Carrie Washington I think is was that terrifically difficult or was that like yeah we just put some because for me I was like this I it this is lit perfectly I don't know why it's it stuck out to me but I was like I I love this and that one and the

scene of Benoit sitting at the picnic bench outdoors both of those oh nice a look to them that I was like I would like to know how these were lit and like what what what what were the challenges there well thanks yeah no I'm glad you like those um hell I don't know um you know those are pretty yeah well that was I mean that was the most fun thing in the movie is all of the um and the biggest thing to figure out was how to do all the the changes with the sun going behind the clouds and bursting back out and so good and all was it yeah but yeah those two scenes that you mentioned i mean it wasn't any kind of specific weird day to do hard you know i mean it's every day you're um solving a puzzle and you know um i mean we weren't doing anything

weird in there um you know like in the kitchen I think part of what was going on is just that and we had some weird things to deal with like there was I think some skylights we had to cover I'm not sure you know but we're really just trying to in terms of the overall principles of how are we lighting this I mean we're just trying to make it feel like

what it really was which is there's the really big window that's you know the door window that the one that's side goes out of and it's supposed to feel like there's a big push coming from there so obviously the sort of the the basis

of all the decisions was it should feel like what the room is except we should feel like that except look awesome and yeah and I think you know the the room has a lot of white walls

which you know I think that's what stunned to me was that still looked good with all that what would like for me I don't know if I could pull off anything close to that just with all the white which I'm faced with constantly. Yeah, I mean, you know, it's just one of those things where, yeah, I mean, you know,

I always remember, you know, near the beginning of my career, like, that, you know, just a big white box is the absolute enemy of cinematography. Like, it's just the most boring thing. But also because of that, I'm just, you know, like I'm always aware, like, since that, you it, you know, became an obvious challenge to me. I'm, you know, I'm sort of, you know, uh, you know, especially at that time, I mean, I'm still aware of it, but especially at that time when it was such a frustrating challenge, I was hyper aware when I saw a movie where a room's all white and it looks fantastic. Like, what, what is it that I like about that? And, um, the, the, the weird thing is

that I don't, again, it's not a one weird trick thing. Like, I don't even know what the, you know, it's not like, and then what I discovered was it's this, you know, it's, it's more like, you know, it's, it's more like a sum total of all of these things. But I mean, one awesome white room scene is an Empire Strikes Back when

when Darth Vader shows up at Cloud City and they're all in that big white box. The room is kind of a big white box, you know, and it looks absolutely fantastic. So that's that's one for inspiration. Windows and that one too. I think so, yeah.

Yeah. Trying to remember. Nighting here. Yeah. Yeah, but were there in the original? Yeah, I don't know. I think that might have been in the, I think that might have been one in the special edition. They might have added some of the windows. I think it kind of was, at least to an extent.

Let's put it this way, it's very much a big white box. Like there might be facet stuff. I don't remember. But it's a good one to watch the next time you have to shoot people in a white box. Watch that one for some inspiration. Well, and I guess with the, because yeah, unfortunately I saw it. like last week so you know the memory starts and they don't they don't take kindly to you sitting

there with your phone and yeah but uh i think the thing with benoit on the picnic bench it speaks to a larger uh thing i noticed that i very much like the way your lighting faces um which i know is everyone's like you're supposed to light spaces but we all have key lights um and i it was something that I couldn't first of all the delicacy of the light is nice like it's it's present but it's not overbearing uh the density on skin is like perfect in terms of the color grade and then uh but i could never tell if it was bounced or diffused and i'm assumed it was a combination but it it was it was a fun little project to try to tackle when i was sitting there in the like question zone well one of the interesting things about that scene is

firstly, if I'm remembering correctly, I think we shot that on day one of the shoot. And, you know, so obviously the, you know, similar to the kitchen thing, of course we're lighting it. We want to make it look theatrical and cinematic, but it's got to be based on what's there so it doesn't look. So it's also, because it also has to be evocative of the space and time of day and everything. So they're under that tree, you know, so the tree's kind of like a dark roof. and you know so the lights clearly coming from the side you know the roof continues way on in the one direction so the lights obviously coming from the side that's that's open and we were talking about the LED light control stuff and the fine-tuning colors and one thing that was really exciting

about this movie is this is the first movie where I felt like we were properly for real able to do that with day exterior stuff because of you know like in the past there's been you you know, like just LEDs aren't bright enough, so you have to go back to HMI's or whatever. So the fact that, and I'm not saying you couldn't do it at all, but like, you know, things have been changing and then just the stuff that we physically had on our job as opposed to what I had on other jobs, even if it existed in the world, you know, it's kind of the first time we could do that. So, you know, given that that was on day one, that was actually the first time that we're lighting.

So that's kind of interesting that you brought that one up. was kind of the first time that we're doing day exterior stuff where and so so it was like I was metering the the sky and putting the you know setting the lights to exactly the sky color so that you know you could make this wrap of the light and this is and this is huge because this is one of the things that that's always been you know there's tons of movies from before the technology that look fantastic but you can kind of see it a lot where, you know, if you make a wrap outside, if you do it with HMI's even if you try to put the right gel on it, it's never going to be exactly the right color, right? Like if it's,

whether it's supposed to be exactly the same color or even if it's supposed to be cooler, because this part is just sky and the other parts, a mix of sky and sun and it needs to look a little cooler or whatever it is, whether, again, whether it's supposed to be exactly the same or whether it's supposed to be different in a specific way. You can't, you know, you either can't control it at all or you can only control it in a very clunky way and you know also HMI's like if they're if they're new they're magenta and when they're old

the bulb is new it's magenta and when it's old it's green or vice versa I don't remember which one and you know it's hard to you know the jealous to control that are in clunky increments they're not in fine tune you know uh yeah exactly so

ass okay yeah so you know um yeah so we were just and and we were shooting that seen it was getting darker as we were shooting it so we were chasing it so as the you know as it got bluer out then i you know i'd meter it again and adjust the lights again and we'd adjust the stop and also adjust the white balance because we can we could also do that in post but um just because we can also adjust the white balance it continues to look the same even even right now

on set not just later and um yeah so it was just a way to um you know we're able to have these lights through diffusion and kind of a big rap that's soft but directional and can be made the right color so that it doesn't feel like just some

artificial movie light interfering with the light. Yeah, I guess when you say that that LEDs can be powerful of it, that literally just like, I was like, oh, duh. Like the past two, three years, all these companies have been coming out with

okay now it's the size of a moon but like who cares if you if you need that much light generally you have a budget for that you know yeah yeah well yeah and even since even like even when we were doing that um even when we were doing wake up dead man they didn't even have the vortex 24s yet oh really yeah i mean like yeah we were doing it with um uh we called them vortex which was uh it was four vortex vortex vortex eights in a in a single you know in a single yoke um you know so we had a we had a couple of those i think and not seen i think it was two of those those uh four texes yeah and so it's just uh to simplify it's just that through some diffusion yeah balance the exposure hit record yeah see that that's that's the thing i think that is is missing

a lot when uh we talk about all that you know at the first half of this there's a lot of like technically, but that's all in prep. When you're actually doing it, it can be just that simple. And it tends to plan for it. Well, I think that's the thing that sometimes people get confused when they're saying like, why are you doing all of this stuff? And I'm like, no, I'm doing all of this stuff in prep so that we don't do

anything and it just works when we're shooting. Like when we're shooting, we put those things up with some diffusion on them, meter the sky, put it at the same color, or if we want it to be a little cool or put it at the same color and then cool it off a tiny bit or whatever. And then, yeah, and then, you know, also, you know, and also the fact that we are fine-tuning the exact, you know, if you do that with an HMI, what are you going to do? Like if you scrims are huge increments, you know, backing it up and pushing it in, that, you know, that was a grassy, Hilton, like, uneven. You're not going to do that, you know, like the fact that it's, you know, it's easily going to be the case that it's either going to look too lit or to, or not do anything, you know. And, and. Because, you know, even just because scrimms are a, you know, a very coarse and not fine increment, whereas here you get it set up and you just do the tiniest little increments until it really feels like you've sculpted it, but it doesn't, it neither looks like there's, you know, on the one hand, it doesn't look like there's a huge movie light hitting them, but the other hand it is doing what it needs to do to make that, that beautiful light wrap on them. And again, you can chase it as it's, as the light's going, you know, because sometimes even.

if you could chase it with scrimms, even if you had scrimms that were in super fine increments, you're just not going to wait for it. I mean, the actors are in the middle of doing a scene, and you can't just every time be, hey, stop and let me, you know, and whereas here, it's just on, I've got it on a dial, and it just goes down a little, and I can do it during the shot, even if it's a long scene, and it's getting, you know, and it's getting darker. I can just, okay, it's getting darker. I can, I can, I can nudge the light down and nudge the iris open, you know, you know, slowly so that you can't see it. then you know the uh do you where am i going with that so is is your i i'm just reminded now of

the um sort of outdoor confessions there's like a couple of them uh is that kind of the same process just one key light over here everything else they they were actually really different it depended on the on the scene um some of them were i mean some of them were literally nothing at all just like totally natural and some of them was some of them yes is you know exactly what you said where it's just a you know just a a diffused light on the side but then there were ones where like if we'd already started and it looked a certain way and now the sun's over the top we would put a you know diffusion over right so that there were a lot of different things and there were some I think there were probably also some where it wasn't just one light for a wrap or we had a little

more light surrounding them just you know just partially because of once you're already in a scene and it can you know it's not that it couldn't have looked good without doing that it was more just we need to keep making it look like we're on the same scene here that you know it did seem i don't know where'd you shoot it uh London London so okay because my first literally the first thing in my notebook was like there's a overcast look to a lot of this London yes yeah we had we had a lot of light changing though where um you know it's funny because obviously in the movie we have all those light cues you know that uh that we're doing on the stage with the sun coming in and going away but um when we're doing a scene i mean obviously you can't have that happen for real because

if if it's happening for real you can't make it happen on the same line for every take and every shot so um you know obviously if you know when we're when we've actually got sun coming and going um in scene that's supposed to be continuous, we have to deal with that a little bit. And, and, you know, that's, you know, yet another thing we can do by having this finesse control, you know, right at my fingertips is, once that diffusion is over top, like if it's not over top and the sun's coming and going, obviously, that's a, you know, that's a different thing. But once it is over top, I can, I can, you know, finesse it. I can either do it manually or I could even just get two settings, like, when the sun's out, you know, our light gets bright, you know, our light gets brighter and

warmer. And when the sun's, you know, when the sun goes behind the clouds, you know, the light's got to be cooler and dimmer. And I could, so I could either do it as two set cues or I could do it manually where I'm adjusting it. They can literally do it where as the sun comes out, you start nudging it towards the other, you know, towards the other queue and stopping down. And then when the sun starts going away, you start nudging it the other way and opening the iris up a little bit and if you, you know, and if you do it slowly, you can't, you know, it's, you know, it's certainly less of a change than what would happen if you didn't do any. Yeah. You know, I think, yeah. I think that was the big thing that I came away with. And again, I was watching

that, I was enjoying the film tremendously, but I am watching it like to ask cinematography questions. So simply your brain, it does happen. But I was, I kind of was chuckling to myself because it did feel like, even that type of stuff, like makes it feel like a magic trick where you're watching it, and you're like, well, that was effortless, and then you're over there, like, you know, key frames on her. It's like very cool. Speaking of magic, was it just production designer? Whose choice was it to put the Ricky J poster in the bar? I'm sure that was Ryan.

I'm sure Ryan asked for that. I mean, I wasn't, like, I wasn't there when that decision was made, but I mean, Ryan loves Ricky Jay, was friends with Ricky Jay. I mean, I'm sure, I'm sure it's something he asked for and wasn't exactly. We got it. Yeah. It was a phenomenal film. I got to let you go here soon, unfortunately. But I did. I guess we could, well, just have to have you back to.

I'll do it. I'm enjoying this. That's all you really care for. Honestly, like the low key, the main thing that keeps this going is like, I know a lot of people are on press tours and it can get real samey. Like what if they just had like an hour in the middle where they just got to like kick back and then take a breath I love it

kind of the idea but the one I remember you know I don't know about famously but you guys had designed like for glasses and eye reflections you know you'd like gaff taped up so you know the right shape oh for nights out

yeah and I caught it in this one with some fire and I was like and and I only noticed because I was thinking about it but I was wondering in What other ways did VFX help you achieve what you guys are trying to do? Like, were you doing any, like, you know, painting out of equipment or, you know, because it's not obviously there's, I suppose there's a little action, but when Phil's kind of

simpler like this, it's fun to know like where those little hidden scenes are. Well, the fire reflection of the eyes is not visual effects. We actually did that. Oh, really? Yeah, that's, that's the next evolution instead of putting tape on, lights, we use monitors. So with the same light control software, I've got videos.

So in prep, we shot fire. We actually shot flames at 200 frames a second. So the flame was actually roiling really fast. And then it's at 200. So that's why it has that kind of smooth look to it because it was one that was really going crazy, but it was shot fast. And then, so I can play it back.

And I'm only using the, from the, video, I'm only using the luminance, the shape of the fire. Where is it dark versus where where is it black? So I'm only using the contrast from the video. And then the light control system is actually setting the color. Because if you have the color rendered video, that doesn't look like a flame because some of it's blown out and it's white. So it's not actually, the real flame is orange, but when it's photographed, it's white. So we don't use any of the color rendering. So I'm setting the, it's just like a movie light. I can set the exact chromaticity, the XY chromaticity coordinates

of the fire, and I can make it brighter and darker, and then the video is just generating the shape that you see. So we did that with several things. We did that. There's also a thing where there's a scene where in Vera's glasses you're seeing

reflected the church windows. That was a monitor with the church windows being played back. Obviously, we could have taped up a light because that one wasn't moving, but we had the monitor. Yeah. Is it? Another one that could have been taped up, but why do it?

We've got a monitor. You don't need to waste a punch of tape. Is, you know, Grace, in the Grace flashbacks, there's one shot where you actually see a cross in her eye, and that was reflected in her eye, and that was from the monitor. Yeah, okay. So this is exciting because I'm a big, you know, catchlight nerd. There's a whole artistry to catch lights. You know, you know, where you, people look sad, happy, whatever.

Were you just using a normal monitor or like one of those, you know, mimics? No, it was a very cheap monitor. The whole idea of it was that because you're not looking at it or anything, you know, so it's just the, I mean, it was literally the cheapest monitor I could find that was bright enough. Because, you know, to get a very bright monitor, it's in the world of lighting, a very bright monitor is still not very, very, bright so it's not worth spending $20,000 or something. You know, so it was literally like a $150 monitor or something like that. And we just in prep, we just had, we're like, let's buy two monitors that are lighting monitors. They're in the lighting department. They're in the lighting gear. And it's the monitor plus the little router and decoder. And so it's just, I'm just running

it on the same system that I'm running all the lights. And yeah. But to answer your visual effects question. I mean, sure, there's things here and there all throughout the movie, but not tons of stuff. But one that was really fun, specifically with fire, though, is in the, and I think we did this in both, because there's two firelight scenes. There's the one with Blanc and Judd, and there's the one with Judd and Martha. And I know for sure we did it in one, and I can't remember, I think it might only be the Blanc and Jud one, but there's a wide shot where you're looking right at the fire and everything. And we, you know, there was practical fire, but it's, you know, it's indoors and it's not ventilated, so we're not allowed to make it big enough that it could, like, actually

light the scene, you know what I mean? Like, if that's with lighting the scene, the scene's going to be black, you know, just literally no exposure on anything. So we put movie lights in the fireplace, and I had my fire effect going on the movie lights. And then we took them away and gave the visual effects people an empty plate. And then also our second unit shot some fire elements. And then they just put the fire. You know, they took the movie lights out and put the fire in for just that wide shot. Yeah.

Well, it's inspiring stuff because it's technically all very achievable on most levels. You know, and that's always cool. When you see something that you're like, I really like that. to hear like that's you can do it like that that's always nice for me yeah well i'm glad you now know that those were 150 dollars and not a not a visual effect not inexpensive i don't have a use for it but lord knows that's going in the back pocket like yeah um well i'll i'll let you get on to uh whatever you've got next but uh i would absolutely love it if you could come back i got plenty of stuff we could uh keep chat yeah love to thank you so much for having me what

What a blast. Frame and Reference is an Albot production, produced and edited by me, Kenny McMillan. If you'd like to support the podcast directly, you can do so by going to frame and refpod.com and clicking on the Patreon button. It's always appreciated. And as always, thanks for listening. Thank you.