Hi everyone, I'm Kali from Mixing Light and I'm here today talking to Peter Postma who's the Managing Director for the Americas at Film Light. Thank you so much for joining me Peter, it's such a pleasure to have you on. Yeah, thanks for having me. So you've had such an interesting career. I met you in your capacity as Managing Director at Film Light, but you have worked at Kodak as a colour engineer prior to that. And you are an alumni of Rochester Institute of Technology.
So I just wanted to begin by talking to you about this career trajectory moving from Rochester to Kodak and then to Film Light. These are three real staples of our industry. So can you tell me a little bit about what you studied at Rochester and how that led you to Kodak? Yeah, I mean some of it was just kind of being in the right place at the right time, I guess. But initially, my interest was actually more in just kind of computer graphics and video games in particular.
So I initially was studying computer science at RIT. But I quickly found that on the technical side of things, I wasn't learning anything in school that I couldn't, I kind of quickly learned myself just by reading textbooks and playing around and stuff like that. It was the creative side of making images and making movies that I needed feedback on.
So after a year in computer science, I actually switched to the Department of Film and Animation and ultimately graduated with a Bachelor's of Fine Arts in Film and Animation. But as part of kind of my electives, of course, we have the Center for Imaging Science right next door to the School of Film and Animation says, "Oh, there's some interesting classes about colour and vision over there."
So I started taking those and then, yeah, just kind of saw how I could kind of really tie together my interests on the technical side of things and my interests on the creative side of things by really focusing on colour for our industry. Because a lot of the resources out there for learning about colour and stuff like that are much more focused on print press work and stills and stuff like that. There wasn't a lot focused on the motion picture industry.
So it's kind of an exciting area to study. Definitely. And it still feels a little bit like that. I mean, I think, like you said, right place, right time. Rochester is one of the only universities that I'm aware of that does teach colour science and imaging science in a really meaningful way. I certainly know that here locally, we don't have anything like it. And so after that, like Rochester and Kodak have a bit of a crossover of some kind. Is that right?
Yeah. I mean, Rochester is an interesting town because it is kind of in the middle of nowhere, upstate New York, but it's the headquarters for Kodak, the headquarters for Xerox, the headquarters for Bausch & Loem, which is a big lens company of Corning who makes glass for all kinds of optical devices. So it really is this kind of hub of kind of optics and colour science and stuff like that.
So, yeah, luckily, it was actually one of my professors in my last year at RIT who worked for Kodak and was just working as an adjunct professor who said, "Hey, there's this opportunity at Kodak you might want to apply for." And so I did. And yeah, I ended up being with Kodak for five years.
But my first job there actually as just kind of an intern was actually looking at different kind of digital image processing techniques for their laboratory because it was really right at that cusp where things were starting to go from film to digital and what kind of tools that they want to have just for their own internal team to kind of assess and process images. Oh my goodness, that's fascinating to be on during that transition.
And without implying anything about your age, can you tell me what year it was then? Yeah, this was around 2000. Yeah, around 2000. OK, so like a good 12 years before it kind of that that change to digital really happened in a sort of commercial way. Yeah, so it was actually 2001 that I moved out to Los Angeles with Kodak because I was initially just in Rochester.
And I worked I was still officially in Kodak, but I worked right next door to Cinecite, which was the place where some of the first DIs were done. So it was right after 'O Bropther, Where Art Thou?' and one or two other films had been finished. So really right at the forefront of that whole DI thing. And I remember, yeah, I was on the team that helped build and deploy the first hardware that could apply a 3D lookup table in real time.
So it was this crazy sun microsystems computer with all these FPGAs in it just to be able to process a 3D LUT in real time so the colorist could see what the film would look like once they went to record and print it out. So which is such a simple thing that every color corrector does now you apply LUTs all the time, you apply three or four. But yeah, back at the time, that was quite novel. So that's unreal. I mean, we really take it for granted now, don't we?
And was it during that time that you were part of the ASE committee to develop the CDL? Yep. So right when the ASE Technology Committee formed, luckily, the first meeting just happened to be on Kodak property. So I managed to get in the room, so to speak, and start talking about it. And yeah, because I already had some experience supporting digital intermediate at that
point. Yeah, I was part of those early discussions as to building the CDL because it was quite clear at that time that things were going digital. But also just for supporting film, you know, it used to be you would actually have you'd either rent out a movie theater on location or have like a trailer with a projector and it's actually do film dailies. And so this is around the time when film dailies were going away and digital dailies were starting to take over.
So you were scanning the film on a telecine every day. But the directors of photography saw that they were kind of losing a key bit of feedback that they got from the lab because it used to be, you know, you could look at what your printer lights were every day and know kind of how things are going if you needed to like overexpose a little more underexpose a little more just if the film from the last day was good.
But when you just get, you know, a DVD or something with your dailies on it, you don't really know, did the telecine operator have to do a whole bunch of work to make those dailies look good or were they just, you know, sliding right into place as they should.
So the CDL was originally kind of thinking as a way to be able to give feedback to DPs to say, here's some specific numbers so you can see if it is in fact being consistent every day or if something changed and then, oh, maybe there's a problem with the film or a problem with the camera or something else with exposure, we need to go back and check to provide that kind of key bit of feedback.
I love that idea that the CDL isn't, you know, it's so much of a grade as it is a communication tool between the departments because, you know, that kind of ties into how CDLs are used now throughout the whole process. They're communicating that idea from set through to VFX and through to DI if you want to if you want to start there.
And to have people who could look at those numbers and say, oh, this is what they must have done, you know, it just never occurred to me that that was happening on the camera side. So, yeah, that's really cool. I mean, with printer lights, it does seem a little bit more straightforward to say, oh, they put in, you know, a stop of red or took out a stop of blue or what have you.
But, you know, being only 10 numbers on a CDL, you could you could look at it and say, oh, yeah, they changed the contrast and they offset it up a bit. So. Right. Well, and that's actually one of the reasons why we went with a slope offset power for CDL rather than lift gamma gain is because in normal like lift gamma gain controls, there is no straight offset. So if you wanted to do just printer lights, you actually need to combine, you know, two or three controls to do that.
But when you have offset, that can be your printer light. So if you if you do want to kind of stay more in that kind of printer light mode, only have three numbers to work about. You can just use the offset part of the CDL. And so you only have to track three numbers still. So that would all still work. And the beauty of that as well, that it's completely invertible. You know, anything that goes up can just come down again or vice versa.
So you maintain the linearity of the negative, which is, you know, just a really great and useful thing to have for other departments anyway. Right. Yeah. Yeah. So it's interesting to me that you've kind of been a part of these committees along the way that have in a really large way shaped the workflows and the way that we work every day. And yet you're sort of not like a household name, so to speak. But you've certainly been there and kind of developed the way that we work today digitally.
And you're continuing to do that. And I know I was skipping ahead a little bit, but you're part of the ACES project at the moment as well, aren't you? Can you tell me a little bit about that? Yeah, I mean, actually, I started with ACES again very early on, probably over 16 years ago. I mean, when the first seeds of ACES were planted. So it was actually while I was still at Kodak. And there was already this recognition that, again, things are going digital.
We should have some kind of industry standard for how to process images. And it took a good I think it was close to 10 years before ACES 1.0 finally came out and everyone could agree on, OK, this is kind of at least a good first step, if not, you know, everything we wanted it to be. So, yeah, that was really just kind of encapsulating some of the best practices of the industry at the time to how to set up a color managed workflow.
And that work continues because, again, when ACES started, it was still very, very film centric. It was keeping in mind, like a lot of time was spent on, OK, how do we bring film scans in to ACES and how do we get back out to film? And of course, these days, that's not so much of a concern as dealing with all the different digital cameras, dealing with high dynamic range display devices and all that kind of
stuff. So, yeah, the ACES committee is getting pretty close to releasing ACES 2.0, which is a kind of major upgrade that more directly takes on those concerns of, yes, super wide gamut devices, both on the camera and the display end and dealing with different dynamic ranges and all that kind of stuff in an even better way.
And that would be I mean, there's recent developments in LED walls, you know, and Unreal Engine and, you know, utilizing backgrounds that have very saturated and potentially like out of gamut kind of luminosity and colors to them. Is that part of what you're looking at with ACES 2.0 as well? Yes, I mean, ACES was always designed to be able to kind of accommodate any possible color. So any color the human eye can see, you know, is able to be put in an ACES container.
And there's certain issues that, you know, ACES can't solve for you. But certainly it could be, again, that good foundation, that color managed workflow that things then fit into. So it's not going to have like a push button simple fix to make your screen match your camera and stuff like that.
But it's part of the workflow you set up that makes that stuff easier and should make it consistent so that you can use the same approach no matter which LED wall you're shooting with or which camera you're shooting with.
That I think is the biggest thing is to not have to be dependent on like, you know, different specific manufacturer solutions so that like, OK, if I'm shooting with a Samsung wall one day and a Sony wall the other, I have to do like completely different techniques just to bring it again to a basic industry standard, then you could put your special sauce on top of that. But at least the basic workflow is the same no matter what you're doing.
It's fascinating, like the complexity of that and just trying to deal with all of the variations. Has there been anything particularly challenging technology technology wise along the way, like, you know, in terms of film going out and digital coming in or any particular technologies that have been tough nuts to crack? That's a good question.
I mean, I consider myself very lucky to have started my career when film was still around so that I could learn like a lot from the limitations, really, of the photochemical process, but how much the engineers at Kodak were able to do to get like great looking images out of that. And to me, you know, being the young guy in the room, a lot of times in those days, it was quite clear that digital was the future. That film only had so much life left on it.
But there were a lot of lessons learned that could then be applied in digital and digital should be a lot simpler because, you know, there's so much intricacy involved in the photochemical process for film. You know, there's this whole industry that supports all the different technologies that you need to be able to lay down the chemicals on a strip of film to expose it. But then also the laboratories which can then develop it and turn that into an actual image and everything.
You know, if someone were to try to build that infrastructure today, it would just be impossible. Whereas digital is supposed to. I thought digital was going to be a lot simpler. You know, it's just like you've got pure numbers you can look at. You can go back to those same numbers. You know, you have that magical undo button that we have in in digital. But yeah, because as you say, there's constant technology changes, there's new sensor technologies, there's new display technologies.
That's, I think the trickiest part really is not any one particular device that has or technology that's that's presented a problem. It's just you have to keep up with all these different things that are being produced in all these different corners of the workflow that then keep pushing it and evolving it.
And I think the biggest change, obviously, in the last few years has been high dynamic range displays, both, you know, laser projectors in the theater that can get a lot brighter and a lot darker, so much better contrast and HDR displays in the home. And we did some very early work with Filmlight with Dolby, you know, on their early HDR kind of technology and stuff like that. But the biggest hurdles, I think, for HDR actually ended up not being technical.
They ended up being creative. It's like, OK, we now have this widely expanded palette. How are we going to take advantage of it? And I think that's something we're still kind of figuring out as an industry is, yeah, what's the best way to use HDR? And that answer is often different for every project. You know, how much of that extended highlight range, how much of that extended saturation range do you want to take advantage of?
And when do you know when to kind of be reserved and not pushed it just because it's there? Yeah, those aesthetic, you know, the aesthetics of standard dynamic range have been around for 100 years. And, you know, we we kind of have a sense of where culturally we think skin tones should see it and where highlights should sit. But we're kind of like you say, it's a new frontier creatively as well.
And I think another level of complexity, which I'd love to understand as well, is that, you know, working for companies like Kodak and now Filmlight, you're in a position to be supporting and contributing globally. And, you know, different cultures and different locations have different standards. So, you know, sometimes I forget that there are a whole other workflows out there and whole other displays, you know, in China and in, you know, the East and other places.
Can you comment at all about dealing with those kinds of different cultural culturally used technologies as well? I don't know how to say that properly, but yeah, no, I think, you know, because I am more on the technology side, you know, helping the creatives do what they do. My biggest thing is just to provide the tools that people can take it so they can take advantage of those new technologies and kind of guide them a little bit as to how we think they're supposed to be used.
But then listening to the creatives once they say, oh, I like this, I don't like this and just constantly evolving that. And yeah, you're right. It can be very different for different cultures. I mean, even if you look at just like stereoscopic, you know, which had a big resurgence a couple of years ago, it's still actually quite huge in China. For whatever reason, in China, they really like their stereoscopic 3D movies.
So we still have to maintain our 3D tools, even though it's much less common, you know, in the US and Europe and some other parts of the world. So, but yeah, so we maintain all those tools. And if, you know, someone in our sides of the world wants to take advantage of it, they're there to take advantage of it. But if they don't, they don't have to. So we've talked about being at Rochester and Kodak. What was the work that you were doing at Kodak as a color engineer?
Like, what would a day look like in that role? So initially, as I said, it was kind of just evaluating software tools that were in the industry for people at the research lab at Kodak to use. So it was because Kodak, of course, had developed the the Cineon software, which was on the first digital compositing software. And that used to be their own internal tool as well. But once Kodak abandoned that, you know, again, technology and industry moved on. So they need to needed to move on to.
So I was helping them evaluate, OK, what can they do with Shake, which is another compositor or Raise, which was a new compositer at the time that was coming out. And also just different tools for dealing with digital still cameras.
And so actually, once I had kind of done my initial evaluation, those different devices, that's when I got involved in a project which eventually be called the Kodak Look Management System, where you could take an image with a digital still camera and then put that through a film emulation to see, well, what would this look like if I shot with, you know, 5218 negative? Or what would it look like if I shot with a slower speed film?
What would it look like if I did a bleach bypass or things like that? So we had these digital emulations of all these photochemical processes. And so when the director and the DP are on set deciding, you know, what stock they want to go to, what film process they want to use, they could shoot an image with a digital still camera on the day, put it through the software and emulate all those different things.
So for a couple of years, I was involved in that project and managing, yeah, kind of profiling different SLR cameras to get them into that software and then making sure we're accurately emulating those different different processes. What kind of tools did you use to do that? Were you like, you know, looking at things through a microscope or, you know, like how does one do that? It really is a mix of things.
And a lot of it was kind of just homegrown by some of the color scientists I work with and just kind of evolved over the years. So a lot of it was just standard densitometry. So, you know, like you record out a bunch of film patches or different densities on film and then measure that with a densitometer.
And you take that information and or you measure it with a spectrophotometer so you can get the full spectrum of the light that's shining through the film and capture that and put it into your model. For profiling like the digital still cameras, we actually came up with this little device that had a piece of dichroic glass in it that you could slide up and down to basically capture the full spectrum of light.
So we had a full spectrum white light behind this dichroic glass that you could shift in like exact increments and basically take a whole bunch of stills to profile. So basically how it was seeing each wavelength of light and get a full profile then of how the sensor, how that digital sensor reproduced light, match that up with your measurements of the film. And that's how you kind of got the emulation between the two.
So that's unreal. I mean, that kind of physical thing, you know, I just I don't think a lot of us actually take into account that there are like prisms and bits of glass. Like, you know, that feels like something that happened in like the 1800s when they were first learning about the different colors of light and things, you know. But obviously it's still a part of practice today, just in a more sophisticated way.
But I've never once in my life, you know, gotten a prism and looked at looked at light through it. I just think, wow, that's really interesting. And so what when you moved over to FilmLight, were you doing similar kinds of work there or was that a bit of a shift for you? It was a bit of a shift, but not not radically. So my first job at Filmlight was supporting our Truelight color management software. And it was going in and helping people calibrate their D.I.
Theaters to match, you know, what they were ultimately going to see on film, because at that time film was still considered the hero deliverable for most shows. Digital cinema was already starting to roll out. But since most people were seeing theater, most directors were used to seeing film in a theater, they kind of considered film their main primary deliverable.
So they wanted to make sure when they're color grading in D.I. that they're seeing as close as possible with the film is ultimately going to look like. So that was, again, yeah, measuring film with densitometers to make sure we're accurately profiling the output of, you know, that facility specific film recorder and the lab that they were working with. And then measuring with a probe the projector to see how it's reproducing light.
And again, you know, meshing those two together to make the digital projector look as close as possible to film. So, yeah, there's a lot of running around. It was just fun, as you know, Filmlight is a very international company and there are a lot of different corners of the world that were starting to get into D.I. So I traveled around quite a bit, just going into these different D.I.
Theaters, helping them calibrate their their rooms and set them up for D.I. I bet they looked good, you know, like once they'd all been calibrated, I bet they were looking the best they'll ever look. Yeah, of course, the trickiest thing with film is that because it's a photochemical process, you know, it's not 100 percent locked down. So that was always a frustrating thing, is that, you know, we could always exactly match what the lab did yesterday.
The question is, is that what's going to come out of the lab tomorrow? So sometimes you have these awkward conversations where you have to say, like, I know what I'm doing in the lab is that fault here or, you know, kind of figure out or are we doing something? And so, yeah, it's kind of funny that in those days, again, towards the end of D.I., like, I think film laboratories are at their absolute best because, you know, you have this rock solid digital projector to compare it to now.
So the post house could say, you know, hey, the lab was a little off today. We got to reprint this. You got to fix what you're doing. So the labs actually got better. They were doing it much tighter controls as well. So yeah. Yeah. Oh, that's so interesting to be able to side by side it. You know, I find today that I'm I'm constantly hearing back from clients that, oh, we just watched the DCP and it's not looking the same as it did in the grade suite.
And then you have to say, well, you know, have you measured the screen? Do you know if the lamps that it's full brightness? You know, you've got so many questions and commercial cinemas, I don't think, are doing that level of calibration. But it's that thing of just, you know, being able to say to somebody, look, when we looked at it during the D.I., it was correct there. And if it's deviating, at least we know where where it's deviating from.
Right. Right. Because you can't control every screen. But wouldn't it be nice to be able to send someone out and say, OK, it's going to get calibrated before you screen for the short film festival or something? Well, I'm lucky enough to live in Los Angeles where some facilities for certainly for larger titles will do that.
You know, if they're going to have a big film premiere or, you know, they know that it's it's an important screening, they will actually send someone out to measure the screen and make sure it's properly set up because, yeah, that is one of the most frustrating things I think about what we do is that you can always trust that an calibrated environment is going to look great. But once it gets onto someone's screen in their home, who knows if they'll actually be seeing what you intended or not.
So, yeah, that's right. Yeah, I do joke that it would be nice to send my Sony monitor out for everyone to look at things on and just ditch their iPads and phones and TVs. But unfortunately, we can't provide everyone with that. At the same time, one of my favorite anecdotes was from a DP I talked to who watched digital dalies on his laptop. And he said it was great because he could grade him himself. If he wanted him a little darker, he just tilt his laptop that way.
If he wanted it brighter just tilt his laptop that way. So he'd just grade his dailies by tilting a screen. Oh, my goodness. Oh, that gives me the chills. I don't like that at all. For dailies, that's fine, I think. But yeah, certainly when you get to the final, you don't want to be leaving it to that. It worries me, especially in short form, what some people are watching things on on their phones, you know, at the pub or something.
Yeah. So going all over the world and calibrating things and using the true light system. I mean, was the true light system sort of in existence before base light was was that like? Before Baselight or part of, how did that work?
So film light actually got started as kind of like the R&D group from Computer Film Company and all our first three main products, the Northlight scanner, the base light color correction system and true light color management are early versions of that all existed in CFC before it split off and became a separate company called Filmlight, where we said, hey, we can, you know, bring these tools to the industry at wide. It doesn't have to be, you know, just for this one, one company.
So Northlight was actually our scanner was the first product that really kind of took off. True light and base light kind of came out at around the same time, but true light definitely got traction a lot quicker because true light would work with, you know, any color correction set up any any kind of theater. And obviously, you know, if you don't have that basic color correction in place, you can't even start to begin to do to do a DI.
So yeah, actually, when I joined Filmlight, I didn't really know much about base light at all. I said, oh, well, this true light stuff is rock solid. I've seen it before. I'm happy to go out and utilize it. This base light thing is kind of interesting, but I haven't seen it. I don't know what it does. And but as I was working with the other people at Filmlight, I said, wow, there's actually, you know, something something really great here.
And and yeah, so base light was kind of the last product to grow up and get widespread adoption. But and you hadn't done color correction yourself prior to that. Only only technical kind of color correction. So I knew how to like drive a DaVinci 2K and how to drive a Pogle, but mostly just for how to set up a, you know, a Telecine the right way. Another one of the projects I worked on at at Kodak.
I've actually forgotten the name of it, but it was a product to to to again kind of make the film transfer process more consistent and set up a Telecine more like a scanner. So you could take, you know, whether it's a spirit scan Telecine or I.T.K. or whatever, we we developed some reference calibration film that you'd put on the Telecine with specific targets that you'd have to, you know, adjust the knobs to to hit.
And then you would know that your Telecine was actually making kind of standard Cineon scans much more like a film scanner that would be used for VFX and things like that. So I knew how to, you know, drive the controls to do that kind of technical grading, but never never creative grading. Did you ever sort of get the bug for creative color grading or have you always felt felt more like the technology was was who you are?
Yeah, I'm very cognizant of the fact that, you know, being a colorist, a huge part of your job is kind of client management and client communication. You know, it's figuring developing that that language with them about what does it mean when they say they want something redder or brighter or darker or that sort of stuff. And I fully respect that. And I kind of decided that that that's interesting, but that's not what I want to do. I want to focus more on kind of what's going on under the hood.
And, yeah, certainly like one of the great things I think about BaseLight is that there's strong color science underpinning all our tools and everything we do. And that's part of why I think we can explain why the tools are the way they are is because we've really thought through, like, we don't just want to make some knob that makes the image looks zazzy. And, you know, sometimes it works and sometimes it doesn't. It's actually like, why are we doing the image processing this way?
Is this the best way to do that? That's going to give you the most control over the image and keep things looking natural and not just like, you know, all electronic and weird. So, yeah, I really like being a part of that process of developing those creative tools and then handing them over to people who can get inside other people's heads and figure out how to actually get the images they want out of it.
Yeah, it is such a different, like, discipline, isn't it, to understand the technology that underpins it as opposed to being able to deploy it in a session. But you've trained a lot of colorists. I'd be surprised if you couldn't sit down and grade something pretty well. Yeah, I could. I've done a few, like, indie projects and for friends and stuff like that. But, yeah, again, I just decided that wasn't what I wanted for my career.
So when I spent some time with you and you trained me, you showed me through the current features as well as some of the new features that are coming out in version six of Baselight. And I just wanted to talk about a few of them because it's actually quite fascinating to me because prior to that, I hadn't had such a huge. Kind of interaction with Baselight and I was kind of taken aback by just how, like you said, thought through the tools were.
So one thing which I really enjoyed were the tools that worked in the perceptual color space. Can you talk to me about those? That was base grade mainly in version five. Yeah. So, I mean, the big evolution, I think, in Baselight over the past few years has been focused on tools that really emulate the physics of light and the way our eye reacts to light.
So as opposed to like our early tools like film grade tried to emulate what happens in a film lab when you adjust printer lights or, you know, vary the processing a little bit to adjust contrast. And video grade emulates what happened in a telecine, which, of course, was originally kind of analog voltage adjustments. So really just trying to work with the limited tools they had at the time to adjust the color.
So with base grade, which debuted a few years ago, that was really one of our first major attempts to say, okay, we have this color management in place now. We can know what the original light coming into the camera looks like. We know how the human eye works. So let's make a grading tool that acts more like a camera does and more like the human eye does.
So instead of grading in some RGB space, whether that's, you know, based on the camera sensitivity to RGB or your displays reproduction of RGB. Let's say we'll just work in human perceptual space because that's that's what color is. It's the name we give to how the human eye sees light. So let's focus on that.
And so, yeah, we brought it into this human perceptual space and we just made the tools work consistently to the way DPs and other filmmakers think about color in terms of, you know, stops of exposure or color temperature rather than just kind of like, you know, arbitrary RGB numbers. So. And so in order to do that, because there has been a lot of research into how the eye sees, did you take that research and translate it into like a digital map of some kind? Like, how did that happen?
Yeah, yeah. I mean, there's there is a lot of research done on on human vision and still there still is stuff that we don't really completely understand. I think like we have a complete, you know, physical model of the eyeball and how the cells and our retina work and all that. But then there's this big squishy brain behind it. And that's the part that you can't exactly, you know, just like pick a pick apart.
So there's still some stuff that happens in our visual processing that we're not quite sure why it happens or what part of the brain it happens in and that sort of stuff. But certainly for the initial part of the eye, that's all very well modeled.
And so, yeah, we were able to take some academic research papers that that model that behavior and kind of trim the bits we didn't need to simplify it to get it running in real time and that kind of stuff to actually do the conversion into the visual space and then adjust the colors in there. And so you were talking about color management and color management workflows have been really important to you throughout your whole career.
And one thing which I was interested in in BaseLight was that you're always working color managed. You can't get outside of that. Right. So within that color management system, you have like another little bubble for this tool. So it's working within another color space, but that's all handled automatically within BaseLight. Is that right? Yeah. So of course, the way the color management system works is at first have to know like what camera was used to acquire the image.
So most of the time for camera original footage, you know, we'll know if it was shot on RED or Sony or ARRI or GoPro or what have you. So we've profiled all those cameras and can bring them from that camera native color space into whatever color space you want. And yeah, for what I call like our legacy tools like video grade for lift gamma gain or film grade or stuff like that, you can choose what color space you want to work in.
So you could say, you know, I'm primarily in ARRI show shooting Alexa. So even though there's a few shots done on Canon camera or a few shots done on RED, I can actually bring them all into the ARRI color space and grade with it there. So my tools feel consistent. But then beyond that, yeah, when you get to Base Grade and some of our new tools like Xgrade and ChromaGen, they work in their own internal color space, which is that human perceptual space.
And so, yeah, that basically lets us emulate those those real world processes, regardless of what camera you shot with at the start. In terms of the way that it looks, you know, it's hard to describe it, talking about it. But when you're actually using it as a colorist, I was so, so happy with how like mainly saturation was mapped.
If you saturate an image in base grade, you never get into those kind of electric colors that you can reach when you've got Rec. 709 primaries controlling your saturation. So it's definitely something that's worth every colorist. If you're no matter what system you use, just jumping in to the Baselight Look software, which is now available, you can you can just download it to learn and play.
It is worth jumping in and just seeing what base grade does and how it reacts, because it is a slightly different experience when you're grading. You know, you might be used to offset grading, but, you know, to have these controls like Flair and the ranges, like you've got Bright and Dim and Dark and Light, I think it is on the two panels. I hope I got that right. Yeah. They all operate just a little bit differently from other tools we might be used to using and especially the Flair control.
I mean, it's just worth having a play. Well, thanks. Yeah, I mean, that's that's good to hear. It means we we did our job. So it is super good fun. And you also mentioned another couple of tools that work in that perceptual color space, which is X grade and chromogen, which are these six tools that are coming out there in beta at the moment. Is that right? That's correct. Can you tell me about those because they are just something else?
Yeah, so X grade is a tool that's, again, kind of plots all the colors in a human perceptual space. So it's not exactly the same as the space that base grade uses, actually, but it kind of plots the colors out so that an equal amount of change in that tool will be perceived the same as your eye, because we're, for instance, more sensitive to changes in like blue and purple hues than we are in green hues or for saturation the same way.
We're more sensitive to saturations in some color than another. So that color space, first of all, kind of smooths everything out so that you know, if you're making the same amount of change in X grade, you're going to see this perceive the same amount of change with your eye.
And it basically when we built it, we looked at, you know, other tools which allow you to manipulate the full color space as kind of a mesh to be able to like, you know, basically just push color colors around and mold them like clay. And we found it was a really interesting way to adjust colors, but also a really quick way to destroy the image because you can quickly get images crossing over with each other causing banding artifacts and tearing and noise and stuff like that.
So we said, how could we do this, you know, more smoothly and more consistently to get a good result? And the first was moving it from an RGB color space into this human perceptual space where everything is evenly weighted for all the colors.
And then the second was rather than, you know, like just manipulating a 3D lookup table directly, let's make it so you have kind of areas of influence almost like you are molding clay and pushing stuff around so that, you know, when you push really hard in the green and then you push, you know, the red direction in another way, it kind of pulls a little bit of it with it. So you don't get that tearing, you're just constantly kind of molding the colors around.
And yeah, just ended up kind of being this really organic way to quickly do what would normally be secondary color correction. So normally, if you wanted to make, you know, your reds a little bit more orange and make your greens more saturated and make your, you know, blues darker, you'd have to do three different keys and a bunch of windows and stuff to protect for exactly what you want it to affect.
But with X-grade, you just kind of grab that part of the color space, push it the direction you want to go and do the same in the three other color spaces and basically automatically interpolates and smooth between them so that you're not getting any of those artifacts. Yeah, I mean, I've tried pushing it every way you can imagine. And like you say, like one point will always affect the others and it'll protect you from from going into crazy town.
Another feature that I really like about it is the neutral line that runs through the center. So it did take me a little while to get my head around exactly what that was doing. But it looks almost like degrees Kelvin running through.
Yeah, exactly. We kind of highlight where the standard black body white points are so that, you know, if you do want to keep stuff more neutral, of course, there's nothing as no such thing as being truly neutral because you're always neutral with respect to some white point. So you can decide kind of what that white point is. But we also have, yeah, kind of like an anchor so that it keeps stuff that is close to neutral more neutral.
So when you're pushing other stuff around, you're not suddenly pushing, you know, your grays towards red, your grays towards green and stuff like that. So that was kind of another key component we realized early on is that normally you want to keep your your neutral scale kind of grounded and be pushing the colors around it rather than shifting the whole image. But you can take that pin out if you do want to go crazy and push everything around.
You can do that, too. So it's also really nice to just have that representation on screen because it can show like I was playing with it the other day and looking at, you know, I wanted to grab this warm color and which direction do I want to push it in? Well, if I push it down the neutral line, it's cleaning up and becoming neutralized. If I push it the other way, it's getting more orange. So, you know, for me, it's like almost a part of a map for me to to know where I am in space as well.
And it's also got ranges, which which I was having a play with and discovering the other day as well. So you can do a little bit of work more so in your dim zone or more so in your bright zone if you switch it over. So I think there's like like any of these tools, they start off looking quite simple, like, OK, it's color warper. And then you start to change a few parameters and you're like, oh, wow, OK, this can really do some pretty, pretty cool stuff in terms of look.
But yeah, definitely the, you know, utilizing it, that's that's another thing which I really like about the tools in Baselite as a whole, is that there are a lot of tools in the main grading options that can save you from going to secondaries. So Hue Shift is one that I would use every time. It might be my favorite tool.
Being able to alter the value of a hue and, you know, brighten skin tone a little bit and perhaps darken a background a bit rather than adding a window for a face or darkening off a background with a vignette. You know, I can just do that right there in the main page. So, you know, is that something that you guys are interested in particularly is keeping away from the secondaries?
Yeah, that was definitely a thing we consciously worked on, because any time you do a secondary, you're literally like carving out a part of the image, treating it totally differently and then trying to slap it back into that original image. So that's always going to be problematic. That's why, yeah, any time you pull a key, you have to go in and watch it through, check for noise, smooth it out, blur it, you know, throw a window on it. So it's only affecting the part of the image you want.
You know, there's all these kind of extra steps you have to do that eats up a lot of time. And again, it's just not the best thing for the image because you're really now kind of taking things down a different path. And that's something that, you know, ideally you should never have to do. You know, if you're able to capture an image you're happy with, you should be able to just kind of mold it to what you want to do without having to tear it apart.
So, yeah, we're not going to completely replace secondary color correction. There were always be those times where you do want to be super isolated with what you do or do have to attack a very specific part of the image where they weren't able to get light in on the day on set or that sort of thing. But for the vast majority of stuff, you know, if it's lit pretty well, you shouldn't have to carve up the image like that.
And not having to worry about that to constantly have to go back and check, oh, is the key noisy here? Oh, is it tracking through the whole length of the shot? Just saves a tremendous amount of time in color correction because you quickly with these primary tools, you quickly develop the trust that, OK, if I grade it well on this frame, that's going to translate well to all these other versions of the shot.
And you don't necessarily have to carefully QC them all when you're setting the look, you just know it's, you know, it's going to work. So you're definitely time savers for sure. I can see X grade adding a lot of productivity to people's workflows. ChromaGen is another new tool in V6, which to me is just like its own beast. It could be its own program. Can you tell me a bit about chromagen? Yeah, so chromagen is a look development tool.
So as opposed to X grade, which is very much for kind of shot by shot grading, you know, adjusting each shot individually, chromagen is a set of tools to help you develop a look that you might apply to a whole sequence or even to a whole show. And we developed the tools again in a human perceptual space so that when you're manipulating colors, you're doing it based on sound photochemical photographic principles and the way the human eye works.
And it's specifically designed to kind of be broad enough strokes that they do translate well across a variety of different shooting conditions and images. So you can't be too isolated with the controls. Like you can't do what you would do with a color key or stuff like that specifically so that it will hold up to a wide range of shots.
But what we found was that even with all the color tools in Baselight, there are some color operations which are often a part of look development that aren't easy to emulate with current color tools. So particularly when you think about like emulating a film look, there's a lot of crosstalk that happens between the color channels. So like when you expose something that's green on film, it's not just affecting the magenta dye layer, which is where green is recorded.
It's actually bleeding into the cyan layers as well, a little bit. So we developed matrices and things like that which can emulate that process on a technical level. But artistically, that wasn't something there was really a tool in the color correction that can handle. So we specifically want to make sure, OK, we can adjust those kind of crosstalk effects between the color channels. But then also just kind of where do you put the bumpers on your look?
So like if you want to be able to know that the look for this show has a certain red quality to it. Like you always want to skew towards a certain red. And that could be for a commercial where it's like a particular brand color you want to make sure you're staying on. Or it could be just when you're developing a feature film or TV show or something like that. Like there's certain kind of signature colors you want that look to have.
And that's something you can do shot by shot with Hue Shift and tools like that to kind of consistently push things towards the same flavor of red or towards the same flavor of blue. But ChromaGen lets you do that as part of your basic look development to say, OK, I want all my reds to look like this. And I want my dark reds to be this way and my brighter reds to be this way. Or I don't want to allow my reds to be too saturated, but I do want my blues and greens to be saturated.
So you kind of can define the color space on which you're going to grade. And then once you've set that look for the show, you're just kind of grading, doing your adjustment shot to shot underneath that. But the broad strokes of the look are then set.
But yeah, it did end up becoming almost a color corrector within a color corrector because it's got a whole bunch of tools for how you set up that color space that you're going to work within or that color palette really that you're defining for your show. But the idea is that these are the stages, aren't they? The stage that you have with all of the different sort of would you call them tools within the stages or options?
Well, I mean, we could we could just as easily call them layers, but we didn't want to confuse them with the layers that are already within your your main grading stack in baselight. So that's what we opted to call them stages. And each one of those operators, I guess you could say, just kind of a slightly different way of tweaking the color and you can combine them in different ways to kind of cover the whole color space you want to cover.
Yeah, so I suppose, you know, where you might otherwise use a look or perhaps a LUT or, you know, create some curves that you're going to put across everything, you would use ChromaGen instead. And using those stages, insert different tools that are meaningful to the project you're on, like color cross talk or contrast or... There's a few that I'd never heard of before. Like, can you talk me through, do you remember off the top of your head what the stages are?
Because there's I'm putting on the spot here, but there was some that I was like, I wouldn't know what that was if I hadn't seen it here. Well, yeah, some of them are named just because we had to come up for the terminology, which isn't something that regularly gets talked about. So in some ways we're choosing the terms. They're not, you know, industry standard terms. I thought I'd missed a memory.
No, but yeah, a lot of it came out of, you know, for a long time in Baselite, we've had this look tool which allows you to apply preset looks. So if you want a Kodak film look or a Fujifilm look or a reversal as negative look or a bleach bypass look, you could apply those. But they were like look up tables. They were kind of, you know, could turn them on or off. You could vary the strength a little bit. But that that was it.
Like, if you really like the film look, but you just didn't like the way the greens came out, you wanted the greens to be, you know, more yellow or less yellow or whatever. It was kind of hard to tweak that look because everything is all tied together in this one LUT, basically that gets applied or not. So with ChromaGen, we're able to reproduce those same kind of looks. We can produce different film looks.
We can produce different photochemical processes or just, you know, other other creative looks. But now they're broken down into steps so you can tweak just the part you want. So you can say, yeah, I love the Kodak film look, but I don't want my highlights to be warmer. I want them to stay neutral. Or I do want my greens to be a little more green and a little less yellow.
So you can tweak just those parts of the look that you want rather than it being kind of all or nothing thing that you would get with a LUT. And are they going to be presets? Because I feel like if it was a preset, I'd be reaching for it all the time. That's that's the way actually I recommend most people introduce themselves to ChromaGen is, yeah, we have a set of presets. Some of them match the existing looks in base light. Some of them are totally new.
But yeah, I think that's that's certainly the way I learn and am able to wrap my head around things often is just by kind of just looking at examples of what works and then picking it apart and seeing how it got there. So yeah, I definitely recommend people look at the presets and then just kind of just turn things on and off, adjust the knobs, see what each step is doing. And then you start to really wrap your head around it. I mean, that's another one that's well worth having a play with.
I know it's still in beta, but there are various places that are running the beta that you may be able to get your hands on it or, you know, of course, there's a lot of materials out there on the FilmLight website about, you know, showing you what it's going to do. And who made that one? Whose baby was that? It's Daniele Siragusano. Siragusano. So it's Daniele Siragusano, who's one of our image pipeline engineers based out of Germany.
He was really the chief architect on that working closely with Richard Kirk, who's another one of our color scientists in London. So, yeah. Wow. Okay. So I imagine that he's got a range of resources that you could look at. Is that I seem to have seen a couple of things flowed into my inbox that he's doing talks about it. Yes. Yeah. So there's a number of videos already.
And certainly once version six launches, there'll be a bunch more resources out there for to kind of explain what all those different different steps in the tool do. And you mentioned Richard Kirk as well. And just a side note, if anyone is interested in color science at all, which you probably are, if you're listening to this interview, his book, Color Sense and Measurement is well worth a read. And I believe you can actually get a PDF of it online.
So, yeah, there's information on that on the Filmlight website. And it's definitely a great read because, yeah, there's very few, as I said, early on kind of resources for color in our industry. You know, a lot of them are focused on pre-press and photography and stuff like that. So it's great that he wrote a book that's very focused on color management for film and television. And it's actually really readable. Like I was surprised. I thought, oh, my goodness, what am I getting myself into?
You know, I'm going to be hit with a bunch of equations and my eyes are going to glaze over real quick. But it was actually a page turner. So he's got a really good kind of personality behind it. I think it's not dry at all. So it's so well worth a check out. And I saw that recently you did some talking about cloud grading, grading in the cloud. Is that something that's new for version six? They're going to be new tools for that.
Because I haven't actually watched it, so I'd be really interested to have a recap on that one. Yeah, no, it's not new for version six. So, yeah, we've been able to run Baselight in the cloud for a little while now. Initially, it was just the render engine. So if you just wanted to render or transcode a bunch of media in the cloud, that's been available for a long time.
The new thing that we partnered with Amazon almost a year and a half ago now was to be able to run actually a fully interactive Baselight session in the cloud. So all the storage, all the image processing, everything was done in the cloud and you're just getting the output streamed down to you. And of course, the big challenge for color correction is that it needs to be super interactive. If you're adjusting a trackball or a knob, you need to see the change to the image right away.
If you're spinning a dial and then a second or two later, the image changes, you're going to constantly be like overshooting or undershooting or not knowing what you're doing. So that was the biggest thing we really had to overcome was getting that latency down. And we basically done it. So we now have high quality image streaming with very low latency out of the cloud. Color grading in the cloud on its own, I think, doesn't make much sense.
Like you wouldn't do a production the traditional way and then push everything up with the cloud just to color correct it. The economics and the time involved to do that doesn't make any sense. But as studios are looking at doing more and more of their work in the cloud already, there is a lot of VFX work done in the cloud. That's pretty standard.
But if you're also doing your editing and your sound mixing and everything else, then it makes sense to keep it in the cloud for color correction because it's possible now. And ultimately to the colorist, it shouldn't really feel any different. They shouldn't know the difference between whether they're working on a baselight that is in the machine room down the hall or in some cloud data center hundreds of miles away.
And so that's what we worked on is just making it possible that you could work either way. And that's pretty much all working. So there's definitely, again, some economic things you have to consider with how you deploy the resources in the cloud and the time to set it all up. But, yeah, the technology is there and ready. That's amazing. I just do use a virtual machine to color grade seems like another another world.
I mean, one thing which is unique to base light is that it's turnkey system and you've got certain configurations that you support and you know that that is going to work. So so those are the ones that you sell. Is it the same in the cloud? You found the configuration now that works and that's the one. Yeah. So it's still kind of preset configurations. And a big reason for that is really just for the support so that we know if there's something that's not working right, it's on us.
Whether it's a hardware issue, a software issue, a config issue, you know, you have one number to call and we'll we'll sort it out. So I mean, that support is pretty incredible. While I was in the office with you that week, I was seeing the phone ringing all the time and you were plugged into the community and helping people in real time. And the experience of using a base light is is having that plug in to this community of people who, you know, you're not talking to somebody who's selling it.
You're talking to someone who's actually made it and who has a stake in it. You know, like I remember actually an anecdote of one of my friends saying, oh, I was complaining to Bob about something to do with the control surface. And he's like, oh, I actually did some of the electronics and now what is it that what's going on? You know, can you talk to me at all about that support aspect and just being there for people?
Yeah, I mean, that's a huge part of what we do and a huge part of why I like working for Filmlight is we are, you know, a relatively small company who's focused on color correction and focused on this industry. And it's it is a very innovative industry. So as we said, there's new cameras, new display devices coming out all the time. But also filmmakers are looking to push things in different directions and different ways.
So we're constantly involved in new projects and new things and new ways of doing things. And yeah, so it's great that we can be a resource there to filmmakers or, oh, hey, someone just walked in the door with this camera. I've never seen it before. Is there anything I need to know, any kind of caveats to what we could tell them? Well, here's our experience here that we've discovered. And, you know, if you learn anything new, let us know, too.
So sometimes we are acting as kind of like a little bit of a middleman to be able to get firsthand experience with with new cameras and new display devices and then share that with other people who are who are working on them later down the line. And also, like, you've got all of these backwards compatible tools, like you've got all of these legacy tools there. So, you know, I imagine that it's not just the new technology that you're supporting.
Like, if I was to unarchive something from 10 years ago now and try to get my head around it, especially if it was somebody else's job, I think I'd go crazy. So, you know, is that also something that you're supporting people unarchiving old jobs as well? Yeah, that's kind of one of the founding principles with the way Baselite was built, is that you can go back to any Baselite you've seen, no matter how many years ago, and upgrade it.
And it'll still work and produce the same image in the current release of Baselite. So, yeah, I was actually just talking to someone a few days ago who had an archival project from I think it's 12 or 13 years ago, they wanted to bring back online and they were able to. So they were able to recover that job, see exactly what was done to the image and then tweak it from there for the new new pass that they wanted to do.
Wow. And is it true that they would be able to use the databases saved undo's to go back and undo everything in that 12 year old job? Yeah, exactly. Again, that's just the way Baselite stores its project files. It's basically a history of every key press and every change you've done. So you can literally hit undo and see every step that was made to get to where you were.
And that could be great for us in support too, because if someone kind of paints themselves in a corner, we can actually undo and see every step they went to get there. So we can actually pick it apart and see what was done. So what does your typical day look like? Are you fielding support calls a lot of the time? Are you working on special projects? Yeah, all of the above. I mean, I think I'm pretty lucky in that I don't have a typical day. Sometimes I'm doing training with the colorist.
Sometimes I'm just kind of jumping in on support calls, picking up the phone with my background in color science. Like I tend to focus more on the support calls that are much more kind of color workflow oriented or color calibration oriented, where we have other people on the support team who will focus more on technical issues integrating with storage or things like that. And then sometimes it's just completely new projects.
I actually just got back. I was in Las Vegas last weekend for the premiere of Deron Aronosky's Postcard from Earth at the Sphere in Las Vegas, which is this giant LED wall. That's I forget how many stories, but it's like a 16000 seat venue with a giant LED wall that completely surrounds your vision. So there were some unique challenges in getting that project over the finish line for color calibration and also just dealing with the geometry of the space and the screen and all that.
So yeah. Wow. Oh, can you tell me more about that? What was your involvement in the sphere? Because everyone's talking about it with U2 playing there. Yeah. Yeah. So some of it under NDA therefore, I can't go into too much detail. Of course. Of course. But yeah, so the colorist for the first kind of traditional film that was produced for the sphere was Tim Sippen and Andre Rivas was the assistant colorist on that.
So he's supporting those guys and what Sphere Studios, because they built a studio specifically to produce content for this this venue, Sphere Studios built their own camera called the Big Sky Camera, which had a fisheye lens on it to be able to capture an 18K by 18K source image to put into the into the sphere.
So, yeah, there was making sure that the color from that camera was coming into base light correctly and then developing the color profiles for the venues so that when they project itself on the screen, it had all the right colors. And of course, we can't do so that the ultimately the playback in that venue is a 16K by 16K image at 60 frames per second. So it's a huge amount of data being thrown up on that screen. Obviously, you can't color correct that in real time.
So most of the time, Tim and the team were working off of proxies just on a standard Sony X310 monitor. So again, the color calibration, making sure that that monitor matched the color of the venue and then also just dealing with the spherical mapping to be able to look at the image in different ways on that flat square to get an idea of what it would look like when you're surrounded by it in the sphere. So, yeah, a lot of fun and fun new challenges in that. Wow, I can't imagine.
Just I just had this like image in my mind of them just grading and looking at the sphere, just like standing out there in, you know, Las Vegas on the strip and with a little desk in front of them. Yeah, we we we hope to get to that point at some further down the road. But also, you know, of course, that that the sphere is now very much in demand. There are U2 concerts in there. They're playing the film regularly there.
So it's not unfortunately a space you can just kind of take over for a few days and do what you want. And it's it's a, you know, it's a live venue. So, yeah, it's just building the workflow around it. It seems like something that belongs in Dubai. It's just such a bizarre, such a bizarre thing. Was it interesting to see in in life, in real life? Like what what did it look like as a viewer? Yeah, no, it's definitely it's a very immersive environment, very unique.
You know, in some ways, it's an evolution of previous dome projections like OmniMax. IMAX have this. They still do actually have an OmniMax, which is basically IMAX film projected on a dome or, you know, some planetariums will have digital shows that they can project. So in some ways, it's an evolution of that. But because it's a LED wall, you know, there's no optical artifacts or anything about you have to deal with in projection. It's just the light directly coming at you.
It's much higher dynamic range so it can get much deeper blacks, much brighter whites. And of course, the resolution is pretty insane. So, yeah, when you're sitting in that venue and it's completely filling your peripheral vision, like any time the camera moves, you feel like you're moving. Yeah, it's just totally immersive that way. And when you're yeah, when the so Postcard from Earth has shots from all over the world and all these different environments.
So, yeah, like when you're in the underwater sequences and the waves are over your head and the fish are down in front of you. It really is, you know, much more immersive than a traditional theatrical experience. So it's really unique. I'll definitely be heading to check it out next time I'm over. That's for sure. And you guys have a system that can deal with that that large file. Yeah. So, I mean, that actually was nothing new for baselight.
We had always architected baselight to be kind of resolution independent and we'd always, you know, so whatever resolution image you could throw at it, you know, we can't promise how fast that it will be. But it will at least, you know, produce an image and work. So, again, a lot of the real time adjustments had to be done off proxies. But, you know, there are 4K proxies. So 4K is the proxy and 16K was ultimately what was what was rendered. Wow. I bet that took a while to render, but it's OK.
You don't have to tell me how long. Well, look, I reckon I've probably covered most of most of the questions that I had for you. But I just find it super interesting what you've been able to do in your career and how much everyday stuff that we take for granted, you know, has actually been the result of committees and companies that you that you've worked in.
And I think really it's it's, you know, people who are doing all of this work in the background to make sure that things really look the best that they can. That gives us colorists the ability to do what we do. So I can't thank you enough for your contribution to the craft and just for for being available and having a chat and, you know, being at the other end of the phone support. It's it's awesome. So thanks a lot. Well, thanks.
Yeah, that's really nice to hear. And yeah, like I said, it's I if, you know, the creative process is just able to go smooth and you don't even think of that technology. Then, you know, then we've won. That's that's really our goal. So yeah, it feels a bit like that in coloring, too. It's like if no one notices what we've done, then we've done a good job. So we're very self-effacing in this field. Yeah. Well, thank you so much.
Peter Postma, I'm Kayleigh Bateman from Mixing Light, and I'll see you next time.