 Hi everyone, this is Brian Bailey. This is my logic fest Quick tip. This is something I presented at a Dallas user group a couple years ago And I tried to recreate that demo For you now, so I'm gonna go through it pretty fast because I don't want this to drag on The basic idea is what I typically get from clients is they've already shot and now I need to figure out how to fix what they did so This is something that you could execute a lot better if you had more input earlier But that's not what I get to work with so They shot a room and it's just locked off and it's a still image and They want that room to be in this phone That she's looking at and you know as she turns you're gonna see different parts of the room so The basic idea and I'm gonna go from left to right here just showing you some steps What I did was I got a perspective grid and I don't have as much luck doing this with gmask When they're tracked, but the perspective grid does this really well The perspective grid is tracked on and If you look at like a top view or a side view you can see that the perspective grid is actually moving in 3d space and It should be pretty accurate The gmask planar tracking isn't always accurate for what we're gonna do. That's why I use the perspective grid So it's tracked on there What I'm gonna do is actually add a camera below the perspective grid and Then if we look at you know like the side of that You see that the camera is moving with the perspective grid Now if we take that camera and zero this out so that it's right on the I'm gonna want to do it to this camera Move it to where it's right on the perspective grid. So what does that look like? Well, that's actually like the camera on the iPhone that is on The back of the screen Or the back of the phone You could actually go so far as to move that And you'd have to get it lined up just right to where the camera is where the Camera on the phone is If it's kind of like the top left or the top right I Just left it right in the middle because I thought that was accurate enough for what we were doing So anyway, if we look through that camera Change this to camera one and I can't hide that now for some reason don't know why look through camera one We're actually getting this movement That mimics what the camera on the actual phone is doing So next add a sphere So I've imported just a basic geometry sphere. I've added another output That is this camera So if we look through that second output and we're have that sphere lined up I made it really big Hopefully you can see what's going on here So that looking through that camera looking at the sphere we can get an idea of The environment that camera is looking at if we texture that with our still image and Turn off wireframe We start to get something that Kind of looks like the camera moving around in the scene So the next step I did a few little things Just a little lens to start to make the edges look even more fish eye There's my output of That camera And if we look at the results of that and you can see that I made that action 1080 by 1920 since that phone is vertical And this is a little rough and like I say if you were had input beforehand you could actually get a spherical of the room But because I didn't get that I'm kind of cheating it and What the result is did a quick key To get that hand in there So I've got a layer that's the hand cut out and then the result of that second camera Tracked on to The actual phone moving a few more little tweaks and stuff The result is I Hope passable. I could work on that some more. I could smooth that out Cheated even more to get exactly, you know what I want, but you're seeing What's moving in that camera that we actually used as the camera as if it was on the back of the phone That's my quick tip. Thanks