The Man Who Wrote The Book on Compositing with Ron Brinkmann

The Man Who Wrote The Book on Compositing with Ron Brinkman | TVAP EP56

When I embarked on my compositing journey, Ron Brinkmann's book, "The Art and Science of Digital Compositing," proved to be an invaluable resource. It provided a wealth of visual information and included a DVD with footage for hands-on practice. Even today, this book remains a seminal work and continues to be relevant for artists, thanks to its emphasis on developing the skill of visual perception.

Therefore, it was a tremendous honor to have Ron Brinkmann as a guest on our show. He served as the Compositing Supervisor at Sony Imageworks for films such as Speed, Contact, James and the Giant Peach, and more. Interestingly, despite being widely recognized for his expertise in compositing, Brinkmann initially started his journey in the field of 3D as a demonstration artist for Wavefront. (Wavefront later merged with Alias and played a significant role in the development of Maya, a renowned software).

Subsequently, Brinkmann co-founded Nothing Real with his colleagues and created Shake, a compositing software that held sway in the industry for many years, particularly in high-end compositing. (I believe it was most recently used by Dneg for their work on Inception).

As always, if you liked this episode, please 👍 like, 📝 comment, 📢 subscribe, and 📣 share!

Listen to all episodes on our website
https://www.thevfxartistspodcast.com/

You can watch all our episodes on our YouTube channel.

• Subscribe & Watch all episodes on our YouTube Channel
• Visit our website

Thank you for your support! We appreciate you!

Transcript

I just couldn't have written a book that didn't talk about why you're doing what you're doing. What's the point of doing this compositing stuff is you're trying to make stuff look real. How do you even talk about that without talking about learning to see what real looks like? But speed was kind of... it came in and I remember reading the script for it. I read through the script, the fast read is a fairly short script and it's mostly action, right? I sort of read it like, "God, there's not much here, right? I mean, I see the action stuff." But in terms of a movie, who's going to want to watch this? Because it didn't feel like there was a whole lot there. And of course, then when the director, Yandeban, came out and put the whole thing together, it was this amazing action movie.
 When When you're building something that's complex, there's a compositing tree, the ability to be able to kind of narrow down on some small pieces of it and understand it. Putting in things like being able to group a bunch of operations together and collapse them into something that looks like a single operation, just so you can understand.
 I I mean, you know, you've seen anybody that's done big compositing scripts that can just be enormous. Hi, so welcome to the VFX Arthur's Podcast. We often say some of these experiences written the book on something and in fact, our next guest, Ron Briggman, has written the book on compositing. So, welcome to the show, Ron Briggman.
 Great Great to be here. It's been a little while since I wrote that book, but hopefully I keep hearing that at least a few people are finding it useful still, so hopefully so. Yeah, I think there's certain things that are timeless. I mean, you kind of wish there'd be a third edition, and I think it'd be cool to see what the kind of third edition might be. But we'll come to that later in the show. Maybe first I'd just like to kind of roll right back and sort of talk about how you yourself got started in VFX.
 I I mean, I'm going to have to, we need to put some sort of a quota on me saying, I remember when, or back in the day, because it's been a few years now. But yeah, I mean, I came out of university and bounced around at a couple of hardware and software companies, and I ended up working at a company called Wavefront, which I doubt anybody remembers it now, but it was one of the very early CG, you know, 3D CG.
 So So Maya, Maya before wasn't it? So basically, yeah, well before all that, there was back in the day, Jesus, I'm going to do that too many times, back in the day there was Wavefront and there was Alias, and they were the two sort of primary competitors in that 3D space. And at some point, they merged, or they really spot them, and then all of this bought them, and somewhere along the line, that all turned into Maya.
 So So yeah, that's the very short history of what those packages were. At any rate, I was working for Wavefront mostly in sort of a, not really a sales, but a tech support kind of a job, and there was this little company called Sony that was like, we need to have an in-house effects group, and you know, we'd like to look at your software. And so I came out there, and I think they were already planning on this. As soon as I sort of showed up and started showing them stuff, they said, "You know, what really needed somebody knows how to use this, and we're just starting off." And they hired virtually nobody at this point, four or five people, and so they offered me a job. And it was definitely diving into the deep end of the tank we had, especially the very beginning, nobody that had really done very much in the way of film production. They started hiring more people than doing film production, but the guy that started it was sort of a big dreamer, but didn't quite, he had no visual effects background. He was just sort of said, "This looks cool," and he managed to sell the studio on the idea. And so for many years there, we were working out of what they called the TriStar Building, which was one of the production offices at Sony, and just started doing visual effects. And we fortunately hired a few more people that knew some stuff, and we made up a lot of stuff ourselves that we were always like, "Oh, I've come up with this great idea," and then later we realized, "Yeah, you know, I've always been doing that for ten years, but good for you to come up with that idea too." So, yeah, so I ended up spending a number of years doing film production at Sony, which became Sony ImageWorks. That was a name that came on after a year or two after we got started.
 So So that was where, so that was James and the Giant Peach, which was one of the shows you... Yeah, no, James and the Giant Peach was probably about midway through my time at Sony. I mean, the first shows we worked on, I think my first credit was on the fabulous movie Last Action Hero. Holy cow! Are they even living with me?
 With With Arnold Schwarzenegger. It was, you know, Arnold was at the peak of his powers then, and this was the big movie from Arnold Schwarzenegger, and there was all the talk about it. And then there was some other little movie that came out at about the same time called Jurassic Park, which got a little more notice than Arnold's movie did. Damn it!
 But But yeah, it was... that was my first one. There was a few different movies there. One of the earlier ones I worked on also was Speed with Keanu Reeves and doing things like making the bus jump over a gap in the freeway.
 So So yeah, it started off as kind of pure visual effects stuff, and even though I'd come on board sort of as a 3D person, I knew the Wavefront software, so I knew 3D a lot more than I knew 2D. I'd done some compositing stuff. There was a really early on compositing package called, God, I think it was called Video Composer at the time, and it was literally...
 Initially, Initially, it would only do video resolution, and then they added the ability to do something higher resolution than that. But I hadn't really done a whole lot of compositing to that point. I'd done mostly 3D, but as a new startup kind of studio, most of the work we were getting was little kind of 2D kind of stuff.
 So So I kind of learned how to do compositing there, really, as did all of us. I think at that point, I sort of learned how to do it, and then started teaching other people how to do it. And that was kind of the genesis of showing people how to do compositing related things, I guess. And you were also developing a compositing software there as well, right?
 So So eventually, I mean, the timeline was... I was working at Sony, I was purely a production guy, eventually to the level of supervising the visual effects stuff, was the CG soup. And we had in-house... So just how... Go ahead. Just to get... what sort of period are we talking about? Like, kind of getting the chronology sort of... What year was it? I don't know.
 So, So, from when you started to turn it to Sony, and then you sort of over a certain amount of years moved to Supervise, I was just curious. It was sort of a little on and off stuff. I mean, like, there was a couple really early... I mean, I think Speed was sort of my first... Is that right? I'm not sure what my first Supervisor credit was. I think... I know on Speed was probably the first one where I could really say I earned it.
 It's It's funny because Speed was... You know, there was a lot of little movies that came in. As part of the studio, we weren't necessarily... We were sort of the pickup guys in some ways, especially early on. It was just sort of like, "You guys have people that can fix this sort of thing?" Or, "Can you do a little bit of this or a little bit of that?" Just kind of proving what we could do and learning how to do it. But Speed was kind of... It came in, and I remember reading the script for it. I read through the script. It's a fast read. It's a fairly short script. It's mostly action, right? I sort of read it. I'm like, "God, there's not much here." I mean, I see the action stuff, but just in terms of a movie, who's going to want to watch this? Because it didn't feel like there was a whole lot there. And of course, then when the director, Yann de Vant came out and put the whole thing together, it was this amazing action movie.
 But But that was probably... So that was a few years into it that I sort of started supervising people. There wasn't a clear growth path. These days, I think most studios have a pretty clear growth path for people where you start off and maybe you start off grabbing coffee for people. But you certainly start off doing fairly straightforward, simple tasks. And you just kind of show up and you show that you're capable of doing whatever people toss at you and you're willing to do it, the right attitude. And if I ever talk to people that are just getting started, that's the thing I really try to underscore. Yeah, you've got to be smart and know your tools and really learn this stuff. But it's so much about... This is true of any industry, right? It's so much about working with the other people and making sure you keep your bosses happy.
 And And it's doing great work, but it's great work in the context of making sure you're doing the great work that people want you to do. So a few years in, I was starting to supervise bits and pieces. And for a long time there, I was back and forth, I would do shots. And most supervisors would still be doing shots as well.
 But But to your question of sort of when the software was coming about, it really wasn't... I mean, we were always developing in-house software at the studios, as any studio would do. And so I spent a lot of time giving feedback on a variety of the tools. But I wasn't anything more than just one of the artists sort of saying, "Gosh, it'd be cool if I could do this." Or, "It would make a lot more sense to me if I could do this with it." Kind of thing. And I think most artists, at least technical ones, I mean, I come from a fairly technical background. I've got a degree in computer science way back in the day. So you can kind of say, "I understand how this could work, and here's an idea of what we could do." That kind of thing. On Flickr, you had a lot of index cards with the software GUI on them, which I thought were amazing because there was all these hand-drawn sketches of the whole interface. Yeah, back to you. The software, which in case people are unusual is Shaped. Yeah, so Shaped. So I'll tell the rest of the story then. So after a number of years doing production, I just finished up a big show, Robert Smex's Contact, with Jodie Foster in it.
 It It was a great show. It was a hard show. That was right after Ken Ralston had come down from ILM to be the head of Sony Image Works. So he had convinced Robert Zemeckis, "Yeah, we can do this work, do this level of high-end work here." But it was a big show, and it was definitely stretching Sony's capabilities at that point.
 A A lot of hours on that show. Boy. And it turned out great. And the movie turned out great, but it was one of those things where, "Okay, I'm going to take a little bit of a break after this one." And sort of in the interim, a couple of guys that had, well, sort of one of our lead programmers had decided he was going to go off and start it up. And he went off with one of the other animators, and a few other guys were kind of getting ready to be done with production for a while as well.
 And And so that was a company that eventually called Nothing Real. And the core of it was originally just a set of command line tools for compositing called Shake. And they kind of convinced me to come on board. Like I said, I was just sort of finished up with contact and a little burned out with doing production. And so they had convinced me to come on board and sort of spec out, number one, how a user interface to a compositing package would look like. And then also just the ability to talk to customers that, you know, from the perspective of somebody that knows how to do it as opposed to just, "Ivery Tower, we're a developer and we've got this compositing package, what should it do?" kind of thing. So it was a good fit because, you know, I had a technical background in coding from way back when, although it's been forever since I've written code.
 And And I had a background doing visual effects and doing compositing. And I knew the guys that were doing it. And I'm somewhat familiar with what we had built in-house that was pretty similar to what ended up being Shake. So yeah, back then there was no Figma or anything that you could sit there and mock up user interfaces really quickly with.
 So So it was all, I mean, you know, I drew it all by hand and me and a couple other guys and then the developers pretty much coded a whole user interface layer as well by hand. So these days you'd be gluing a lot of pieces together instead.
 Amazing. Amazing. And so Shake, I mean, a lot of our listeners have never used Shake. It was the first compositing software I used and that I actually got a job with. I mean, I used After Effects before actually. But the first thing that I actually was employed to use and I was just at the tail end when Apple killed it. So Apple bought it and killed it. It was just a sad end of that software. But the question I had is, it's very similar to Houdini in terms of the interface, although Houdini is a 3D app, but in terms of the way you connect the nodes and you look at them. And I was wondering who was inspired by who. Oh boy. Good question.
 I I mean, Houdini was around before Shake came out and I had certainly used Houdini. I used Houdini a lot on content actually doing some of the space stuff at the very beginning of the show of the film. There's this giant fly through the universe kind of thing. And we used a Houdini quite a bit on that.
 But But I mean, before that, node graph kind of stuff was around in Maya and that sort of thing too. So it certainly, the node based compositing certainly wasn't something that we had done ourselves. The idea was around. But it just sort of makes sense. It's one of the things that just drives me nuts when I have to use After Effects is that it doesn't really embrace that sort of thing. And you get these stacks and stacks of stuff that's just for me really hard to sort of see what's going on half the time. So it just sort of made sense, I guess. And I think all the developers knew that that's what it needed to be too. It's just the only sensible way to kind of represent the kind of complexity that you can have in a compositing script of how things need to wire together.
 Well, Well, I said the similarity. It was also the fact that it's node graph, but it's got the similar idea that you click on the button on the left to view the output of that node that was separate from the selection. Yeah, all the little... Which is quite different to how Nuke does it where it's a viewer node. Yeah, I mean, it's... Yeah, I couldn't tell you where exactly all of these little things came up from originally. But it just makes sense, right, that you need to understand the flow of it. I mean, a whole... I mean, I could talk for hours about, you know, my personal opinion for how people need to interact with software and the ability to understand. You're building something as complex as a compositing tree. The ability to be able to kind of narrow down on some small pieces of it and understand it. And putting in things like being able to group a bunch of operations together and collapse them into something that looks like a single operation just so you can understand. I mean, you know, you've seen anybody that's done big compositing scripts. They can just be enormous and ridiculous. And the thing that Shake was able to do and packages like Nuke can do and all that is be smart enough with their kind of memory management where you can do it all in one big script. Because there were other tools prior to that, but you'd be doing so many little bits and pieces and pre-composites and, you know, you get one little piece working and then bake that out as a layer and save it off and then bring it in. But if you realize suddenly that something was wrong with that, you go back to your original process and you have to put everything through it again.
 So So the ability to have it all on one big script, you know, sometimes for better or worse, but the ability to have it all on one big flow graph was pretty key when it came to doing serious production that was efficient. And that's kind of what it's all about, right? One of the famous traits of Shake is that there was no undo or no reliable undo. I was kind of curious about that because it was right to the end. I mean, it wasn't just nothing really got bought by Apple. Some other platforms were used on huge shows. You know, I think Dine last year's are on inception with their in-house one and all through this entire history. So in one sense, the node graph kind of replaced your undo because you would copy the node and you had a copy of where you were if you worked on something. But it's kind of interesting that this production level software still shocks people that you couldn't really just press control Z and get back to where you were. It shocked me too. It certainly was not a design feature, I'll tell you that.
 Oh Oh God, brings back nightmares of me fighting with the developers and saying we got to put it in there. And I don't even remember what somebody said to me. Well, there's no undo in life is there? It's true, but doesn't mean I don't want one. So yeah, I mean, at some level, you know, we were a small group of developers and we were constantly chasing, you know, what's what's the feature that needs to be in there? And you could look at undo as a feature, but, you know, there were ways around it and careful composers could do exactly like you said, where you just make a copy of some nodes and stick them off to the side and know that you're going to do it. But no, I mean, it was terrible. I mean, I'm not going to I'm not going to apologize for, you know, any sort of grand scheme of we intended it to not have an undo. It's just for what I can't even remember what it was. But there was some kind of deep technical issues with keeping track of it. I think it had to do with there was too many different places where, you know, do you put the undo at the level of if I change a slider, do I undo it or if I change the topology of the graph, do I undo it? It was just I don't remember what all it was, but I'll apologize many years later for everybody that had to deal with the fact that they couldn't undo some horrible change they made. But to be fair, it bit me plenty of times myself, too. So, you know, I I guess it does. It does teach you good hygiene for saving regularly. Absolutely. And to be honest, the the idea of undo for life is is a high concept movie. Yep. Yes. I work on that. And so contact, since you put up, did you work on the famous mirror?
 Not Not really. I was there when they shot it. It was I mean, I remember when actually I was there when they shot half of it. So we shot in two pieces. Now it was it was that was done in flame. It was all done in flame at that point. Tracking was we didn't have great tracking tools. Flame had much better tracking tools. And that shot was all about, you know, getting the getting the match move to work right and everything. And a really excellent flame artist, Mark Holmes, did that shot. I remember I remember when he put it together and I looked at it and I'm like, oh, that's that's going to be the shot. People remember. So and it was it's just one of those things where I mean, I think everybody was just sort of like, you know, you're going with it and you just and you just sort of stop. You're like, what did I just see? And so, yeah, it was a great shot. But no, that was I can't take any any credit for that other than being one of the people, one of the first people to say, wow, that's a great shot. But yeah, I mostly I mean, in contact, I mostly dealt with the the opening sequence, the giant pullback through the galaxy, which was a great, great fun and a lot of miscellaneous stuff. And then a lot of the stuff at the end when they're kind of on that surreal sort of beach and getting some of that to come together. There was a lot of flame, inferno, whatever it was at the time. Then there's well, this is flame even a thing. Is it inferno now? What the hell do they call it now?
 Flame Flame is still going on flame. Weirdly, it's the one survivor of all that they had all these things, flint and smoke and they all died. Now it's flame. It's biggest in editorial. I was actually using it at the mill, but not for compositing for the timeline. So you'd give you an ad, you would conform it into flame and you would put your shots in your movie, posting in you interesting and Warner Brothers. I know use it to conform movies. So yeah, to conform level online editing capabilities and the kind of level of compositing that you can do within online. It's still rocking and rolling. I mean, it's really sort of interesting. Yeah, I mean, it was, you know, I mean, it was an extraordinarily expensive piece of kit. It was, you know, especially the time when I was running on those high end silicon graphics machines and all that.
 But But it has some cool stuff and just the interactivity of it was amazing. You know, it was fun. You know, there was this brief period of time there where the flame artists were kind of the gods of the post world. I mean, more so in commercials, but just because of the level of productivity and you know, you'd get the guys that partnered up with their favorite commercial director and they were lauded and tossed all a whole lot of money back then. I mean, I remember trying to hire flame artists at, you know, compositing visual effects studio kind of salaries and they were like, do you realize how much I can make doing commercials? So it was a it was a tough. They were getting transfer fees like like like athletes. Yeah. Yeah. The company wouldn't just pay them a big salary. They would pay the company that would take them from a transfer fee. Yep. Yeah. I mean, I remember I was I was dating a gal. It was a flame artist at the time and she was just like, she was paying for it. Exactly. But yeah, it's interesting too to me how flame like you said that, you know, sort of kind of the editorial side of things, how much compositing work is keeps getting pushed to editorial. I mean, it only makes sense, right? And if you make them make it easy enough that you have somebody whose skill is editing doesn't have to know the ins and outs of compositing. There's so much you can do if you have the right tool. I think that's only going to continue. I mean, I'm sure it's I haven't looked at, you know, the latest in in avid or whatnot. But you can definitely see that a lot of this stuff used to be positive. Yeah. Right. Yeah. I mean, Resolve has this whole worldview and everything, right? They built Fusion and stuck it in there basically. So that was your main. That would have been your main competitor when you were running Shake. Yeah. Fusion was definitely one of the fusion was one of the ones we always ran up against. Didn't we tend to. You know, people tended at the higher end to buy Shake more often. But Fusion, you know, stuck around forever and they were they had a lot of cool stuff in it, too. So I feel like where Shake was strongest is actually when I was learning compositing. It just was just made so much sense. And that's kind of why I think it ties. And the other thing about Shake that I was that I really enjoyed was the manual because up until then I read things like the order this the Maya manual, the After Effects manual. They just were like incomprehensible even for simple operations. It was clearly written for engineers, by engineers, the audience being other engineers. And, you know, Shake was like having a chat with your mates and the part about how to composite. But mates that are really good at compositing, you know. Yeah. And I mean, so that the whole manual was pretty much written by my dear friend Peter Warner, who unfortunately passed away last year. But he just an amazing, amazing writer. I mean, a great artist. He was he was an animator that I worked with, but also just an amazing writer. And, you know, he started writing this stuff and that's really tongue in cheek kind of documentation style. And we, you know, we talked about it. I'm like, is this going to fly? And we're all kind of let's find out. I mean, you know, who cares, right? If people enjoy it and it gets the point across. So, you know, it always you look at the shake manual and you look at the book I wrote, which is rather dry in comparison. And but, you know, we sort of knew that that's what we were coming up with. Peter was writing this thing that was going to be fun to read. And I was writing something that was hopefully going to be around for a while, but also kind of wasn't quite quite as much fun. And he was just a funnier guy than I am. So that's probably a lot of it, too. But yeah, the manual was was I still have still had people like getting in touch with me on LinkedIn saying, you don't have a copy of that manual. I just want to go read it for fun. So it's out there.
 Which Which is which is not something you would say about any other manual. No, I don't think so. That I've seen. I think, you know, I think we had to your point about how shake sort of made sense from from the composite. We had a huge advantage. You know, if you look at and I spent some time working for the foundry after I left Apple and looking at Nuke.
 And And there was talk at the time about, you know, should we revamp the Nuke user interface and try and make it a little more. I always found it slightly unintuitive. And obviously, that's because I developed shake. So it was nothing but intuitive for me. But, you know, I think Nuke, I mean, it was a piece of production software that just grew out of production. Right. And so every time they needed something, they would bolt something on there.
 And And I think it got a little unwieldy at times. I mean, it's an incredibly powerful piece of software. There's still stuff about it that kind of bug me with the way the user interface is laid out. But, you know, that's I can't really take that much credit for it because they have, you know, shake had the huge advantage of we put together compositing software. And then we got to start from scratch and build something from the ground up that was really designed to be a piece of software that kind of built on everything that we had screwed up the first time around or building production software at the studio. So a different kind of beast. But yeah, I mean, the other thing that we really tried to do with shake was trying to just make it really kind of thin at some level in terms of it was just a direct representation of what your your compositing graph was doing. It was just, you know, very direct representation of what you were thinking, how you laid it out. Yeah. And one. Well, actually, the new case conor to iteration has got a lot of shake like attributes. So for new users who may have never used or heard of shake, for example, if you're a new user, you'll be connecting masks, the mask input on almost all your great nodes and so on. And that's a shake feature that wasn't originally in you. You can use the pipe, the mats and have different mats and different channels. And then actually, oh, you just plug them out into the side. That's a shake feature. Yeah. And now they've actually recently they've just added the shake to remove your. Excellent. Excellent.
 That That is a new feature in like 13. It's coming for 10, 15 years later. But good. I mean, I think you just I'm curious if anybody listens to this podcast and is still using shake because I you know when it we talk about the the end end of days for shake. But, you know, at the at the end of it all, when Apple decided to kill it, we actually did provide source code to some of the bigger studios. You said you worked at the mill. I think the mill bought a copy of source code. A few, a few places, you know, got a copy of source code and ended up integrating into their workflow. So I'm curious if seems like a lot of the London studios in particular, I'm curious of any place like, you know, Dean Egg or mail or whoever still have bits and pieces of shake laying around in their workflow or is it just too far gone at this point? I think now, yeah, I think I think it's gone from what I understand from speaking to artists who were there and I wasn't there. But Inception was the last shake show at Dean Egg. Got it. And that was it. That was a pretty good song. I mean, you've got to say that is the last concept is the composite inception. I like it. You know, that's a good story. So I'll stick with that one. At this point, you've been doing 3D alias, you've gone to Sony, you've become a compositor, you've been training people in compositing, you've left and developed and promoted a positing software, which was the dominant one for a long time.
 And And then some point you left again, you did something new, you wrote a book about compositing. Wrote a book. Which? Yeah, I mean, the books. How did this? Well, the books started. I mean, technically, it started when I was still at ImageWorks and we did a SIGGRAPH class myself and a couple other guys, Jerome Chen and Gary Jackamuk, you know, kind of, you know, we were all sort of 3D artists. And so this is pretty early on. And we just did a course on compositing because nobody had really done a compositing course at SIGGRAPH, you know, and just given a talk about what is the basics of compositing. And so we put something together that was, you know, very broad overview. I don't remember how long the course was, but I think it was just a half day course and, you know, tossed in a few examples from whatever the state of the art was at the time. Somebody can go back. I'm sure, you know, the SIGGRAPH archives have the course somewhere on there. And I mean, I ultimately kind of fell to me to pull all the pieces together for that course. I did most of the good chunk of the writing on it. And, you know, so I had that.
 And And whenever, I mean, I was sort of thinking about doing a book at that point. It's just one of the things where, you know, if you've worked your way up and then you're to the level where you're supervising and you're hiring a lot of people, you're also, you basically become a teacher, right? You basically have to explain to them, you know, specifics of what this how this studio works, but also just a lot of learning for the, you know, the new the novices about how compositing works or how 3D works or whatever works. I just found myself sort of explaining the same things and the same basic concepts and not just compositing concepts, but just visual concepts, you know, how to how to shadows, you know, add up kind of things and, you know, understanding where the lights hitting to make your composite look real. And all of those all of those pieces that a good artist has to develop the eye for. And I just kept sort of saying the same things over and over again. It's just like, you know, some Sony needs to write a book about this thing. So it was right when I right when I left Sony and started doing the software company. I'm like, I should I'm also going to take this time to write a book on it. And it dovetailed nicely because same thing. If you're trying to sell software, more people that know how to use it and more people that are capable of using it to the best level, the better.
 So, So, yeah, it was just sort of, you know, my world was a whole lot of compositing from a couple different directions there for a while, which is always funny to me. So it's like I said, I kind of started out in the 3D world. But, you know, but I love compositing. I love just the fact that you sort of are the person that gets to put the final touches on on the shot and really make it, you know, make it sweet at the end of it all. I think is a nice place to be. I mean, the downside is, you know, shit rolls downhill, right? So you're at the end of the chain and inevitably it's the compositing guys who are pulling the all nighters to get the shot out the door. You know, the 3D artists have delivered all their elements and, you know, at least back in the day, it was where everybody else was done with their shots. And the poor compositing team was just sweating bullets to hit the deadline.
 I I think that's always going to be a pass. Yeah. Yeah. Pass the comp. Pass the comp. Exactly. And you know, this lighting is good. Yeah. Final. Final lighting. It's definitely not a fun comp. Yeah. Yeah. And one of the things I really like, I was looking at, I mean, I know that this is a second edition, so I presume there were some revisions, but I did really like just the structure of it.
 I've I've just got a copy open here just out of camera. But there was a fact that one of the first chapter of the game gives you an introduction and an actual example of compositing from way before digital, in fact, way before film, which is this photo image. Yeah. Composite the photograph. Yeah.
 But But then I like that once you get second chapter, and probably these are the I think the bits that are most relevant to this day and things like this chapter, whole chapter about learning to see and just observing colours, observing light, observing shadows, observing all of these things, all these kind of observational skills that, you know, often if I find a new artist as a junior coming in, that's what they're missing. You know, a lot of them have like really amazing software skills, and they know all these kind of crazy modern tricks, and they're getting an obviously more and more modern tricks you can learn. But sometimes, you know, they can't, for example, tell a lens just by looking at the, at the kind of arrangement of objects in space and things like that. There's like little observational skills that are just really, really, and I like the fact that that's like, right there front and centre of the house.
 And And it doesn't actually talk about any specific software. So the fact that this book is, you know, now, 15 years ago. Yeah, I mean, it kind of had to be that way that I just couldn't have written a book that didn't talk about why you're doing what you're doing, you know, what's the point of doing this compositing stuff is you're trying to make stuff look real and how do you do that?
 How How do you how do you even talk about that without talking about, you know, look, you're learning to see what, what real looks like? You know, ultimately, that's the skill that can't necessarily be taught. It's just somebody that, you know, and some people never quite seem to get it and other people are just sort of naturals at it. But just basic things, like I said, you know, how many times did I see people have a shadow of three different elements and they just layer them on top so the shadows add up even though there's only one light source? It's just like, no, you know, there's only one light that's being blocked. You don't get to have three times the shadows adding up on top of things. So, but yeah, I mean, I wrote the book intentionally to be this is just being lazy and greedy, right? I didn't want to have to be revising this thing all the time. So I tried to write something that could stick around for a while and it would be as timeless as possible. And that's intentionally why I didn't put any specific compositing software in the forefront of that. I mean, otherwise, it would have been a manual. You know, Peter was writing the manual for shake and, you know, in spite of the fact that the other people would kind of nudge me towards, well, don't you think you should be using shake and all these examples? I just want to have very genetic kind of things and it I think it's done well just because of that because you can still pick it up. And most of this, I mean, there's things in there that, you know, if I ever did do an updated version of this, it would just be, you know, here's the list of things that my book is that's not completely obsolete out of my book.
 And And I think that there's the basic stuff I think is still there. But yeah, I mean, who the hell needs to know how to do a convolution kernel anymore? You've got a button and you don't even have to think about it. You know, for me, the crazy thing is sort of what's what is this space going to look like in another five years with all of the the AI tools? Because, you know, up until a couple of years ago, everything was kind of iterative and building on top of existing concepts, right? I mean, you know, you can you can have a whole bunch of extra channels and there's all kinds of cool stuff you can do. And there would be new algorithms for how things could fit together. But, you know, these days, the sense of just tossing an AI, I mean, I'm seeing papers that are being published that are literally, you know, teaching an AI to do compositing.
 And And I say that not because I everybody is is worried about, okay, when is this going to take my job? And I don't I don't think that's going to happen right away. But I think what it's going to do is just give the artists, the people that know how to see the ability to be so much more productive. I really think that I mean, my God, the amount of time that we would spend trying to pull a clean blue screen or something like that, you know, in certain situations.
 And And it's not like it's trivial for every one of them nowadays. But clearly, I mean, you tell me I assume that most studios have ways of keying things that are starting to become based on AI that, you know, just make it easy. So much easier. I mean, at the moment, it's still it really, and it still feels like there's yes, you can do it faster, and you can iterate faster, but you're still using a lot of the brute force. The techniques aren't sort of massively. Like you said, there's an iterative improvement, you can definitely see everything's a little bit faster. You've got a few more tools, you've got a few more things to do with your edges and so on.
 Having Having said that, this is when the AI is rolling in. And actually, I mean, most of the compositing software is pretty old. I mean, all of it, you know, is built on a stop of structures that were around at the same time as shake, right? Yeah. Like Nuke, they used from around the same time. I'm sure, I'm sure, you know, yeah, there's no reason to do that. It's still around from those days. I do. It's kind of amazing in some ways that Nuke in particular has had as long of a run as it had, and nobody came along. I think that partially speaks to, you know, it's not that profitable of a space that somebody can come in and devote the resources to recreate or create a new piece of software that's as deep as Nuke is. I mean, there's just so much in there and, you know, it would take a massive team to try and come up with a new compositing software that's duplicating what Nuke is capable of doing and why bother because everybody's got their workflows based on it.
 But But I think, I mean, I haven't, I'm not completely up to speed on this. I know that like there's the company Runway is a company that's starting to produce pure AI kind of tools for doing similar sorts of work. But it's gonna take a whole different workflow. It's gonna take a completely different mindset for some of this. And I mean, that's gonna be fascinating to me to see where this all ends up. But I think the artists are still gonna have to be involved because, you know, the key thing that, I mean, I've spent a lot of time playing in Mid-Journey and it's cool. It's cool as shit what comes out of it, right? I mean, you look at that stuff and it just blows your mind.
 But But then when I'm like, oh, but I really want to tweak this one little tiny thing, this bug in me, I mean, forget it, right? You can't come up with a series of text prompts that says, you know, I need this highlight decreased by 10% and I want the angle of the head to be tilted by 5 degrees. You know, it's just eventually we'll get there kind of maybe, but I think it's gonna be more of this kind of world where, you know, you have control images and you may generate, you're gonna be generating. I mean, this is already what people are doing, right? You generate sort of a mock-up using a basic 3D posing package or something. And then you let the AI do some work on it and it's gonna be some back and forth. It blows my mind how much is gonna change in the next, you know, five years. I mean, the next five years are gonna be so radically different than the past 15 years in terms of the change that's gonna happen. Yeah. I mean, you see a better nuke, there's a copycat note. So essentially, one example is you can, you might do, you might work on a few frames and use that as your training data. Yep. Rather than tracking the patch across, you use that as training data and then you run the training on that shot, on other shots. And then this is where the cool thing about, of course, AI-based workflows like that is if you have 25 shots that are pretty much the same thing. I mean, I'm good omens, I'm just saying, because I've got the jumper on, but he had these contact lenses that are like cat eyes or reptile eyes. But while he's acting, they just kept rotating. So there were a lot of shots, I mean, they took 20 minutes, they were pretty quick and easy to do, but they were just, you know, all through the show. There were hundreds and hundreds and hundreds of these shots where his contact lenses have rotated around. Now, that's the sort of thing where an AI tool would storm through it, right? You would do 20 of them and you would train it and then you would let it rock and roll and you might even have a tool that you can just give the editor and just go apply this. I think that's a good point that the editor will start to pick up more and more of these things too. In some ways, it's a better place for some of this just because if you can have a single button that an AI can do something and the editor doesn't have to understand much behind it, it's... Yeah, I could see it ending up in editorial as much as anything.
 Or Or you can even resolve, I mean, there's the tools to remove, I mean, in SAP there's the tools to remove noises and so on and, you know, they're pretty crazy what they can do. So if you had the guys from Nothing Real or another startup came along and they said, "We want you to build a new compositing package, like now."
 Well, Well, my first plan would probably be to see if somebody would let me go sit back in production for a few hours and figure out what the hell the world looks like these days. Yeah, I mean, it's obviously AI-based. It's something where I think you would go back to first principles of what are the problems that are being solved, you know, what are the roadblocks that you hit when you're putting together a shot. Because part of me feels like it's probably not always going to be a compositing solution. I think you're going to find something else where there's this feedback loop between here's the image and... I mean, there's kind of two things, right? There's compositing is used for fixing stuff, right? There's a whole lot of, like you talked about, the contact lenses and all that kind of fixing level of work is sort of in one bucket.
 And And I think you just sort of need to figure out what kind of problems there are. And then the other half of it would be full-on image creation, you know, full synthetic imagery or nearly full synthetic imagery. And I think that's probably where it's really going to get weird, where you're going to be...
 You You know, we always used to joke back in the day of... because most directors and producers didn't really understand this technology that well back then, right? And so for them, the joke was always, yeah, you know, they want us to press the generate a dinosaur button. And it's easy, right? Just hand it over to the VFX guys. They press the button for generating a dinosaur and there you go.
 But But man, these days it's sort of like there's going to be some kind of crazy feedback loop, a lot of it may be on set, of I need to see my creature in this scene. And you can give me a quick rough comp that's probably fully articulated and animated and ready to go. And just, you know, having the AI produce images and you can sit there and spitball it. And I mean, I probably wouldn't do a classic compositing package so much as I would come up with more tools for, you know, on set AI based integration of putting cool shit on the page, right? Putting cool images on the screen and seeing how they all kind of fit together. And I think what's exciting is sort of the idea that, you know, we can actually, I'm probably more excited about where animated movies can go with all of this that, you know, it's, it's, and sort of where that line between an animated movie and a live action movie, it feels like there's such a hard line right now.
 And And you either say that looks real or that is looking stylized and, but even the style, you know, every Pixar movie, I mean, beautiful stuff, but they all kind of look the same. It's all kind of the, you know, the same basic style. And I want to see some cool ass thing that maybe looks like a watercolor, but, you know, has true volumetric sense to it. And that kind of stuff is, I would find that more interesting. You know, tell stories that need a different look. There's so many, I mean, I guess you got into the spider versus very different and the Mitchell's versus machines that are doing this pantily look. And then you had, you know, what I am did Rango, which used to have some photo real approach and made it look really weird and easy. So there's, there are elements of those kinds of different approaches going. And I think this Pixar's got this house style and it works. It's successful. So I guess they're going to stick with it, but it's good for other people to come in and do. Yeah, I'm certain. Yeah, and I'm certainly not, you know, Pixar is fully capable of pushing the envelope in a whole lot of different directions. But like you said, it's sort of it is a house style and it's sort of.
 So So it's interesting to see what you know, who's going to come out with something that's very different. I mean, into the spider versus a great example, though, where, you know, there's bits and pieces of that, you know, just little tiny snippets where they do something that's completely different style. And like, yeah, let's let's see more of that. That looks so cool. Another one was the Lego movie where, you know, they just went, there's no motion. You know, the motion below is bits of Lego. Everything's made of bits of Lego. All the explosions are bits of Lego. Right.
 It's It's a stop motion movie, but it's CG. It's all CG. Yep. Yeah, exactly. So yeah, so I just think that the the stuff is going to be first of all, it's just going to give people a lot more ability to spitball different weird ideas and different looks just based on that.
 But But then I think, you know, is this it's only going to be, you know, a matter of a year or so before you can put out full animated movies that are, you know, mid journey esque kind of thing. So we'll see. It's cool to me because it's it's certainly democratizing who can make these movies and who can do a feature like movie. And it's kind of trite to say it. And I think a lot of people say that, oh, yeah, you know, anybody is going to be able to make this in their basement. And they're not because most people don't have the visual sense to make something that really looks cool and don't have the storytelling sense to really put it together in a compelling way. But what it does do is take things a generation removed to where you can get a couple or a few talented people together who have their own specialties and can produce something that is going to be competitive with the stuff that the big studios do. But it's not it's going to take some very dedicated people to do that, you know, at least for the next few years. What do you think of if you've been following some of the sort of ethical debates that about day? I know it's I'm no expert. It's it's a really tough. I don't have a strong position on it. I mean, I totally get the sense that, you know, a lot of these artists, their work was used to train the A.I. to do certain things at full stop. There's no question about that. Right. And then again, you know, a lot of traditional artists looked at a whole bunch of other things. Two to develop their own style. I think it's kind of a false equivalency to say, you know, a human does it or an A.I. does it. And that's the same thing because it's not. I don't think that's a fair comparison. I guess my my hope is that as new training models, you're just new models are trained. Somebody is going to do the hard work of figuring out attribution and being able to trace backwards. You know, what parts of this went into that artists are going to have to be able to opt out and say, I don't want my stuff used for this this training data. Honestly, I don't think it's going to make much difference if certain artists opt out of something. And it's probably a kind of a lost cause. But having something where, you know, you could go in and say, you know, there's some micro payments due to a hundred different artists because this really hits in the zone of, you know.
 And And I would hope that maybe if enough of that's happening and the micro payments are in place that, you know, a prolific artist that has a unique style that somebody's kind of wanting to emulate. It just sort of algorithmically spits out that, hey, you know, here's a here's a check for some amount of money that kind of reflects what was spent to create this. So but man, it's a hard problem. So hip hop and sampling of old 70s artists and that kind of thing. Yeah, very true. And then that took a lot of lawsuits and it's still ongoing stuff. And I think there's been some weird decisions made that aren't necessarily fair on that one either. But at least people are, you know, are trying to. Yeah. Who was it that I mean, it was some artists that looked at the credits of another artist's work, you know, the credits on the album and how many writers had been credited. Right. And they were like, I don't understand how many how can this many people have been used to write the song? And they explained it that, you know, this is this is the samples that were used and these people are getting credit because samples was used. I think that's cool. I think I think that's fair. It's it gets a little unwieldy when you're trying to figure out where you put that credit list. We don't have album artwork anymore in the back of an album to look at.
 But But yeah, I just hope that somehow they can kind of be able to credit back to the artists that sort of define the style. I mean, you would think it would be able to even hyperlink you to the artists. Right. I mean, we're in I mean, we don't have the album, but we do have the ability to kind of find any artist or anyone in the world if you want to.
 And And so I guess as we sort of come towards it before we sort of move to the end, I really wanted to ask you what you're doing now, because I my last the last podcast I heard you on, which was a long time ago, which was on a fixed guide. You had left VFX and you had moved, I believe, to Amazon. It is. It works for Amazon for a while.
 And And now you work for something else. So I was curious about what the post post VFX and Ron Bookman is doing a whole lot of little stuff and trying not to have a regular nine to five. Yeah, I mean, I was at Amazon for a while. That was a completely right turn. It was I mean, make a long story short, I was at a conference.
 I I was kind of talking about visual effects, the sort of things, but it was after I'd mostly moved out of the visual effects world after I'd left Apple and some guy named Jeff Bezos happened to be in the audience. And we chatted afterwards and I, you know, I think I was doing the second edition of my book at the time and I was like, you know, it'd be cool if as a book writer, I could have some of these tools for writing and connect with the audience. So anyway, I wrote, you know, some half deranged manifesto and sent it over to Bezos because he'd given him his card and, you know, he writes back to me or I see email back, you know, we've been thinking about a lot of this stuff, which I later realized was probably his legal disclaimer. We've been thinking about a lot of this stuff, but we need somebody that can kind of implement some of it. Is that you? You know, it was a very Jeff kind of way of hiring somebody, I guess. So, so yeah, I worked at Amazon for a while doing things that were related to book technology, you know, and bits and pieces of it kind of touched on what you can do in a Kindle and all that kind of thing. It was, it was sort of a weird experience. Amazon has a very Amazonian way of doing things.
 And And I didn't last more than a couple of years there. I was just kind of like, yeah, this is not, I mean, it was a combination of things. It was up in Seattle and I'm used to sunny Southern California. So, you know, too much rain. But and I got family back in all kinds of, you know, a lot of good reasons, but also just the Amazon kind of methodology was was fairly rigid in some ways.
 And And I was just getting tired of working for big companies, too. So, so I asked her what I'm doing now. I mean, I've sort of, you know, bounced around. We mean, some friends had done some iPhone apps a while back, but those, you know, we've shut all those down because they never really got much traction. Most of what I'm doing is kind of talking to tech companies. I'm investing in tech companies, you know, kind of doing the angel investing thing with some interesting technology companies that are not necessarily even close to being, you know, compositing or visual effects related. But yeah, I mean, between just like I said, chatting with friends and bouncing ideas off and then doing a little bit of angel investing here and there is kind of what takes up my day. So yeah, keeping busy, though, it sure seems to be keeping me busy no matter what. Yeah, it sounds like it sounds like it very much like, oh, you just you could sit on a beach and do investments in the morning, but there's stuff I guess that keeps you keep your head most of the day. There are days that well, you know that I mean, I've got a couple of couple of kids that fill up the rest of my time. So, but yeah, it's kind of got a late start on having kids. I've got some somewhat younger kids that are still keeping me plenty busy. So between that and yeah, just sort of trying to keep my fingers and interesting things. That's kind of what keeps me busy. Sounds like a good place to be, though. Can't complain. I cannot complain.
 Very Very cool. So as a closing note, I mean, you did actually touch upon some advice, but what advice would you give someone starting out in VFX or someone just leaving VFX and thinking of moving on to something else? I mean, you know, starting in VFX is kind of like I said, you know, it's most of the advice is stuff you should get with any job stuff that for whatever reason, we never got taught in at college at university, which is, you know, think about the people you're working with. Think about the people you're working for. Think about what they want. Show up, do the work. Don't be a prima donna, you know, all that kind of stuff. I mean, I worked with some incredibly talented visual effects artists or artists in general, right, who have no desire to work with them again because it was just like, this is too hard. You're too much of a pain in the ass. So don't be a pain in the ass. That's my number one piece of advice for anybody starting out. Don't be a pain in the ass. Be good, but don't be a pain in the ass.
 And And yeah, either for somebody starting out or somebody that's wanting to move on, I just, you know, it's too easy to say it, but figure out where some of this artificial intelligence stuff is going. I mean, it's incredibly fun is a thing. There's so much cool stuff happening and it's changing.
 Every Every day I'll see something new. I'm like, that kind of blows my mind. So yeah, I'd be, you know, in my spare time, if I was starting off with this, I'd just be playing with all that kind of stuff and seeing what's because the bottom line is, you know, see what you can make, make stuff, right? Make cool things. We didn't anybody that's gotten into the visual effects industry or the film industry in general.
 They They want to see cool stuff on the screen. They want to, they want to tell cool stories. And I think these days just give yourself time to do that. Give yourself time to just explore cool stuff. Brilliant. That's awesome advice. And we've already got like a potential plan here, right? Like your whole micropan system. That's almost like the kind of step from Napster to Spotify right there. That's true. Yep, it's true. I'm not, for me, that sounds like a little bit of too much accounting, accounting work versus making pretty pictures. But somebody should do it. Somebody go out there and do that for me. Okay. Somebody go out there and do that. Brilliant. So well, and lastly, do you think you'll ever do a third edition?
 Oh, Oh, man, I've been asked that so many times. The short answer is probably no. But if somebody really wants to do a third edition, I will happily hand over the reins. I mean, I'm serious here. Somebody should probably do a third edition.
 If If they really think they are up for doing it, they could get in touch with me and we could figure out something where, you know, all the existing content, I'm sure my publisher would be like, sure, let's have somebody else do a third edition and, you know, build on this existing content. I have no problem with that whatsoever. You know, you do all the work. I'll keep my name slightly somewhere on the cover of the book. And, you know, Bob's your uncle. I like it. Very cool. Thank you so much. Have a lovely rest of your day. All right. It was great chat. Thanks.