WEBVTT 00:00.420 --> 00:03.360 Welcome to solving camera emotion. 00:03.510 --> 00:09.690 So this is going to be all about setting up camera movement and indeed working out whether the tracking 00:09.690 --> 00:13.410 we've just done is any good for our scene. 00:13.800 --> 00:15.870 You'll need the right camera details. 00:15.870 --> 00:18.000 I don't know if I've mentioned that before. 00:18.000 --> 00:19.140 I think I have. 00:19.410 --> 00:25.260 It is important whether you've got the camera details by faffing around for hours to try and get them 00:25.680 --> 00:31.110 or indeed you know the camera details because you've got the spec it is important to have the right 00:31.110 --> 00:36.260 camera detail so go ahead make sure that's one of the first things we get sorted. 00:36.420 --> 00:40.460 We're going to check the accuracy of our markers as we go through as well. 00:40.590 --> 00:46.110 Now I was quite thorough but I might have missed one as I was going through because it may have looked 00:46.170 --> 00:49.510 on a sort of zoomed out scale like it wasn't jerking about too much. 00:49.560 --> 00:55.040 But we will go through and check the accuracy of our markers and dump any that are causing us problems. 00:55.080 --> 01:02.250 Now an average camera solve error of less than one is ideal but that's not always necessary. 01:02.250 --> 01:08.840 We may have a camera solve error of six and it work out to be absolutely fine. 01:08.850 --> 01:12.990 We do have to test and assess as we go through. 01:13.080 --> 01:16.380 You may have to refine your markers more. 01:16.440 --> 01:20.550 It may be that one of your markers is causing a big problem see how to get rid of it. 01:20.550 --> 01:23.770 There's lots of options and alley ways we can go down. 01:24.060 --> 01:28.820 But let's go ahead now and set up the camera movement over in blender. 01:29.130 --> 01:34.380 Okay so I've just re opened my file and the first thing I've noticed is my lovely light blue line where 01:34.380 --> 01:36.750 it was cash before is no longer cash. 01:36.750 --> 01:41.910 Now that's not surprising because obviously if you shut down blender is going to lose anything that's 01:41.910 --> 01:47.820 temporarily stored so I'm going to go up to the top on my tall shelf and instead of just playing the 01:47.820 --> 01:53.830 clip as we've done before I can go ahead and pre fret fetch and the short cut there is. 01:53.830 --> 02:00.180 P It will prefetch the frames from the disc for faster playback and tracking and we can see that is 02:00.210 --> 02:04.730 a lot quicker than it was just hitting play. 02:04.920 --> 02:10.590 And now we should be able to scrub through we can see it's consuming around seven were seven point two 02:10.590 --> 02:12.610 gigs this time round. 02:12.660 --> 02:15.970 I can hit reload now reload is a bit pointless. 02:15.990 --> 02:22.750 In certain cases it's only if you've edited the file on the disc itself on your computer. 02:22.920 --> 02:25.000 And now I'm going to have the prefetch again anyway. 02:25.380 --> 02:31.890 So we've got that now in our buffer and ready to go and we can just go ahead and go to the next tab 02:31.890 --> 02:35.510 down on the tall shelf which is the sole tab. 02:35.850 --> 02:39.990 Now before we do anything else let's just click solve. 02:39.990 --> 02:42.960 Camera motion and see what happens. 02:43.580 --> 02:44.270 Wow. 02:44.390 --> 02:45.200 Okay. 02:45.200 --> 02:47.010 I was not expecting that. 02:47.030 --> 02:53.030 After all the trials and tribulations are been through trying to actually work out this scene I've got 02:53.030 --> 02:59.180 to solve error and we can see on the head above the clip editor of three point six which is a great 02:59.180 --> 02:59.720 start. 02:59.720 --> 03:04.350 It turns out I may need do nothing more but will go through and refine it anyway. 03:04.520 --> 03:08.560 Oh and before we do anything else you know what I haven't done. 03:08.560 --> 03:12.590 I said it was the first thing we're going to do I got distracted by this blue bar. 03:12.590 --> 03:21.800 We need to go over to the propertys and go to camera and set it to full frame and scroll down to lens 03:21.830 --> 03:29.800 and set it to 20 because that is what the dji Phantom for's lens configuration is said to be. 03:29.960 --> 03:35.450 Let's go ahead and click solve camera motion and it goes up to six point five. 03:35.630 --> 03:41.840 Okay we can work with that is not great that it's gone up but we can play with the settings now and 03:41.840 --> 03:44.390 work out what we can do. 03:44.420 --> 03:50.690 Now one of the things we can do now we've got a solver error is we can have a look at individual tracking 03:50.690 --> 03:55.240 points and they will all have an average error attributed to them. 03:55.250 --> 04:02.570 So if you've got something that's got a high average error you may find and you say may find because 04:02.570 --> 04:09.950 it does vary removing it from the scene can improve your particular odds of getting a good solve. 04:10.010 --> 04:15.530 So after clicking through all of those I can see there that one seven point six before I do anything 04:15.530 --> 04:18.520 else I'm going to save my work extra. 04:18.610 --> 04:20.810 I'm just going to delete this Bush entirely. 04:20.900 --> 04:25.280 So we got a average solara of three point six now before we do anything else. 04:25.280 --> 04:31.790 I do want to just highlight occasionally when you hit solve camera motion again you get a different 04:31.790 --> 04:32.780 result. 04:32.810 --> 04:39.050 So let's go ahead and remove this one that's rated on an average of an average error of seven point 04:39.050 --> 04:47.510 six on the propertys panel here we can see if that by removing it we get a better results of three point 04:47.510 --> 04:48.280 six solve. 04:48.290 --> 04:52.490 Camera two point nine two that's much better. 04:52.490 --> 04:57.700 Now I'm going to guess here if I go through and remove the one that's five point six. 04:58.400 --> 04:59.250 All right. 04:59.330 --> 05:02.910 Is there anything else that's above 4 that's going to be my metric here. 05:02.910 --> 05:04.530 And of course I need eight of these. 05:04.550 --> 05:06.500 Ah well that's really low. 05:06.500 --> 05:11.100 So let's see anything here that's above for I think there was. 05:12.450 --> 05:13.660 Ah there's one that's five. 05:13.660 --> 05:15.270 So let's get rid of that. 05:15.310 --> 05:20.740 Now again getting rid of it may not have been the answer we may it may be better to refine it but let's 05:20.740 --> 05:23.850 solve now were on two point nine one point eight. 05:23.860 --> 05:26.290 This is getting better and better. 05:26.310 --> 05:35.470 Incidentally you can look at this graph Ed and it gives us an idea of the actual what's going on in 05:35.470 --> 05:41.260 terms of these markers and if they're jerking about too much and if they are jerking about that may 05:41.260 --> 05:42.360 cause us problems there. 05:42.370 --> 05:45.180 I'm really happy that we're down to one point eight. 05:45.280 --> 05:50.140 So I'm not going to do anything else for the moment because I need to check and see what that looks 05:50.140 --> 05:52.820 like being lean about it. 05:52.870 --> 06:00.020 So I'm going to scroll down on the tool shelf to scene set up and click Set as background which scan. 06:00.040 --> 06:04.270 This particular video has the background on the camera view. 06:04.410 --> 06:07.280 I'm going to set up a tracking scene as well. 06:07.690 --> 06:14.530 I'm now going to go over to the default layouts and from this view I'm going to watch the camera and 06:14.530 --> 06:17.740 make sure it's not doing anything crazy. 06:17.740 --> 06:23.860 If it starts doing crazy things and wobbling about that can cause issues out there are a few jerks at 06:23.860 --> 06:26.020 the beginning and a few around. 06:26.020 --> 06:27.590 Let's just watch here. 06:29.390 --> 06:31.050 Around 180 or so. 06:31.250 --> 06:32.910 Let's look at it from the camera. 06:32.910 --> 06:36.590 And now this will slow down tremendously. 06:36.590 --> 06:40.550 Now the important thing to note here is I'm going to have to hold down Shift to move. 06:40.780 --> 06:46.410 But I want to have a close eye at the cube itself or the ground. 06:46.430 --> 06:52.850 What I want to make sure is as this is playing through the timeline that it's not jerking about horribly 06:53.630 --> 06:59.630 and then it's consistent So if the cubes covering up part of our scene that it remains covering up part 06:59.630 --> 07:03.410 of that scene and not sliding and jerking all over the place. 07:03.470 --> 07:08.090 If it does slide and jerk all over the place it just may look horrible. 07:08.120 --> 07:14.930 In the end now because I'm going for an aircraft flying over the surface I can probably get away with 07:14.930 --> 07:16.850 a bit more jerk than I would do normally. 07:17.740 --> 07:20.050 Now I'm going to pause it there and go back to the beginning. 07:20.070 --> 07:21.800 Now is sluggish doing this. 07:22.960 --> 07:27.530 But I do want to just highlight this beginning bit if we have a look at the scene now. 07:27.570 --> 07:30.190 It jerks all over the place. 07:30.340 --> 07:35.640 You see that twitching it was only for a couple of frames but it's still very noticeable. 07:35.650 --> 07:37.280 Now there's something I can do here. 07:37.280 --> 07:38.650 That's really easy. 07:38.650 --> 07:41.740 I get my starting frame and I set it at 10:00. 07:41.920 --> 07:48.400 I'm not quite sure actually where that wobble occurred but I'm pretty sure as buelow 10 hour the 10th 07:48.400 --> 07:53.430 frame so that's one way of dealing with that just cut it off it was a problem right at the beginning. 07:53.440 --> 07:57.100 Who cares if no one sees the first half a second. 07:57.100 --> 08:03.640 However if you do need to make sure that that works we need to investigate why it wobbles and it will 08:03.640 --> 08:04.870 wobble like it does. 08:04.870 --> 08:08.690 There we can see that it's shifting from one place to another. 08:09.050 --> 08:14.520 And we as a whole scene is shifting because of some reason I can't tell you from here. 08:14.520 --> 08:17.720 I need to go back into the motion tracking you. 08:18.160 --> 08:23.980 And let's have a look at the graph editor of this point and see if we can gleam any information as to 08:23.980 --> 08:25.940 what is going on here. 08:25.960 --> 08:31.170 So one of these I can see is jerking and moving in the preview pane here. 08:31.180 --> 08:32.890 In fact the one we've got selected. 08:32.890 --> 08:35.660 So all of a sudden it jerks from one place to another. 08:35.770 --> 08:41.410 So we can literally go in and fix that for this one particular frame that it's wrong on. 08:42.160 --> 08:43.110 And place it. 08:43.270 --> 08:45.860 Not sure where each frame was eeva side. 08:46.000 --> 08:51.140 And that's always a good place to check a case that has fixed that one. 08:51.250 --> 08:55.420 Could it be an Nach just that one was jerking all over the place. 08:56.250 --> 08:58.640 No that also has a big shift to it. 08:58.680 --> 09:03.440 I don't know what was going on at the beginning so I'm just going to move it over here. 09:03.600 --> 09:09.090 So it's literally frame one frame to frame three is fine. 09:09.360 --> 09:14.220 So let's go to the next one and I may have to repeat this several times in order to refine it. 09:14.220 --> 09:15.360 Does this also jump. 09:15.360 --> 09:16.830 It does. 09:16.860 --> 09:22.560 How bizarre that each one happens now because it's every every time. 09:22.560 --> 09:25.640 And rather consistent annoyingly consistent. 09:25.980 --> 09:31.740 I can pretty much go through and fix all of these because it looks like they just need to be moved over 09:31.740 --> 09:32.920 ever so slightly. 09:32.970 --> 09:36.990 Does this one whatever it does does it do the same as some knee jerk. 09:36.990 --> 09:38.570 Yes it does. 09:38.580 --> 09:43.360 So there's something in the footage there that's causing it to jerk about and I don't like it. 09:43.410 --> 09:48.330 I'm going to go away and fix this little issue and I see you guys back in a few moments. 09:49.920 --> 09:58.530 In fact it looks like here that the issue stems from the camera not moving for one particular frame 09:58.560 --> 10:03.600 or perhaps are a problem with the video it could be something along those lines. 10:03.750 --> 10:11.310 But it looks like all of these frames are just shifted over by a tiny tiny amount. 10:11.460 --> 10:18.690 It looks like just placing the marker on top of the previous one is working absolutely fine which seems 10:18.690 --> 10:21.460 a bit of a cop out really out of thoughts. 10:21.540 --> 10:23.970 There was a fundamental reason why it was doing that. 10:24.030 --> 10:30.960 But it seems to have affect every one browser is a skip in the video but it doesn't take too long to 10:30.960 --> 10:35.430 go through and just move everyone over everyone else. 10:35.430 --> 10:37.840 So I think that's now sorted. 10:37.890 --> 10:38.690 Now I have done that. 10:38.690 --> 10:44.400 I'm going to have to go through and solve the camera motion again way down to one point eighty two as 10:44.400 --> 10:48.300 a couple of other things we can do to make sure things are a bit more steady. 10:48.300 --> 10:53.290 Now we've got a key frame a in a key frame B in this is where your 8 frames have to exist. 10:53.430 --> 10:59.400 If I said Keen frame A to 0 and press solve camera motion you'll get the error at least eight common 10:59.400 --> 11:02.730 tracks need to be there for reconstruction. 11:02.730 --> 11:07.520 What do you pick in order to what key frames do you pick. 11:07.680 --> 11:11.960 Well usually you pick the cleanest area of your footage for. 11:11.990 --> 11:16.440 So from 1 to 30 is only that much. 11:16.440 --> 11:17.370 Is that enough. 11:17.370 --> 11:19.910 Well we could pick from 10 too. 11:19.920 --> 11:26.130 Where does this stop moving around 70 let's do tend to 70 and see if we get a better solve error because 11:26.130 --> 11:28.440 of that two point six. 11:28.440 --> 11:33.540 So the answer there is no by picking that one but maybe 70 to a hundred and fifty. 11:33.540 --> 11:37.800 This is one of these ones where you can go through and you can check. 11:37.840 --> 11:45.960 So we get about a result one point a to have the whole lot from frame 1 to frame what our 230 solve 11:45.960 --> 11:48.270 caromed motion three point three two. 11:48.270 --> 11:50.550 So there are various ones that are good. 11:50.550 --> 11:51.990 There are various ones that are worse. 11:51.990 --> 11:57.510 One to 20 gives us two point six one to 40 gives us one point eight. 11:57.510 --> 11:58.850 So back to that one point eight. 11:58.860 --> 12:04.890 If that looks like a pretty good setting and then once more hop back to the foreign again if you got 12:04.890 --> 12:11.370 a 3D view open you can hop back to us was around for can review and see let's see what happens now over 12:11.370 --> 12:13.590 those first few frames. 12:14.670 --> 12:16.620 It's no longer jerking about. 12:16.660 --> 12:21.000 So now we have a 3-d scene set up being tracked. 12:21.000 --> 12:26.860 That means we can start placing things in this scene and giving them things like shadows. 12:26.900 --> 12:27.260 Oh. 12:27.320 --> 12:29.180 Oh there was a little jerk there. 12:29.190 --> 12:32.430 This is why this playback is very important. 12:32.460 --> 12:34.610 Is that going to make a difference in the playback. 12:34.620 --> 12:35.880 Possibly so. 12:35.940 --> 12:40.240 I'm not sure where that was but I'm going to go and play it back and have another look. 12:40.260 --> 12:43.990 And of course it's time for you guys to have a challenge. 12:44.770 --> 12:48.820 I'd like you to go ahead and get a good camera solve. 12:48.830 --> 12:51.200 So aim for a low error rate. 12:51.220 --> 12:56.760 Now I have been rather fortunate in this particular scene that I've managed to get it so low so quickly. 12:56.890 --> 12:58.230 That doesn't always happen. 12:58.240 --> 13:05.380 And you often do have to refine and that just comes with practice and getting used to how the marker's 13:05.470 --> 13:07.900 act what bounds you set them in. 13:07.900 --> 13:10.370 And sometimes it's just the luck of the footage. 13:10.390 --> 13:13.850 Some footage is a lot harder to track than others. 13:13.870 --> 13:17.060 Remember to check the actual footage speaking about it. 13:17.110 --> 13:21.600 You can actually see little wobbles as it's going through in the preview. 13:21.600 --> 13:24.190 Pain in the properties panel helps with that. 13:24.190 --> 13:25.270 Is it stable enough. 13:25.270 --> 13:29.200 Now this refers specifically to the 3D editor. 13:29.350 --> 13:36.100 So when you're looking through the camera with the background turned on are you seeing the cube or the 13:36.100 --> 13:37.860 ground plane that's placed there. 13:37.900 --> 13:42.880 Slide abouts on your scene or do they look like they're part of the scene. 13:42.880 --> 13:45.880 If they look like they're part of the scene that's probably stable enough. 13:45.880 --> 13:52.890 I've got a couple of those little jerks I can go through and fix and I would want to fix them for completeness. 13:53.080 --> 13:59.440 However thinking about it at the same time if I've just got a plane flying over a over the surface a 13:59.440 --> 14:03.110 little jerk may not actually be apparent in the video. 14:03.380 --> 14:04.830 So it's worthwhile knowing. 14:05.620 --> 14:11.920 Refine your markers add new or even delete more markers if necessary. 14:11.920 --> 14:14.920 If you find that brings the solara down. 14:14.950 --> 14:17.320 Then go ahead and sort that out. 14:17.530 --> 14:22.260 Now check the track preview in the properties panel isn't wobbling about too much. 14:22.360 --> 14:26.950 That should look pretty static throughout the entire time line playback. 14:27.100 --> 14:29.450 If it's not then you know there's somewhere else. 14:29.470 --> 14:35.620 You can focus your eye a word and it's certainly great as previewing each one of the markers you've 14:35.620 --> 14:39.130 got in that panel and making sure it's not jumping about. 14:39.130 --> 14:41.100 Now I think I've spotted another jump. 14:41.110 --> 14:45.550 I'm going to have to have a closer look at that and I think it's similar to what happened before the 14:45.550 --> 14:46.690 one we fixed. 14:46.930 --> 14:51.890 So go ahead pause a video now and go get yourself a good camera solve. 14:53.830 --> 14:54.450 Okay guys. 14:54.460 --> 14:55.430 Welcome back. 14:55.570 --> 14:58.070 Let's hop back on over in a blender. 15:00.150 --> 15:04.630 Okay so let's go and see if we can refine this even further. 15:04.890 --> 15:08.550 I'm going to go straight back to the motion tracking side of things. 15:08.580 --> 15:12.110 And let's have a look and see what we can do now. 15:12.120 --> 15:17.060 Obviously selected everything there don't want everything selected can be a dangerous thing to do. 15:17.100 --> 15:23.130 And I was playing about's and I managed to get an average solver of one for five when I was playing 15:23.130 --> 15:23.780 about. 15:23.820 --> 15:25.280 So just a word of warning. 15:25.290 --> 15:33.310 It is quite easy to screw things up and I set my key frames here back to 40 and solve camera motion. 15:33.320 --> 15:37.340 There we go back down to a sensible level down back down to five point six. 15:37.350 --> 15:42.450 Now one of the things I did notice and I was doing a bit of reading on camera sensors and really into 15:42.450 --> 15:50.730 my photography and I realize that camera sensors when they're doing video often don't use the full sensor 15:50.730 --> 15:53.100 with they are smaller. 15:53.100 --> 15:55.290 They use part of the sensor. 15:55.290 --> 15:57.660 Now that will have the effect of each other. 15:57.660 --> 16:03.300 We could either lower the sensor width making it smaller or we can go ahead and change the focal length 16:03.300 --> 16:06.720 because they're interchangeable sort of. 16:06.720 --> 16:11.760 So the smaller the smaller the sensor the larger the larger the apparent focal length. 16:11.910 --> 16:17.760 And we can see that these things are linked because as I move one the other moves as well. 16:18.100 --> 16:24.390 So it's going to work out to be slightly smaller do we make the sensor smaller all that would make sense. 16:24.420 --> 16:29.360 So let's make the sensor with 35 mil and do solve camera motion. 16:29.370 --> 16:36.180 Five point six let's do thirty three point six thirty two five point six. 16:36.210 --> 16:38.580 It looks like we're locked at five point six there. 16:38.580 --> 16:39.430 Does it work. 16:39.450 --> 16:44.490 If we move our way a minute right now is moving that of course it will be locked in and I just told 16:44.490 --> 16:47.270 you guys that it moves about it will move this other one. 16:47.370 --> 16:49.430 So it makes more sense to change the focal length. 16:49.620 --> 16:55.740 So now with the focal length it would go the other way it would go up apparent size. 16:55.740 --> 17:01.340 So let's go through 22 has gone to 16 that's go to 24. 17:02.860 --> 17:07.020 And as God to all one point nine now that's starting to look good. 17:07.240 --> 17:13.840 So 24 Let's try 23 because these numbers do jump about are point seven nine brillion. 17:13.840 --> 17:16.450 I'm really happy that that's worked out. 17:16.630 --> 17:19.300 So I'm just going to Prescot solve Kara motion again. 17:19.300 --> 17:21.000 My eyes are not deceiving me. 17:21.040 --> 17:25.420 So it does look like with the information we've got here. 17:25.420 --> 17:32.530 We can say that the focal length when taking a for a video on this particular device works out to be 17:32.530 --> 17:35.490 roughly twenty three millimetres give or take. 17:36.250 --> 17:41.020 I'm not going to take that as a given for the moment but there we go we've got that sorted. 17:41.660 --> 17:46.610 Let's see how this looks when we go back to our 3-d view. 17:46.610 --> 17:52.620 Let's go back to the faults and play that through now we've got the camisoles starting from frame 1. 17:52.790 --> 17:57.170 So still a little bit of a jerk at the beginning that I need to go back and sorts. 17:57.680 --> 17:59.130 And there was another one. 18:00.100 --> 18:01.910 You need to be keen eyed. 18:02.700 --> 18:04.070 There it is. 18:04.110 --> 18:08.060 And that's around frame 45 something like that sir. 18:08.070 --> 18:09.920 I need to have a closer look at those. 18:09.930 --> 18:13.520 I'm going to go do that now and see you guys back in a few moments. 18:15.550 --> 18:22.420 Hey so I'm busy going through these particular key frames and taking a close look at them as it as we're 18:22.420 --> 18:25.890 approaching I'm going to jump through with a timeline here. 18:25.990 --> 18:28.900 We reach frame 48 49 50. 18:28.900 --> 18:32.860 It jumps and then at some point it comes back again. 18:33.570 --> 18:39.830 Now what we can see on the graph editor at the top is we can actually see this the Vission this this 18:39.830 --> 18:44.520 jumpier the marker we've got selected and we've got this green party here. 18:44.520 --> 18:46.380 That is what we're looking at. 18:46.440 --> 18:53.640 So as we come through we can see that that is where all of a sudden jerks and then jerks back again. 18:53.640 --> 18:59.370 The other way as it corrects itself now in this particular case it does look like hucking. 18:59.370 --> 19:05.940 Just go back in time to where the error occurs and just re track forward from that point. 19:06.060 --> 19:09.210 And that does appear to solve the issue. 19:09.210 --> 19:15.480 Of course I still need to do that for all of the markers in our scene but that's somewhat better than 19:15.540 --> 19:17.670 having to manually place them all. 19:17.670 --> 19:21.840 So again this seems to go wrong frame 49. 19:21.840 --> 19:23.910 Does this also have a dip on the graph. 19:24.060 --> 19:25.060 Yes it does. 19:25.080 --> 19:31.070 So again I'm going to play that forward how we can see it smoothes out those great big dips. 19:31.080 --> 19:34.640 Now we've got another dip here so if I go to frame forty nine. 19:35.350 --> 19:39.430 I'm just literally going to jump there at this point and just play forward from here. 19:39.430 --> 19:44.520 It also jumps there are play that forward and that smoothed out the dip as well. 19:44.590 --> 19:49.510 So it looks like during the tracks it has mucked up somewhere. 19:49.510 --> 19:54.450 It's as simple as that does this jerk about around there or not. 19:54.580 --> 19:58.230 Yes something happens a bit around the same area. 19:58.240 --> 20:03.700 In fact now is not as pronounced because the movement isn't as pronounced. 20:03.790 --> 20:10.160 It's as simple as that so if I play forward that should have solved that and it should have read tracts. 20:10.180 --> 20:11.430 Yes it has. 20:11.500 --> 20:13.660 And again I have to go through the recipes and sort that out. 20:13.690 --> 20:15.530 I'll be back in a few moments. 20:18.830 --> 20:19.870 And I'm back again. 20:19.880 --> 20:23.450 And I think I've fixed all of my little problems. 20:23.450 --> 20:27.370 I've toyed about with the focal length for twenty two point eight. 20:27.440 --> 20:34.820 With all of these fixed all of these markers face I'm getting a Solve error of point 6 three which is 20:35.090 --> 20:36.090 excellent. 20:36.110 --> 20:42.140 I can now go to my default view and just play back what I've got. 20:42.170 --> 20:48.120 I've also decided just truncates the very beginning at five frame five. 20:48.290 --> 20:54.920 And as we go through I can't see any sudden jerks which is exactly what I'm looking for because now 20:54.920 --> 21:02.150 I can go ahead and build up the 3D scene with the plane flying through it's exciting stuff. 21:02.150 --> 21:03.680 How did you guys get on. 21:03.680 --> 21:10.100 I hope you've managed to place your cube and plane your ground plane in this scene so that it looks 21:10.100 --> 21:11.410 as if it's part of it. 21:11.420 --> 21:17.610 Obviously it doesn't look like it's at the moment because we've got nothing applied to these models. 21:17.690 --> 21:22.690 However we're well on our way now to making this work. 21:22.970 --> 21:27.670 Shea a work in the discussions perhaps a little animated gif of what's going on. 21:27.860 --> 21:30.830 And I will see you guys in the next lecture.