NAB 2016 Rewind - Jeremy Cox: 3D for Game of Thrones and National Geographic

- [Jeremy] I'm Jeremy Cox and I work at Imaginary Forces in our New York office. We are a bicoastal conceptual design studio, production company, all sorts of things. And I could talk for a while about what we do. I'm just going to show you a reel of some of our work, so you get an idea of what we do. You might already know of us, in which case, don't leave. We'll get into some fun stuff soon. Here we go. ♪ [music] ♪ So I'm going to show you a couple projects that we did recently that we actually used Cinema 4D for. And we use a very broad range of different pieces of software and tools for the work that we create, but Cinema 4D is one of the keys ones, especially that I use, But a lot of us at Imaginary Forces use it, and it's very prominent in a lot of the work you just saw. So, as you just saw, the first project I'm going to talk about is the cover we created for the February issue of National Geographic. And this was a huge honor to get to do this, and, to be honest, it was something I didn't even know that I could do. It was like it just sort of came up that they wanted to work with us to do this cover, and it was just a huge honor to be asked. And I'm sure that there's some photographers out there that are kind of mad at me for getting this opportunity when they've been working their entire lives to maybe get to do something like this, but I'm sorry. So here's the final product. The brief that we got from National Geographic was they were doing an article about the expansion of the London subway system. And they were discovering all of these amazing artifacts literally under the streets of London. London has such an amazing history that as they were tunneling, and as they were building this extension, they were just finding all of these artifacts everywhere. And so they were doing an article on all these artifacts. And they came up with this idea that they presented to us, which was, "Let's kind of create this stylized representation of what's under the streets of London. Let's go through the history from hunting mammoths in the Bronze Age and the Roman period and Battle of Hastings and the great London fire, like all eras in London's history. Let's find a way to represent these in a more sort of exciting dynamic way." So, I'm going to break down a few of the things we did here. With any project like this, the first thing that I actually really love doing is researching it and really going in. And, in this case, this is kind of an aesthetic research. This is like, "Okay, we have this idea it's going to be this kind of cross-section of the streets of London. How do we represent that?" Part of it is sculptural, so you can see on the top row, these are sort of like low-relief, almost freezes. Here's how you can represent figures in that method. And then on the right side, these much more dynamic sculptures. I really love sort of the intertwining of them and the idea that you get these different eras almost like flowing into each other. Then even things like at the bottom here. We actually researched the actual geologic construction of what is underneath London, so the London clay, aptly named, and then chalk underneath that, and really trying to make it as accurate and sort of realistic as possible. Then I actually grew up with all of these cross-section books, and I loved them growing up, like seeing all these like cross-sections of planes and cities and all these things. So I actually was really fascinated just to back through all those books and find these examples of sort of these really interesting cross-sections showing the infrastructure underneath the city, and I think that was something I was really keen to try to get into this cover. The next step is obviously let's figure out composition, and to that really just started sketching. So we used just loose sketches trying to figure out what eras we want to cover, how we lay them out, how they intersect, what are the scales, how many figures can we fit in? Just what's the overall, as I said, just composition. So here are some examples. Here's some more. You can see this one we were actually going for that kind of Trajan's Column look where it was kind of this stratified layers. It was a very sort of clean delineation between the eras, and then on the right side sort of a mix of some of these things, But ultimately, we really liked bits and pieces of all of them. So, in our final illustration, this is sort of the client approved version we sent that we then based our sculpt on. This is what we arrived on. So there are bits and pieces from all of them, and really it was finding the best part of each that we really wanted to create. Once we had an approved design, what I did is I actually just took it into Cinema 4D, and I actually just sculpted just an extremely crude version of it, and this was to get the sense of volume. It was to get the sense of like how much space do we need each of these figures to occupy? It may not look that challenging, but when you actually think about it, where we have like a mammoth down there, and then behind that we then have these Roman soldiers, and then behind that we then have this knight on horseback. But how do you actually get those sort of to sit on a flat plane and still occupy a space? Like naturally you'd sit these one behind the other, and they'd go way back in space. But how do you sort of like tilt them and get them all fitting in together in this way? So that was really the important part of this step. And I'd actually never really done sculpting in Cinema 4D before, so this was kind of just starting from scratch for me. But really it's just you can do it on anything. Let me switch to the sculpting layout here. So, really, essentially what I was doing was... Let me take my materials. So I'm going to make a material out of the illustration that we created. You can actually just drag texture into the material manager, and it'll actually make a material for you, if you didn't know that little trick. Throw it on. That's scaled about right. So, you can see, I'm going to turn on the line so that you can see the density of the mesh that we're working with, and hit C to make it editable. And then you can see in my sculpting layout, I just have subdivide active here, and really you can just hit it. All that does is in your object manager it adds a sculpting tag to it, and this just allows you to non-destructively sculpt on this object. Then this isn't very useful to sculpt on such a primitive plane. So I'm just going to hit subdivide a couple more times, and you'll see it'll actually just add more and more geometry. So obviously I don't want to see that grid, so I can just go back to my normal shading, and really you can just use these brushes to just start pushing and pulling the surface of it. I mean, it's really nothing more complicated than that. It was just me going in and sort of saying, "Okay, I want him to be higher. I want that to be lower. I want him to come up." And it was just going in like that, which is, again, just super nondestructive and very easy to do. So, again, like this was really my first foray into sculpting. There are all these great brushes in here. If you go into the attributes on it, you can actually just load all these presets. Apparently not here, but there are lots of good preset brushes that come with it that have all these like amazing textures and if you want to add bolts or rivets to things, they're all built in there. So once I had this figured out, we actually sent this all back to an amazing sculptor named Meats Meier. He came back to us with this, and this was sort of the like the bones of what we were trying to do, so it kind of had everything where we wanted it. It had all our figures in there. And really, this was kind of the jumping off point where we got this back from him. I don't know if you've followed the publishing industry recently, but let's just say the budget for this project wasn't what we would usually want for sort of a full CG pipeline project. So there was a lot of figuring out how to do things more efficiently than we're used to. So in this case there was actually a lot of me going in and doing it myself. So we got this back from him. As I said, I hadn't really done much sculpting, but it kind of came down to me to figure out how to actually like bring the final 20% of it and really bring it to life and add in the details that we needed and kind of finish it up. This is almost a time lapse. It's like I was doing maybe a couple renders a day as I was just sort of noodling with it and sculpting. And it's a little bit chaotic. But if you actually just look at individual areas and see how different things evolve, I think it's really interesting to see how, say the horse and knight over there evolve through the process, or how the blacksmith in the Iron Age over there. Sort of the posing changes we made it as it was decided we needed to be more dynamic. We needed to change some of the details. So I think this is a really interesting look at just kind of the process that something like this takes, where it's just an evolution. You're just changing one thing at a time until finally it looks the way you want. So this was the final sculpt. So, this is actually all geometry. There are no displacements, no bump maps, nothing in here. So this is, I think I counted earlier, and it was 35 or 40 million polygons, so it was an extremely dense mesh. Actually, I have it open here, and I'm just going to show you some of the little tips and tricks that I figured out. Essentially this is going to be how we evolved it from that kind of more rough stage to the final result that we ended up working with. So you can see, even if you look at the open jail there, that's like that a very high number that I kind of can't even tell. I think that's 48 million? It's a lot of polygons. I'll just say that. You can see Cinema 4D is actually being very responsive. I can just tumble around this. That's ridiculous. And then people would come up as I'm working on this and like want to see the wire frame or something, and then it's like a joke to go like that. There's my wire frame. But what was extremely important about this was organization, because ultimately we couldn't work on it at this level. As much as it is fast to work with, we really needed like even faster than that to be able to sculpt things. So really having an efficient layer system was extremely important. So with this, I can just solo individual elements in here and just work on those one at a time, and you can see that even gets much, much faster. If I had the frames per second down here, you'd see it's jumping to like 150 frames a second or something, from 20 frames a second. Just to see how I'm breaking this up I can turn on options, layer color. So you can see we're just taking every single sort of major section and just segmenting it up. And I think that this is sort of a very overlooked portion of Cinema 4D is your layer manager and just being able to quickly organize and have control over your scene when it gets this complex. So one of the things we actually needed to do a lot, we would need to repose these characters. Because they were actually originally done in Maya and ZBrush by our sculptor, but I was, again, bringing them into Cinema 4D, and I needed to, like I say, make a pose more dynamic. So how do you do that? Again, that was kind of a learning process for me. So let's take this guy as an example, this blacksmith. Snd I'm just going to copy into a new scene, just to make it a little more efficient. You can see one thing that's actually going on here is there's a deformer on here that's actually doing some pretty crazy stuff to him. And this is one thing we did, as I said, because we were trying to squeeze these characters into kind of unnatural poses, where they looked good from the camera's point of view, but they don't need to look good from any other point of view. So we were doing all these like squashing and sort of warping. And this was using the freeform deformer, or I guess that's a little bit redundant, but freeform deformer, FFD. This, you can just take these points, and you just distort a character just by kind of moving him around. So if we wanted to make him a little more natural in depth you just pull him around like that. This allowed us to dynamically, or interactively, fit these characters into these somewhat unnatural positions. What I wanted to talk about actually was these sort of different ways that we were able to adjust the pose of the character. So, for example, I want to adjust his hand a little bit. One thing I can actually do--I'm going to just go back to--I'll stay in the sculpting, because I can actually do it that way. Although I talked about like hitting subdivide and sort of doing it that way using a sculpting tag, you actually don't need to do that. All of these sculpting tools, they work just fine on normal geometry. So let's say I want to adjust his hand a little bit. I can just sort of like push and pull this, and this isn't a parametric thing where... It's actually just changing the geometry. So it's a very interactive way just to adjust your model slightly, or do normal sculpting. Maybe I just wanted to go in and just kind of add some more details in here. I don't need to go through a process of then baking it out or making it. It's like you're just working on the geometry. And if your mesh is dense enough, I found that to be a much more efficient way to work. So another way, let's say we want to turn his head, like this is kind of just a dumb way of doing it, but just select some polygons. Then one thing that I actually didn't know... one thing I need to do is turn off select only visible elements, because I want to select all the way through the model. Let's get there and there. So I didn't know that you could actually hold down L, and then when you drag your axis, you can actually reposition it dynamically. So it's essentially, you're changing your axis this way, but rather it's just a hot key where you hold down L and move it. So that's an extremely quick way just to move your offset point. I'm Actually missing some polygons there. So we'd actually do this quite a bit where it's just like cheating, because ultimately, as I said, we're not animating these characters. They're just ending up on a still frame. So if we want to say rotate his head, we can just put his axis there, and then rotate his head around. And, obviously, you're going to get some weird distortion and things, but because it's a sculpt, we just go in and smooth it out and make it all beautiful. But really, it's just about getting it in the right position. But then one thing that I did a couple times, which I was actually surprised by, because I've always been kind of afraid of like rigging in character animation, and it's actually not nearly as scary as I expected. So, really it's just all about joints. So if I just make a joint, I'm just going to place it kind of like on his shoulder blade. Again, I would, maybe I'm going to move this arm, like this arm I want to sort of move around a little bit. So I'm going to do one in like the middle of the shoulder blades. The way this joint works is you just duplicate it. So I'm just going to duplicate it in the hierarchy, and now this is going to be the kind of shoulder itself. So let's move that over to there, and then, again, another joint underneath. Move this to his elbow. Of course, you'd want to spend more time on this and actually make it work well, but right now I need to do it quickly, so I'm just placing it roughly. Seeing your four up view is very important to actually get a sense of where they are in 3D space. There's one for the hand. There we go. So this is really all you need. You just need a joint chain which is this, and then you need a geometry, and if you select all of them and go Character, and go to Commands, and then go to Bind, this is going to do all the work for you, and it just binds. It's going to create skin, and it's going to create this weight tag, and if I double-click the weight tag and click on my joints, you can see I get this heat map, and this heat map is which joints are affecting which pieces of geometry. See, I'm getting a lot of overlap here so what you can do is use the weight tool, which is what's highlighted now. And if I set that to erase, I can just kind of erase it around, and I'm not going to go and do that for all of these. It's going to take time. But you can see, it's going to have a lot of weird deformation in the body. But with the arm you can see. I can go through joint by joint and repose him to be in the more dynamic, or whatever the note is that I want to address. I can just very quickly adjust it. And, again, this is something I was kind of afraid of, and it was like, "I really don't want to have to do character rigging. It's like a day's worth of work just to do that." And it's like we just did that it a couple minutes. So, yeah, so I would say that most of this, it's like the sculpting. It's like the sculpting tools, and, again, like I hadn't really used them. I started using them. I'm really impressed, like the project right after we finished this, and then like jumping in, sculpting, just doing it. It's really easy. It's right there. It's right in Cinema. You're not exporting to ZBrush or anything, and it works really well. So, as I said, I mean, it's just a still frame that we're creating. It's almost a matte painting, so ultimately we are rendering out all sorts of different pieces to assemble this thing. And we actually had a Maya artist rendering this out of V-Ray out of Maya. We had myself rendering things out of Cinema 4D, and ultimately, I'm just going to go through these quickly. We were just creating all of these different renders that in Photoshop we were then just comping together. It's not like we needed to do a proper CG comp, because all we were doing was just, it's a matte painting. So, like we were adding displacement to things, different amounts of displacements that we could just like in Photoshop dial in. Okay, we want the face to have very little displacement. Then we want the background to be really displaced and have sort of feel like the sculptor had spent more time on the face and less time on the background. Even moving the lights around, so if we did a little ping on this side on this character, we could just paint in little bits here and there. We did some like marble passes, a lot in the faces and things. We did some more sort of like high end feeling sculpting. This was actually just a subsurface pass out of Cinema 4D that I actually think is super beautiful. That's a close up of it right there. And I'd love to do something just in this style. I think it's really nice looking. So, here's the final product again. You get a bunch of detail shots, just so you can see the amount of, just the resolution we were working at and the amount of detail we were trying to put into this. Again, it's really just a lot of Photoshop in the end. It's like it's really just a still frame. You can do whatever you want to it in Photoshop. Because we could, we had the geometry, and then we finished the final cover, so we just projected it back on it and added a camera move to it. This was just for fun. It was just to create something that kind of showed it off and really like showed the volume and dimension of this thing. We actually even created... I don't have it here, but we created an anaglyphic stereoscopic version that was kind of funny. If you had the old red and blue glasses you could look at it, and all the characters were popping out at you, and that was a really fun dynamic way to see it. So, yeah, as I said at the beginning, this was an amazing honor to work on this. And the folks at National Geographic were amazing collaborators to work with. So I'm really proud that I was able to work on this. And I actually have a physical magazine with me. So, after, if you want to come over and see it and actually look at it in person, you can come see it. All right, so next up. I'm going to talk about some teasers we created for the latest season of Game of Thrones. It's a little show you might have heard of. One or two of you might have seen a couple episodes, I feel like. But I hear the new season's airing this weekend. So, mark your calendar. As I said, we created three different teasers for this new season, and I'll play the teasers, and then I'll go through some of the process we went through in creating them. ♪ [music] ♪ - [High Sparrow] Everyone of us is poor and powerless, and yet together, we can overthrow an empire. ♪ [music] ♪ - [Ramsay Bolton] Winterfell is mine. Come and see. ♪ [music] ♪ ♪ [music] ♪ ♪ [music] ♪ - So, HBO approached us with this idea, and they really liked this concept of using these banners, and if you've seen the show, you know that those are banners of three of the Houses in Game of Thrones, and it's the Lannister, the Targaryen, and the Starks. They were looking for this sense of kind of almost like everyone is in it together; it's this kind of dire straits where it's like they're all tattered flags, and it's all like this sort of apocalyptic landscapes, even like the sunset feels kind of foreboding. But everything having this sort of feeling of dread to it. So, when we got this brief we jumped in, and we just started creating style frames, and so these are our initial style frames for the Stark snowstorm. They talked about sort of this kind of lightning and snow, just really this crazy storm that was kind of engulfing the banner. Then Lannister, we put it like it's on the parapet of the Red Keep in King's Landing looking out over Blackwater Bay, and this kind of, again, almost like an apocalyptic sunset behind it. As you can see, these are actually extremely similar to what the final result was, which I think speaks to really we got just an amazing reaction from the guys at HBO marketing who were looking at this stuff, and they just loved everything we were doing, which was extremely gratifying and great to work with them. So then here's the Targaryen, and, again, it's like this kind of fiery landscape that this banner exists within. I mean, one of the interesting things in creating these was they're very pure teasers, so we didn't actually know anything to tease. It's not like they told us any plot points that we could spoil, but really they didn't want to imply anything, in a weird way. They kind of wanted this... they had this tone they wanted to give the audience, but they didn't want to sort of imply any plot points. And if you've seen the last season of it, there are certainly a lot of plot points that are up in the air that they didn't want to spoil, and they want to wait until this Sunday to probably reveal some things. But we'd have these great phone calls where we'd sort of like be pitching ideas and throwing things around with them. Then suddenly they would put us on mute, and they'd be talking and talking and talking, and come back on the phone. "No. No, we shouldn't do that." We still have no idea what they meant by most of those things, but it was really funny to kind of like be a part of that secrecy and understand a little bit of what they're going through in trying to make this show. So, I'm going to go through each of these spots and break down some of the techniques we used to make them, and just a little bit of sort of a high level thinking on this. Generally, we were thinking of these as kind of like glorified matte paintings with CG banners in the foreground. So the banners were created in Maya and rendered in V-Ray. If you've heard me talk here before, you've probably seen that as a bit of a thread that we use all sorts of different software, and it's kind of like whatever, whoever we have working on it, like use whatever you feel comfortable with. So in this case, we had a Maya guy doing the banners, and it was myself and another compositor doing the backgrounds. So, as I said, it was really, especially the backgrounds, it's a glorified matte painting. It's starting out as simply as possible, and then just adding in the elements you need to really bring it to life. So, in this case, looking out over the Bay it was really about, okay, we can do a matte painting of a sunset; that's easy. We'll make the clouds move a little bit. Okay, but we need a bunch of water. How do we do that water? We ended up doing that in Cinema 4D as a reflection. So it's sort of a bunch of noise shaders, and I'll show you that in a moment. Then the island is actually a fully CG island that I created in Cinema 4D using some of the trees in the content browser, and then modeling out some of these other elements, then adding in some stock footage of birds. And really it's just it's finding these little details that really bring it to life and make it feel real. So I'm going to jump into Cinema 4D here and show you the scene for that Lannister. A couple things I want to talk about here. So, as you can see, it's kind of what you saw there. It's the shot. But since I sort of come from a little bit more of a compositing background, and kind of a design and compositing, so kind of a 2D background in a weird way. So I always think about these things from the view of someone who's trying to kind of assemble the scene as a composite. So one of the things that actually, or I would say this is one of the biggest features in like the last five years in Cinema 4D for me, which most people probably don't care that much about, is the take system. If you've ever used Maya or something, it's like the render layers are a huge way of rendering things out, and finally Cinema 4D has an equivalent. And I actually think it's an amazing implementation of the idea of render layers and actually goes way beyond that and allows you to do different takes of animation, and all sorts of different things. But I'm going to undock this really quickly, so you can see it. It's not very useful if you're actually covering. I'm going to change this back to the standard layout, and undock that. So, you kind of need to see your object view at the same time as you see your takes, so I like to keep it... I usually have a second monitor that I have it on, but you can see it here. The idea of takes is each of these, it's almost like a preset of your scene. It's like it's a recording of the various settings you have at any one time. So if I go here to sun, if you actually pay attention over here you can see a bunch of things being turned off and on, and some of it you can't see because I have compositing tags where I'm turning off scene by camera, because if I still want the reflection of it, but I don't want to see it in the camera, I'm doing things like that. As I toggle through you can see different things are being turned off and on. And really, you can see in here, these are just all the things that are being recorded in each take. So, the entire scene stays the same, except for these little things that want to change in these different takes. So in this case, I wanted a pass where it was just the reflection of the sun, because I wanted to be able to composite that separately from the reflection of the sky. So getting that control where I could sort of like change the shape of the reflection just by doing some roto in it, having that flexibility. In the past, we would have needed to save out different project files for each of these. Like if we're putting it in a render farm we need to save out one for the reflection, one for the island. So you'd sort of like split it, and then if suddenly you want to change you either need to remake every one of those files, or you need to go back in each of the scenes and make that change one at a time, but now you just make it in the main take. You just go up here. You make the change, and it propagates through all the rest of your take. Again, like sun, so the inverse of that one I just showed. No sun. So this is a reflection of just the sky in the water, and then, again, I can just add the reflection of the sun onto the reflection of the sky, and you just get so much more control that way. Similarly, as you might guess, if I go to the end of the shot you can see there's the wall in the foreground coming in, and, again, it's just splitting into as many layers as a compositor I would want it split into, and the great thing about this is once you deal with depth of field and things like that, you get so much more. Like you just get nice edges. If you ever dealt with depth of field normally, where you have a lot of overlapping objects, you're always going to get funny little aliasing issues and things and artifacting and the blurring. So anything you can do to kind of separate each layer out is going to be extremely beneficial in the comp. That's the wall. There's some towers of the island, in the background there. So really it's like just logically breaking up your scene, and this is very easy to set up. So, let's say I want my towers, but I only want this tower in the scene. So I'm going to make a new take, which is this button. And right now this button right here is active. This is, I think it's called Auto Take. Yeah, Auto Take. So essentially this is like auto keyframing, except it's actually automatically recording the states of objects that you change. So, it can get very dangerous, because you can suddenly start actually making changes that are important to your scene, suddenly they will only affect that one take. So you need to be careful about when you have it off and when you have it on. But if it's on, really all I need to do is say, okay, I'm going to turn off the sky. I'm going to turn off the water. I'm going to turn off the right tower, because I only want the left one. And I'm going to turn off the island. I'm going to turn off the island trees. Let's do all those things. So I'm just turning things off until I have what I want. Flag, don't want that. Okay. So I'm going to call this Left Tower. Now if I go back to any of these others you can see they're still exactly as I left them. Go back to the main. Each of these, it's an entirely nondestructive way of making different versions of your scene, is a way of thinking of it. You can even have animation that is unique per take. So if you start keyframing animation with this auto take on, that animation will only apply to that one take. So even if you're auditioning different camera moves, you could do different animation on your camera. Talking about cameras, you can even assign this button right here. You can tell it which camera you want to render that take. So when you submit it to render you can actually have one camera from over here, one camera from over here, one camera from over here. So there are all these different ways you can use this system to make it a much more kind of modular rendering process. This button right here, this is your render settings. So you can actually have different render settings per object. So even if you want to, say, render one object in V-Ray and one object in physical render, and then maybe you want to render just a depth pass in standard renderer, you can actually separate your render settings out and select different render settings for each take. It's just a huge amount of flexibility. One last button over here. This, I kind of made fun of MAXON earlier about this, and they're probably a little mad at me. But this button is like the most unintelligible icon ever, and I needed them to tell me what it actually did when you clicked on it, because that icon is so meaningless to me. What this button actually does is when it's toggled, which is the default, as you can see, any of the objects, you can see how everything's grayed out. That means, it means it doesn't want you to change it. You need to manually override it to make a change. But it also means that because I'm in this take I can't change... Let's say I want to take the position of this tower, I can't change that. If I do, if I override it, it's only affecting the position in this take. But if I want to affect the position of that tower in every take, that's actually what this button does. So if I uncheck that, you see how it un-grays that out? Now I can change it, and say I want to move this tower up to there. If I go back to the main take, you'll see now it's up there. So it doesn't record it just on that take. It records it everywhere, even when you're inside a take. The inverse of that, or just to demonstrate, I guess. If I go back to the left tower, and now I actually do have Auto Take selected, and move it down, now if I go back to the main take, it'll still be in that higher position, because now that's the default, and when I recorded this new position now it's down there. So, as I was saying earlier, I just think this is an amazing new tool for those of doing sort of like more complex rendering and breaking things into layers, because it's been such a pain in the past. And really, like I've been asking for this for years, and I've probably hugged a bunch of them a bunch of times over this, because it's just an amazing new feature that I love. So the other thing I want to show you in this scene is how I did that reflection in the water. It's one of those things that is actually extremely simple, but it just looks complicated, and it's really it's just using Cinema 4D noise in a bump, in with a reflection, and that's all it is. It's just animating that. So, I'm going to copy... let's see. Let's copy the sunset, and it's projected, so I need the projection camera. Let's do that in a new scene, just to make it simple. Yes. And it's going to look a little weird with the tiling, but let's just roll with it. So this plane, this is going to be our water, so I'm just going to make it nice and big. Kind of start going out to the horizon. That'll work. So, really all I'm going to do is make a material, go into reflectance, add new layer, and just do a GGX. So this is just essentially a mirror, and that's the simplest thing you could possibly do. I'm just making a mirror, putting a mirror on the ground, and if I render it, I'm just rendering a reflection. So that's like a still pond or something. It's like it's not very interesting, but it almost still feels like you're looking at a body of water. So, really what I need is, I need ripples. I need waves. I really need to give it a dynamic quality. So, as I say, I've done this a few times, and I'm always amazed at how well it works, but really it's just you're just add in a bump, put a noise in there, and this is actually still going to look pretty terrible, just because it's... Sorry. Put it on the wrong-- I don't want my sky to be bumped. I want a reflection. Put a noise. I'm going to render it. Generally the default noise always looks bad, so I would always recommend that you not use this noise unless you have a very specific look that requires that noise. But if you actually go in here to the noise you get this little button here, and these are all your different noise types you can choose from. It took me a long time to discover this button, like I used to just like pick from here, and you have no idea what you're picking. But actually going in here and, for some reason I always like Stupl. I don't know if that's how you pronounce it, but this one I always go for, because I think it generally gives a really nice look. So, a rendering is great. I'm going to change it to physical render. It's a little better. So you can see that already starts to look like what you'd expect a body of water reflecting a sky to look like. Really it was just about layering these things together. So generally any time I do noises I almost never just do one noise. It's always multiple layers of noise layered on top of each other. So if I just take the noise, and I put it in a layer, I can take another noise. Like I told you never to use that noise, but I'm actually going to take that noise, the one that no one likes that's the default one, and let's go into it, and let's just make it nice and big, so let's make it like 500. So now we're getting just a really big kind of a soft undulating noise. And maybe with this, I'm trying to get something like if you ever look out over the ocean you see like different areas of it feel like the different textures, because of the wind blowing at the waves differently or different... I don't know if it's like surface qualities, something. Something about the water looks different in different areas. So getting that kind of quality out of this. So, going back into the layer. Right now, all I'm seeing is the really big noise I just made, so I need to lower the opacity of that, and I probably want it really low actually. Too much, and you'll just get these kind of mountains and hills and things. Again, I'm not sure how much of this you'll actually be able to, how much of it will come through on the screen up there, but you're really just adding variety and things. Then another thing that I wanted was, let's add some like longer horizontal waves to really feel like they're like crashing in towards us. So let's, again, maybe do noise. Let's try the same one we did before. Let's try Stupl again. And then I'm going to scale it separately, because, again, I don't want it to be just the same noise. I want it to be kind of stretched out so it's like these waves coming in towards shore. So this is X, Y, and Z, and X is the direction I'm looking. So I want it X. I'm going to make it maybe 500 wide instead of 100, so it'll be kind of elongated in that direction. You can see in the preview here how we're starting to see it. Then I'm just going to, again, lower the opacity to that, because I just want to mix these together. Let's just try it there and see how that looks. Let's turn it up a bit. My scale might be a little small, but you're starting to see it. You see these sort of like horizontal striations through it. If I scaled it maybe 200%, yeah. So it's just this sort of iterative process. You change something. You render it. You change something. It's just a creative process. So the one thing then that I did was, obviously, right now it's just static. You want water to move. So all the textures have this animation speed, and I can't really render it to preview it, but I'll just say, you can change this animation speed. It's usually way lower than you think it should be. It's like .1 or .25, or something. Don't just go to like 100, because it's going to just like go all over the place. So we used that, and even like for the waves crashing on the shore, we used movement. So, again like I said, this is X, Y, and Z. So in Z, moving towards us, you just put a value in there, and then the speed. It'll actually feel like those waves are kind of coming towards you. And that mix of like the waves that are both just undulating and the waves that are coming towards you, you get this really nice feeling that actually feels like pretty accurate water. So let's go back here. So I'll just actually play that once more, just so you can see. Like if you look at the water in particular you see those sort of larger undulations of waves, then all the little rippling, and all these different pieces coming together to really sell it as something that feels like real water. So this was Stark. This was the Stark background. And this is actually relatively simple, although it looks like it's kind of a crazy snowstorm. The snow itself was done in After Effects in Particular. Most of our comping was done in Nuke, but I still use Particular all the time, just because it's a super fast way of generating a lot of particles. So for things like this with snow, where you want depth of field and all that, I usually just jump straight into Particular. The background is all a projection. So we just did a projection, sorry, a matte painting in Photoshop, just like a really rough sort of mountainous glacial kind of thing, and just projected it on some really basic geometry. Then we actually used X-Particles to generate some of the... you can see, there's some little like bits of snow and ice and things bouncing around on the ground. We used X-Particles for that. And they're in the back, in a booth. You should go check out what they're doing because it's pretty amazing. So I'll jump in here again really quickly and show you some bits and pieces of this. The first thing you'll probably notice is it's still have the same thing, the takes. I'm using it everywhere now. Sky, background mountains, foreground hill. Then fog. This is kind of an interesting one. This is a technique I've used a lot, and it looks pretty amazing, but it's a way of-- let's see how quick this renders. It's a way of generating volumetric fog in Cinema 4D that's, as you can see, it's very fast to render. You're not simulating anything, so it's real time. And what I'm doing here is I'm actually using a light. On the light, you can make it visible, and then put noise through it. So essentially, you get a volumetric noise that's affecting the visibility of a light. So I can show you that really quickly. If I make a new light, I can set its visibility type to volumetric. Sorry, visible is actually the one we want. And then if I just render that, it's just like a gradient. It's kind of what you'd expect. It's the falloff of the light, but that's not going to work for fog really. I mean, I guess it could work for like kind of a depth pass, something, but it's not nearly as interesting. So, if I go into noise, this is where I actually break up that gradient with what's actually true volumetric noise. So if I change it to visibility, and then there are different types here you can choose. Let's just try the default one at first, so you can see how you're actually just getting what looks like kind of a clouds thing. What I usually do is turn up the contrast quite a bit, and you can see it's blowing out, so you just need to lower the brightness a bit. Let's try rendering that. Starting to be good. It's still a little bit kind of like fractally noisy looking. So I usually play with the different types, so like hard turbulence and wavy turbulence and things. These get different aesthetics out of it. But the great thing is once you, I mean, this looks like one thing, but once you actually put it in like a ground plane for it to work with, so let's just do landscape. Then scale that up. Just going to take off border at sea level. So we render this. One thing I think I need to do here is this illumination scale. I think that's the setting. Sorry. That's not it. There's a setting in here somewhere. I'm not remembering, visibility sample distance. This is kind of how accurate it renders that fog. So the smaller the number, the more accurate it is. Let's try something like 3 or something much smaller, and it will take longer to render, but you'll get much better sort of interactions with the ground. The other thing I'm going to do here is just... right now we're seeing the actual light illuminating the ground. So I'm going to turn off, make a texture with just pure black. Put it on the landscape. Render that, and there. You see you get this really nice like volumetric light, and if I actually put my camera right in there, you're actually going to get what's a really nice kind of depth pass. And what I'm doing with this is then animating the fractal noise. I'm just animating it so it kind of evolves. Then I'm also putting what's essentially wind on it, so it just moves. So I actually have an example of that here. So this is like a rendered pass of what that looks like. And, yeah, like I was showing people this the other day, and they're like, "You didn't simulate that? That looks like you did some actual fluid simulation or something for fog." It's like, "No. It's just noise, just dimensional noise." Obviously it's rendering out with the environment just as pure black, so you get that as kind of cut out, and you just composite it right on top, and it looks really nice. I was talking about using X-Particles for some of the snow on the ground. So this is just generating a bunch of particles kind of on that ground plane that I'd created, and then really just wind. It's just blowing them across, and because you can actually use proper physics where it kind of like goes over the surface, it just has this really nice like tactile quality where it really feels like the snow's interacting with the ground itself. So the last teaser we did was the Targaryen one. And this actually has the least actual Cinema 4D in it. But really it was mostly a compositing challenge. It was like how do you make this much sort of like smoke and fire and things look good? Usually our answer was, let's find some real fire and smoke and composite it in there. Because generally, simulating stuff is asking for trouble. But we actually did do some simulation in Turbulence FD for this. So that's like a fluid simulation software, like a gaseous fluid simulation for Cinema 4D that you can actually do some pretty amazing things in. And so we used that for some elements. It's a lot of footage, as I said. Then, as usual, some Particular, so the sparks moving through the scene are all done in After Effects in Particular. So here's an example. This is me just playing with the simulation. And again, I could jump in there and try to show you some Turbulence FD, but it's not really an efficient thing to try to show real time, because it's simulation times and things. So the one thing I actually want to show you is, if you're curious about it, the thing to do is go into your applications. Go into your actual install. Find the plugins folder, and Turbulence FD examples. I don't think anyone really knows this is here. But these are a bunch of examples of different kinds of simulations. And I would say just pop these open, and just like hit simulate and see what they do, and change some settings and see what they do. Really, it's just that's how you're going to learn it is playing with other people's files. If you try to follow the tutorials where you make a cube, emit something from it, it's like you're never going to get anywhere. But playing with something that actually looks good from the beginning; that's how you get good at it. Not even good, I'm not good at it. I just mess with things until I like the look. So another tip I have about doing simulations like this is render, like as you're going, just render the same frame over and over. So as you're iterating with different values, so if you're just trying things and kind of trying to figure out what each one does, like there are just hundreds and hundreds of parameters in Turbulence FD, and it