[00:00.830 --> 00:03.530]  Thank you guys so much for having me today.
[00:04.190 --> 00:12.690]  Now, when we think about films, I would say some of the most iconic moments in film and TV are at the hands of practical effects.
[00:13.570 --> 00:25.090]  And some of the examples, of course, are the shark from Jaws, the giant T-Rex in Jurassic Park, and of course R2-D2 and C-3PO from Star Wars.
[00:25.090 --> 00:33.970]  But these are easy examples, and each of which are technological marvels in their own right, but they were created for one very specific purpose.
[00:34.350 --> 00:35.870]  And that was to tell a story.
[00:36.190 --> 00:41.370]  They weren't designed to be technological marvels only, they were designed to tell a story.
[00:42.510 --> 00:48.050]  And practical effects, I would say again, have had an outsized impact on the history of film.
[00:48.050 --> 00:54.530]  And in my opinion, I think this discipline has punched above its weight, both on screen and off.
[00:54.730 --> 01:00.210]  When we think about Star Wars, there's a lot of plot and a lot of characters, but we still think about those plucky little droids.
[01:00.570 --> 01:04.050]  And again, all of which were handmade by these craftsmen.
[01:05.350 --> 01:11.010]  So how do we bring that level of impact off or outside of the world of film and TV?
[01:11.010 --> 01:16.430]  What does that look like when we take what I think is the secret sauce and apply it elsewhere?
[01:18.250 --> 01:23.590]  Now, I've been working in film for almost a decade, and roughly three years ago, I started a YouTube channel.
[01:23.750 --> 01:31.910]  And the only reason I mention that is because one of my builds, and most of which are fun, wacky prop builds, special effects builds.
[01:32.110 --> 01:38.550]  But one of my builds, I got a comment, and I think this comment perfectly encapsulates what I want to talk about today.
[01:39.210 --> 01:43.810]  And it was someone who said, don't try to make things which don't make sense.
[01:44.390 --> 01:51.070]  And the reason that I love this comment so much is because in order to fix it, all you have to do is cross out one word.
[01:51.070 --> 01:54.510]  And I think we have what I would say is the secret.
[01:54.770 --> 01:57.770]  And that is try to make things that don't make sense.
[01:58.210 --> 02:07.470]  And it seems a little bit silly, but at the end of the day, we have to look at things beyond just what they do on screen and think, well, what are we learning?
[02:07.650 --> 02:09.550]  What is the process that goes into building this?
[02:10.370 --> 02:15.810]  And to tell this story, we first have to go back to the first big special effects job I ever landed.
[02:16.050 --> 02:21.150]  And that was as the electronics expert and one of the presenters on Motor Mythbusters.
[02:21.470 --> 02:27.050]  And if you watch the original show, you know that they're no stranger to building things that don't make sense.
[02:27.050 --> 02:28.890]  It's a giant lead balloon.
[02:29.130 --> 02:33.230]  Not very practical, but in the sense of the show, it did make sense in that regard.
[02:33.430 --> 02:35.910]  But again, you never knew what they were going to build.
[02:36.030 --> 02:37.830]  And that was the joy of that project.
[02:38.570 --> 02:40.690]  And this show was fortunately no different.
[02:40.950 --> 02:42.790]  We just took a slightly different approach.
[02:42.930 --> 02:45.590]  They brought back Tory Belleci, one of the original presenters.
[02:45.870 --> 02:49.510]  And now we were focused on testing myths around cars.
[02:50.270 --> 02:58.390]  And one of the first myths that we had to test, and one of my personal favorites, was the myth of, can you accelerate out of trailer sway?
[02:58.870 --> 03:02.910]  And trailer sway, if you're not familiar, is if you have, in this case, a caravan.
[03:03.110 --> 03:10.370]  And if you've loaded a trailer incorrectly, at speed it will start to fishtail, which is to sway violently back and forth.
[03:10.550 --> 03:14.750]  And that's because the weight is not correctly placed on the tongue of the trailer.
[03:15.390 --> 03:20.190]  And traditional wisdom says, well, you're supposed to slow down, which makes sense.
[03:20.330 --> 03:24.570]  But what we wanted to find out is, like a loose rubber band, could you pull it straight?
[03:24.770 --> 03:29.610]  Could you accelerate and take this situation from a problem into a solution?
[03:30.990 --> 03:33.170]  To do that, we would be using nitrous.
[03:33.530 --> 03:38.170]  And if you've ever watched a Fast and Furious movie, you're probably very familiar with what this is capable of.
[03:38.290 --> 03:49.950]  But that was the idea, is that we would get this up to speed, we would induce trailer sway, and then we would see what would happen if you put the pedal to the metal and gave it a hit of acceleration, and would that pull it straight?
[03:50.710 --> 03:54.270]  Now, obviously, this is a pretty dangerous thing to test.
[03:54.450 --> 03:58.670]  And the producers were a little bit hesitant to put someone in the car.
[03:58.850 --> 04:08.530]  So my job was to turn this Dodge Ram 1500 into a full-size remote-control car, including the Nitra system.
[04:08.830 --> 04:10.750]  And the requirements were very simple.
[04:10.750 --> 04:15.910]  For example, I had a $500 budget, I had three weeks, and I couldn't kill anyone.
[04:16.690 --> 04:18.030]  That also extended the truck.
[04:18.130 --> 04:19.250]  The truck couldn't kill anyone either.
[04:19.410 --> 04:29.070]  And if you had seen some of the older seasons of Mythbusters, you knew that a remote-control car breaking loose and getting off and running off into the distance was not unheard of.
[04:29.150 --> 04:38.210]  So there was a bit of weight on my shoulders to make sure that this didn't happen, because we would be testing this on an airstrip that was an active airport.
[04:38.470 --> 04:39.970]  So there were planes and things happening.
[04:40.190 --> 04:42.030]  Our section of the runway was closed.
[04:42.230 --> 04:44.530]  But again, it couldn't just run off into the woods.
[04:45.610 --> 04:49.150]  And now, I didn't have any experience building something like this.
[04:49.250 --> 04:52.990]  There was no manual, and all the previous versions were compressed air.
[04:52.990 --> 04:55.710]  So I wanted to go with an electronic systems control.
[04:56.030 --> 04:59.510]  But again, there's no laid-out path.
[05:00.050 --> 05:04.710]  However, as one of my favorite quotes, the path is made by walking.
[05:04.970 --> 05:13.270]  So I found some scrap parts lying around the shop and got to work working through failure after failure after failure after failure.
[05:13.450 --> 05:15.490]  Fortunately, all of which were on the workbench.
[05:15.910 --> 05:25.090]  But again, none of that would have happened without being able to first look at the problem and go, well, I'm going to have to get the first failure out of the way before any of the real work can get done.
[05:25.250 --> 05:27.530]  And so eventually, this was the system that we landed on.
[05:28.030 --> 05:30.410]  And it was, again, all electronically controlled.
[05:30.510 --> 05:33.730]  But the idea being that we could control the throttle, the brakes.
[05:34.230 --> 05:35.650]  We had a kill switch.
[05:35.830 --> 05:36.850]  We had our nitrous switch.
[05:36.910 --> 05:42.990]  And it was all being controlled at the push of a button, just kind of an off-the-shelf remote control plane receiver.
[05:43.390 --> 05:45.290]  And under the hood, it's nothing special.
[05:45.290 --> 05:48.130]  Again, we didn't have a lot of money or a lot of time.
[05:48.270 --> 05:50.050]  So these are all off-the-shelf components.
[05:50.210 --> 05:53.270]  Arduino components, PWM, motor controllers, nothing fancy.
[05:53.490 --> 05:58.650]  But the important part was we needed to test the myth and we needed to see if we could get the results.
[05:58.810 --> 06:00.110]  And that was the crucial part.
[06:01.410 --> 06:09.270]  And most excitingly, I had basically no time because the day of filming, we had to use this truck for another shoot that we needed.
[06:09.310 --> 06:21.650]  So I had one hour to essentially convert this truck from a normal human-driven car into a fully remote-controlled truck, not knowing if I was going to be able to test it and essentially just having to go into the myth blind.
[06:22.990 --> 06:25.670]  In the end, we only had three tries to get this done.
[06:26.070 --> 06:31.250]  And somewhere in there, we had to get the results that we needed to either confirm or bust the myth.
[06:32.210 --> 06:34.070]  And fortunately, the truck did not run away.
[06:34.230 --> 06:37.970]  And you can watch the full episode if you want to see exactly what happens.
[06:38.130 --> 06:44.110]  But I can tell you right now, when you apply nitrous to a truck with trailer sway, the end result is nothing good.
[06:44.410 --> 06:46.950]  But results were what we were going after.
[06:47.150 --> 07:02.550]  And even though the truck may not have fared tremendously well after the end of this, it was still the beginning of the process where we tackle ridiculous problems and looking at it from, again, that position of try to make things that don't make sense.
[07:03.650 --> 07:12.930]  Now, regardless of whether we were blowing up cars or building them, it was here that I learned the first reason for why we build things that don't make sense.
[07:13.590 --> 07:18.170]  And that's because it increases the number of angles from which you can approach future problems.
[07:18.650 --> 07:21.130]  That was something that sounds obvious now.
[07:21.330 --> 07:29.130]  But at the time, when you think of problems from an engineering mindset, you're looking at, well, what is the first and most commonly used solution?
[07:29.290 --> 07:32.130]  And that's not inherently a bad thing, but this is showbiz.
[07:32.230 --> 07:33.490]  So we had to come up with something new.
[07:34.210 --> 07:39.670]  And again, most people won't be asked to blow up cars or to bust myths, but they will be asked to solve problems.
[07:40.490 --> 07:49.970]  And again, by increasing the number of angles that you approach these problems, you're going to get better at solving them and you're going to get more flexible at solving problems when something goes wrong.
[07:50.370 --> 07:52.890]  And best of all, this is an observable phenomenon.
[07:53.530 --> 08:01.810]  Back in 2014, researcher Dr. Michael Hausman was analyzing thousands, tens of thousands of applications from call center applicants.
[08:01.970 --> 08:09.670]  And what the company wanted to find was what were the markers in these applications that would indicate an individual who would outperform.
[08:09.670 --> 08:20.030]  And what they found was that the single strongest predictor of employee performance, and this is back in 2014, was what web browser they used.
[08:20.390 --> 08:40.300]  Now, it wasn't because the web browser itself was the secret ingredient, but Chrome users and Firefox users outperformed their peers by almost 20%, and they stayed with that company 15% longer than their compatriots who were using Internet Explorer or Safari.
[08:41.180 --> 08:45.880]  Now, again, these browsers aren't magically making these employees better at their jobs.
[08:45.940 --> 08:49.560]  It was really just an indicator of what they were really looking for.
[08:49.740 --> 09:04.700]  And that was essentially, were these employees looking at the problem at hand and accepting the default solution that was presented to them, which in this case was quite literally the default browser, or were they approaching this problem from new angles and thinking, okay, what else is there?
[09:06.820 --> 09:10.240]  What other solutions exist that aren't just immediately presented?
[09:11.620 --> 09:17.120]  And so this idea of asking unconventional questions goes beyond just your browser choice.
[09:17.120 --> 09:23.200]  And to come back to special effects, it leads into one of my favorite builds of all time.
[09:23.800 --> 09:26.240]  And that is, of course, Newt the Robot.
[09:26.580 --> 09:33.960]  This was a robot that I got to build back in 2022, and it was a phone call that I would say every maker dreams of.
[09:34.180 --> 09:43.360]  This was essentially them saying, we are building, or we are making a new movie, and we want to build a robot character from the ground up.
[09:44.380 --> 09:49.220]  And, of course, that's what everybody wants to hear, but what does that really mean?
[09:49.340 --> 09:51.720]  What are some of the questions that you have to ask?
[09:51.840 --> 09:54.780]  And what does that look like as you go through that build process?
[09:55.700 --> 09:58.580]  And what made it so fun is that we had complete creative control.
[09:58.840 --> 10:03.260]  We had a script and, essentially, a blank space wherever the robot was.
[10:03.360 --> 10:05.900]  So we didn't have to comply with any existing IP.
[10:06.100 --> 10:11.840]  We could sit down and go, okay, what does a robot whose sole job is to tell a story, what does that look like?
[10:13.520 --> 10:17.660]  And so, normally, when you build a robot, we think of an industrial application.
[10:17.760 --> 10:21.520]  You're asking yourself, okay, what is the function of this robot?
[10:21.640 --> 10:24.200]  What are the services that it has to perform?
[10:25.100 --> 10:29.580]  And what is really interesting in our world is, what happens when form is the function?
[10:30.100 --> 10:32.920]  And in our case, that's exactly what we were designing around.
[10:33.460 --> 10:36.240]  Everything comes down to, this is a kid's movie.
[10:36.360 --> 10:39.500]  So this robot has to be a lovable, cute character.
[10:39.500 --> 10:41.280]  That is the prime function.
[10:41.520 --> 10:43.820]  It doesn't have to serve anything else.
[10:44.040 --> 10:46.640]  It's just delivering story to our audience.
[10:47.480 --> 10:55.960]  And while it sounds silly, you have to remember, practical effects engineers are not there to make assembly lines faster or surgery safer.
[10:56.280 --> 10:58.820]  Our job is to tell stories better.
[10:59.120 --> 11:03.640]  And so, again, it goes back to this idea of approaching these problems from new angles.
[11:03.900 --> 11:06.700]  How do we do these things and how do we tell stories better?
[11:07.520 --> 11:12.600]  And so, we have to ask some really interesting and some really unconventional questions.
[11:12.880 --> 11:17.260]  For example, how do we make a robot feel relatable?
[11:18.100 --> 11:23.480]  And what does it look like when emotions are expressed by an inanimate object?
[11:23.820 --> 11:28.120]  Now, obviously, we as humans are very good at anthropomorphizing things.
[11:28.420 --> 11:33.220]  Ask any kid about their stuffed animal, and that's probably a real character in their head.
[11:33.220 --> 11:38.460]  But more importantly, we have to look at these things and say, okay, what is the goal?
[11:39.040 --> 11:44.880]  And taking an inanimate object, how do we apply more to it than just what our audience is coming into it with?
[11:44.940 --> 11:46.260]  How do we craft a narrative?
[11:46.380 --> 11:49.540]  How do we craft a story around an inanimate object?
[11:50.580 --> 11:53.580]  And we do that by showing, not telling.
[11:53.800 --> 12:00.180]  Because everything is designed to basically squeeze emotion out of the parts that we have available to us.
[12:00.860 --> 12:05.080]  And whether you realize it or not, this happens in every movie that you watch.
[12:05.820 --> 12:17.900]  Now, it just happens to be a little bit more distinct when you're working with robotic characters, because humans are already, you know, there's a lot of ways that we can express ourselves, but robots, it starts to become kind of clear when you put them next to each other.
[12:18.340 --> 12:23.200]  When these two characters roll on screen, you immediately know how you're supposed to feel about them.
[12:23.500 --> 12:27.000]  But when you dig a little bit deeper, there's a lot more going on under the surface.
[12:27.400 --> 12:29.480]  We're looking at what are the color palettes?
[12:29.780 --> 12:31.940]  What is the surface shine?
[12:32.200 --> 12:35.740]  It sounds silly, but, you know, things that are well-loved.
[12:35.880 --> 12:37.600]  Think of a well-loved sweater you have.
[12:37.760 --> 12:38.480]  Is it brand new?
[12:38.620 --> 12:38.840]  No.
[12:39.260 --> 12:40.440]  But it's why you love it.
[12:40.520 --> 12:42.420]  It's because it's got maybe a few threaded holes.
[12:42.600 --> 12:44.460]  Maybe it's got a little bit of a hard...
[12:44.460 --> 12:45.880]  You know, it's got some hard miles on it.
[12:46.280 --> 12:49.240]  And so these are the different elements that go into character design.
[12:49.420 --> 12:53.840]  And again, in this case, it's a mechanical thing, but you'll also see this on human characters in movies.
[12:53.840 --> 13:00.780]  And so I encourage you to pay attention to these things, because we can also dial it up even more in the world of animation.
[13:01.120 --> 13:06.380]  Because now we're not limited by physical materials, or physical rules, or even physics.
[13:06.780 --> 13:17.500]  So now we can create characters that really push the bounds of, all right, how obvious can we make it that one of these characters is bad, and one of these characters is good?
[13:18.180 --> 13:21.200]  Everything down to symmetry, to color, to wear.
[13:21.700 --> 13:31.300]  These are all things that, again, the storytellers are not going to sit down and do exposition and sit and tell you these things that you need to know about these characters.
[13:31.900 --> 13:36.880]  They're just going to show it to you as soon as that character comes on screen, and you're going to accept it for that and move on.
[13:37.100 --> 13:39.360]  It's time savings is really what it comes down to.
[13:41.120 --> 13:46.680]  So knowing all of that, the director and I now have to build our own character.
[13:46.980 --> 13:56.600]  And so we start by creating sketches and asking ourselves again, this is a kid's movie, so what are the shapes, what are the sizes that are going to be friendly to a kid?
[13:56.880 --> 14:02.300]  We're not going to build, and as much as I love Johnny Five, we're not going to build a Johnny Five-style robot, which is huge.
[14:02.440 --> 14:04.280]  That's going to be intimidating to a kid.
[14:04.860 --> 14:09.180]  These characters have to exist within the world and within the relationships.
[14:09.180 --> 14:11.240]  that we are building as part of the story.
[14:12.100 --> 14:15.500]  And so this is kind of just the rough sketches that we did for this process.
[14:15.620 --> 14:21.100]  And my job as an engineer is to also kind of be the bad cop, as I like to say.
[14:22.020 --> 14:30.300]  And that essentially means the director comes to me and says, well, I would love to have a one-wheel design where he balances and rolls around, and I say, that would be really cool.
[14:30.400 --> 14:31.520]  And we can do that.
[14:31.800 --> 14:40.200]  However, let's think a little bit more on the logistics side, because we're going to have to film for 12, 13, 14 hours a day, and that's going to burn a lot of energy.
[14:40.420 --> 14:41.520]  It's going to be a lot of complexity.
[14:41.680 --> 14:42.840]  A lot of things could go wrong.
[14:43.160 --> 14:51.780]  So the final design we ended up with was a three-wheel design that, that way, when the robot wasn't functioning, it could essentially sit there and not burn any additional power.
[14:52.080 --> 14:55.260]  These are some of the additional, you know, logistical challenge that we look for.
[14:55.980 --> 14:58.260]  And so we brought in a concept artist.
[14:59.020 --> 15:09.040]  And this is one of my favorite parts, because I'm not a great illustrator, and part of the build process, I would argue, is recognizing what you're not good at and getting help in doing that.
[15:09.240 --> 15:21.560]  And so we brought in this concept artist to essentially flesh out the details, flesh out the small things, and go, okay, well, we like the body from this design, and we like the head from this design, and the antenna from this, and so on and so forth, and figure out what are these little ingredients?
[15:23.840 --> 15:27.440]  What about each design do we gravitate towards and why?
[15:27.800 --> 15:31.140]  And more importantly, what would the character gravitate towards?
[15:31.140 --> 15:34.800]  The character of Miles in the film who would have built this robot?
[15:34.800 --> 15:37.620]  What would be going through his head when he's building this?
[15:37.800 --> 15:39.220]  And how do we recreate that?
[15:40.760 --> 15:43.340]  And eventually we land on our finished design.
[15:43.720 --> 15:47.800]  Everything down to the wear and tear, the soft, round shapes.
[15:48.160 --> 15:50.060]  There's no hard angles in this.
[15:50.060 --> 15:52.940]  There's no shiny, sleek, black appearance on the outside.
[15:53.200 --> 15:59.220]  This is designed to look like essentially your lovable family dog that you've been in the family for years.
[15:59.400 --> 16:00.080]  It's scruffy.
[16:00.220 --> 16:02.020]  He's got some wear and tear around the edges.
[16:02.240 --> 16:06.460]  He's soft, and you kind of just want to pinch his little cheeks, as I would say.
[16:06.720 --> 16:10.080]  But again, these are all very cognizant design choices.
[16:10.360 --> 16:15.140]  And again, when you go and watch movies, think about these things next time you see characters on screen.
[16:15.380 --> 16:17.940]  What are the little details that you might not have noticed?
[16:17.940 --> 16:20.740]  And maybe ask yourself, why are those details there?
[16:21.000 --> 16:22.640]  Why are those colors being used?
[16:22.760 --> 16:24.040]  Why are those shapes being used?
[16:24.360 --> 16:31.880]  Because somebody had to sit down and make a very cognizant decision to say, well, this is what we want you to feel, and this is how we're going to achieve that.
[16:32.700 --> 16:36.640]  And so this also plays into kind of what I was talking about earlier.
[16:36.640 --> 16:38.920]  There are some of the logistical advantages as well.
[16:39.520 --> 16:43.820]  And one of which, famously R2D2 piloting this, there's no moving mouth.
[16:44.420 --> 16:50.640]  And what we ended up learning was this was great because we can now fill a lot of plot holes in post-production.
[16:50.920 --> 16:58.740]  So if there's anything that we might have forgotten to shoot on the day when we were filming, well, now you can just cut back to the same insert of Newt, our little character here.
[16:58.740 --> 17:03.120]  You can put new lines in his mouth, and we can now fill a plot hole and move on.
[17:03.480 --> 17:12.280]  So it's fun to also kind of factor in some of these logistical challenges when designing a robot, even though emotion is the overall driving force behind it.
[17:13.000 --> 17:16.060]  And one of the other things, too, is we have minimal facial movement.
[17:16.200 --> 17:19.520]  So again, I mentioned Johnny Five, you have eyebrows, you have things like that moving around.
[17:19.520 --> 17:24.180]  That adds a lot more nuance to a performance, which, when done well, is really nice.
[17:24.360 --> 17:35.360]  But when you're working on a smaller budget and with a smaller team, you can now use the dialogue, and you can use more exaggerated movement of the rest of the body to drive the emotion.
[17:35.600 --> 17:39.740]  So it helps make my job a lot easier when I'm actually operating him as well.
[17:41.440 --> 17:45.140]  Now, this part is going to look a lot more familiar, I think, to the engineering people in the room.
[17:45.220 --> 17:48.380]  And this is just the part where I now have to actually build this.
[17:48.380 --> 17:53.840]  And as I like to say, I'm essentially a full-stack engineer when it comes to this from designing it in CAD.
[17:54.400 --> 17:59.980]  Everything needed to be 3D printed because I wanted to dial in the exact look of every single part.
[18:00.200 --> 18:05.000]  And that is super easy to do when you can 3D print something again and again and again until it's just right.
[18:05.380 --> 18:11.520]  And speaking of which, you can also now have, as long as you have a 3D printer, an unlimited supply of spare parts.
[18:11.860 --> 18:19.200]  That's a lot harder to do when you're trying to source very specific things that maybe were only manufactured for small runs or are hard to come by.
[18:19.400 --> 18:24.420]  So again, the idea being we want something completely bespoke and something that is serviceable.
[18:24.620 --> 18:27.400]  So even the tire tread was 3D printed in TPU.
[18:27.640 --> 18:28.760]  It was nice and squishy.
[18:28.900 --> 18:30.340]  We were able to get the effect of rubber.
[18:30.520 --> 18:32.260]  And on camera, it sells.
[18:33.380 --> 18:38.500]  And again, just like the remote control truck from MythBusters, my job is to make something that works.
[18:38.500 --> 18:43.540]  At the end of the day, if we cannot film, then everybody is wasting their time.
[18:43.840 --> 18:49.820]  And the whole production is essentially hinged on whether or not this robot, considering he's a pretty large character, can do his job.
[18:50.020 --> 18:55.740]  So having off-the-shelf components makes it a lot easier knowing that I can overnight something from Amazon if something breaks.
[18:55.920 --> 18:58.960]  And I'm not relying on custom parts, custom motor drivers.
[18:59.140 --> 19:06.700]  Because as fun as those are to build, if something goes wrong, I still need to be able to be back up and running by the time our call sheet comes out the next day.
[19:06.700 --> 19:10.660]  So a lot of this is just using, again, like I said, off-the-shelf components.
[19:10.940 --> 19:15.900]  That's just a remote control flight transmitter from basically an RC airplane.
[19:16.180 --> 19:18.400]  And there's a lot of RF interference on set.
[19:18.540 --> 19:22.140]  So being able to use something a little more powerful than Bluetooth was perfect for that.
[19:22.240 --> 19:23.600]  And it's the reason I went that direction.
[19:24.180 --> 19:28.100]  In terms of servos, again, off-the-shelf parts, these are coreless servos.
[19:28.360 --> 19:31.980]  High torque, but easy to get your hands on when something goes wrong.
[19:32.020 --> 19:36.380]  And I did end up having to swap out a few, which I'll detail a little bit later.
[19:36.380 --> 19:39.140]  And again, motors, brushed DC motors.
[19:39.260 --> 19:41.740]  Now, in hindsight, I probably would have done this a little bit differently.
[19:41.740 --> 19:46.020]  But there were some added benefits to going this route that, again, I'll detail a little bit later.
[19:46.340 --> 19:50.740]  And at the heart of everything was an Arduino Nano, which now is probably shocking.
[19:50.960 --> 19:53.260]  Keep in mind, this movie was made almost four years ago.
[19:53.400 --> 19:55.060]  And at the time, it was bulletproof.
[19:55.060 --> 19:56.240]  It worked every single time.
[19:56.240 --> 19:59.760]  And I was using the remote control to do most of the heavy lifting.
[20:00.160 --> 20:04.800]  I'm doing most of the PWM control out of the receiver, not the Arduino Nano.
[20:04.800 --> 20:08.620]  Now, there was a small projector inside his head that had to be turned on and off.
[20:08.780 --> 20:11.940]  Some lights and some motor driver stuff.
[20:12.140 --> 20:15.260]  But most of it was all being done out of just the receiver itself.
[20:15.640 --> 20:17.140]  Trying to keep things simple.
[20:17.640 --> 20:20.500]  And then, of course, our LiPo battery was the main power source.
[20:20.640 --> 20:22.980]  Because they're cheap, they're light, and I had to fly there.
[20:23.200 --> 20:27.860]  So that way I could actually make sure that I had the parts that I needed without going overboard.
[20:29.000 --> 20:30.620]  And, of course, packing everything in.
[20:30.720 --> 20:31.460]  It's a small fit.
[20:31.720 --> 20:32.720]  I just use Velcro.
[20:32.940 --> 20:35.640]  I slap it down on the inside, and then you can Velcro everything on and off.
[20:35.760 --> 20:36.420]  It works pretty well.
[20:36.560 --> 20:39.980]  But, again, I need to be able to jump in, repair things, and jump out.
[20:40.100 --> 20:53.100]  So every single component is designed to be easily replaceable, easily repairable, so that if something goes wrong, we're not having 50 people on set who are all now burning time on the clock waiting for one single robot to get up and running.
[20:54.320 --> 20:57.680]  And one of the most important parts is, again, going back to that wear and tear.
[20:58.260 --> 21:05.420]  Thinking about, okay, any object that exists in this universe, that exists in this world, has experienced some sort of wear and tear.
[21:05.700 --> 21:10.220]  And not to get philosophical, but so have we, and it's what makes things relatable when we see them on screen.
[21:10.220 --> 21:19.320]  And to do this, I'm just simply taking a Dremel, some sandpaper, and trying to emulate the natural wear and tear that an object like this would experience in the real world.
[21:19.340 --> 21:24.740]  If he's a robot, there's going to be some wheel rub where the tire hits the chassis, there's going to be some drips from grease.
[21:24.920 --> 21:25.760]  What does that look like?
[21:25.860 --> 21:33.380]  Again, all these tiny little details, trying to create something that actually sells what this robot is and the world that he lives in.
[21:35.720 --> 21:37.020]  And, by the way, I had to make two of them.
[21:37.780 --> 21:45.200]  So, all of the work that you saw, all of the sanding and the priming, and if you've ever 3D printed anything and tried to make it look smooth, you know how much time that takes.
[21:45.720 --> 21:46.640]  And then I had to double it.
[21:46.820 --> 21:53.840]  So, obviously, one is none, and especially in this world where something can go wrong, the studio, of course, wanted to have a stunt double.
[21:54.020 --> 22:02.260]  So, any time we were doing crazy stunts, we had what I lovingly call our team of Newt, the prime robot, and Crash, our stunt double robot.
[22:02.260 --> 22:08.180]  So, any time something was a little bit more rough and tumble, we would send Crash up to bat.
[22:08.520 --> 22:12.840]  And for most of the control on-screen stuff, we would use our regular hero Newt.
[22:13.340 --> 22:16.340]  But, best of all, if something went wrong, everything was filmed in Nashville.
[22:16.520 --> 22:18.580]  So, I couldn't access my shop tools.
[22:18.780 --> 22:25.160]  So, at least we had a full working duplicate that I could then scoop the parts out of one and put in the other if something were to go wrong.
[22:25.160 --> 22:31.160]  Because, even though the stunt double looks identical on the outside, on the inside, he's missing a few of the crucial components.
[22:31.340 --> 22:32.860]  Just purely out of cost reasons.
[22:34.880 --> 22:41.420]  But the best part of this whole project was, by far, getting to actually puppeteer the robot on set.
[22:41.720 --> 22:44.640]  And this, to me, is... and I'm a theater kid.
[22:44.900 --> 22:49.220]  So, this was an extension of, essentially, all of my theater classes in high school.
[22:49.220 --> 22:51.400]  You are becoming a character.
[22:51.660 --> 22:58.260]  But instead of your voice and your limbs being the instruments you use, it's been condensed into this tiny little controller.
[22:58.620 --> 23:00.520]  Which, in some ways, makes it a lot harder.
[23:00.540 --> 23:05.400]  Because now you have to, again, ask yourself, what does an inanimate object look like when it is expressing emotion?
[23:05.440 --> 23:06.700]  And how do we sell that?
[23:07.060 --> 23:09.120]  And in some ways, it also is a little bit easier.
[23:09.120 --> 23:15.180]  Because now I only have about four or five different ways to express emotion as opposed to everything else that we normally have at our disposal.
[23:16.680 --> 23:20.780]  But, at the end of the day, and this is kind of what I talked about earlier, there were some happy accidents.
[23:21.120 --> 23:27.620]  And part of it, too, is thinking about not just the design of the robot and the emotion there, but also the performance.
[23:28.800 --> 23:34.400]  A industrial robot is always going to pick the shortest path between A and B, but that doesn't really sell a convincing character.
[23:34.600 --> 23:39.600]  So, even in blocking, which, in film, blocking is how a character moves through the scene.
[23:39.820 --> 23:44.180]  In this case, our robot moving through the scene is not going to go in straight lines.
[23:44.180 --> 23:45.100]  He's going to meander.
[23:45.300 --> 23:49.140]  He's going to, as a kid would, take his time and get distracted and look at things.
[23:49.300 --> 23:57.820]  And it is those elements that really sell the idea that this is a living, breathing character, even though, in the back of our minds, we know that it's not.
[23:58.880 --> 24:01.060]  And there are some benefits to that, as well.
[24:01.540 --> 24:07.960]  Because we were working with children on this set, and kids love to be able to interact with a real, actual prop.
[24:08.120 --> 24:09.620]  This isn't a tennis ball and a stick.
[24:09.820 --> 24:10.800]  This isn't a green screen.
[24:10.800 --> 24:17.940]  This is the world that we are building, and it is real for them, because it actually is the robot that they're supposed to be interacting with.
[24:19.480 --> 24:28.900]  And one of my favorite aspects of this was, again, on screen, there's the three-dimensional world that I'm operating in, but there's, more importantly, what does he look like on screen?
[24:29.080 --> 24:37.900]  And a lot of that is thinking about what is this robot's relationship with the characters in the scene, and how does he move, and how does he act differently in relation to those characters?
[24:38.480 --> 24:51.920]  And the greatest compliment that I got on all of shooting was a compliment the director didn't know she gave me, and that was because, in between takes, she would always forget that I was working behind the scenes, and so she would deliver the notes to the robot and not to me.
[24:52.140 --> 24:55.700]  So she would go, oh, Newt, actually, in this take, can we actually get you to roll over here?
[24:55.780 --> 24:58.040]  And I'd be like, hi, actually, I need to hear all of these notes.
[25:04.620 --> 25:07.080]  And when the character was so young, she didn't know that E.T.
[25:07.120 --> 25:07.980]  was a prop.
[25:08.180 --> 25:11.740]  And so the special effects guys, in between takes, would keep E.T.
[25:12.340 --> 25:16.480]  alive for her, because in her mind, that was a real character in the movie.
[25:16.580 --> 25:23.900]  And so it was that that I tried to take into this world, because not only is it very cute, but also it's just why we get into this business.
[25:24.040 --> 25:24.780]  It's movie magic.
[25:25.020 --> 25:29.020]  And as schlocky as that sounds, it's so much fun when you get to see it play out in real time.
[25:30.080 --> 25:37.360]  And after filming was complete, the best part, of course, is we get our red carpet premiere, and we finally get to see all of our hard work on the big screen.
[25:37.640 --> 25:44.160]  But at the end of the day, it's still, there's one final reason why you should build things that don't make sense.
[25:44.260 --> 25:59.740]  Because even though this robot goes against what most of us would think about for designing a robotic thing, designing something that's sole purpose is usually to be functional, there's still something else that I would say in the world of building things that don't make sense.
[26:00.260 --> 26:01.740]  And that's to build it for you.
[26:02.180 --> 26:09.620]  And again, I know it sounds cheesy, but at the end of the day, so much of the building that we do feels like it has to come from a place of being pragmatic.
[26:10.180 --> 26:16.500]  And I would say, you don't need a film crew to be pointed at what you're doing for it to have value.
[26:16.980 --> 26:23.320]  The value comes from learning to work with the ingredients and not just working with a recipe.
[26:25.140 --> 26:34.680]  And in the words of Tim Urban from Wait But Why, I will leave you with this one simple question, which is, in everything you do, are you a chef or are you a cook?
[26:36.220 --> 26:38.380]  And on that note, thank you guys so much.
[26:38.540 --> 26:44.600]  I want to open up the floor to questions, if you would like, but thank you guys so much for attending my talk.
[26:51.530 --> 26:52.270]  Any questions?
[26:55.920 --> 26:56.360]  Yes.
[27:10.440 --> 27:11.580]  Yeah, great question.
[27:11.780 --> 27:17.060]  And for those who couldn't hear, asking a question about operating RF on film sets.
[27:17.240 --> 27:20.300]  And fortunately, I come from a background of camera department.
[27:20.320 --> 27:27.460]  And so much of what we do in camera department is a lot of what you talked about, which is scheduling out where the different RF frequencies are going to be.
[27:27.460 --> 27:31.220]  And on professional film sets now, you kind of have everything scoped out.
[27:31.440 --> 27:35.980]  And to give basically a brief overview, usually you have your microphones in sub-gigahertz.
[27:36.140 --> 27:40.400]  You have all your DMX controls, which is your lighting, in 2.4 gigahertz.
[27:40.900 --> 27:48.100]  Then you have, you know, hopefully if you're lucky and your IT department does their job right, you have Internet and your Wi-Fi at 5 gigahertz.
[27:48.240 --> 27:53.080]  And then our video cameras are nowadays transmitting in the 6 gigahertz spectrum.
[27:53.320 --> 27:58.820]  So it used to be maybe, you know, five, six years ago, everything was really compressed into that 2.4 gigahertz band.
[27:59.000 --> 28:01.740]  But fortunately, now we've been able to separate department by department.
[28:01.900 --> 28:04.020]  Everybody kind of spreads out and gets their entire thing.
[28:04.140 --> 28:09.700]  So in this case, now I'm also coordinating with, you know, the lighting department saying, hey, this is what I'm going to be working with.
[28:09.780 --> 28:10.400]  This is the equipment.
[28:10.580 --> 28:12.220]  What are you guys going to be using on set?
[28:12.420 --> 28:13.380]  Talking with camera department.
[28:13.400 --> 28:16.600]  Are they using older transmitters that are in 2.4 gigahertz?
[28:16.960 --> 28:22.740]  But yeah, very much, you know, there's the backend engineering side of it as well that is looking at it from a logistical.
[28:23.040 --> 28:25.200]  I have to play nice with all the other departments.
[28:25.200 --> 28:34.020]  I can't just roll up and screw everything else up because it is one giant choreographed, you know, machine that all is ideally moving in the same direction.
[28:36.300 --> 28:37.260]  Yes, other question?
[28:43.940 --> 28:45.580]  Yes, FDM printing.
[28:45.900 --> 28:48.200]  And again, it's funny.
[28:48.280 --> 28:51.660]  When you finish a project, you look back and you're like, man, I would have done that completely differently.
[28:52.240 --> 28:55.840]  And this project being no different, again, this was a long time ago.
[28:55.840 --> 29:01.640]  And so the key mistake that I made, and for anybody who does FDM printing, you're going to cringe when I say this.
[29:01.940 --> 29:13.440]  I printed Newt, the original Newt, in silk PLA because I was like, ah, well, I want, you know, a nice, like, silverly undercoat so when I sand things away, you know, you have, like, almost a metallic appearance.
[29:13.520 --> 29:17.800]  Well, the problem is, silk PLA has atrocious layer-to-layer adhesion.
[29:17.800 --> 29:21.720]  So, now, we were lucky because filming only lasted for four weeks.
[29:21.880 --> 29:26.240]  So, in that case, we didn't really have to worry about the longevity of, you know, traditional engineering problems.
[29:26.460 --> 29:33.160]  But again, you look back in hindsight, you're like, well, there's one of the failures I get to talk about now that fortunately didn't, you know, ruin anything.
[29:33.360 --> 29:35.480]  But yeah, in hindsight, you're like, well, I wouldn't do that again.
[29:35.580 --> 29:44.400]  But yes, a lot of that is just FDM printing, sitting down and for hours and hours and hours and hours, sanding it to a nice finish, putting some filler primer, all that fun stuff.
[29:44.860 --> 29:45.160]  Yes.
[29:52.410 --> 29:58.710]  You know, honestly, in the sake of, like, speed and prototyping, ease of just being able to, like, slam through stuff.
[29:58.890 --> 30:06.830]  And again, the one thing I forgot to mention is, both of those robots that you saw, the stunt double and the original, I had four weeks to build both of those.
[30:06.950 --> 30:13.430]  That was four weeks from being handed the design to having to be ready to fly out to Nashville to go filming.
[30:13.610 --> 30:23.250]  So again, and most of that, I only had one 3D printer at the time, most of that was just sitting and watching a giant chassis get, you know, a three-day-long print, hoping that nothing would go wrong.
[30:23.430 --> 30:32.650]  But yeah, a lot of it was just being able to very easily go through, and those FDM printers are pretty bulletproof when it comes to just being able to knock out, you know, plastic stuff like that.
[30:32.910 --> 30:33.270]  Yes?
[30:33.510 --> 30:34.970]  Did you get to keep one of them?
[30:35.330 --> 30:36.530]  I was very, very lucky.
[30:36.630 --> 30:44.110]  So from the beginning of the project, one of the things I was adamant about with the producers was, I do not want this robot to sit on a shelf after we've finished production.
[30:44.150 --> 30:47.590]  So I got to keep the original one, which is sometimes rare.
[30:48.210 --> 30:49.550]  But he lives with me in my workshop.
[30:49.710 --> 30:50.990]  They kept the stunt double one.
[30:51.310 --> 30:52.190]  And so it works out.
[30:52.310 --> 30:53.870]  Now, obviously, it gets a little bit complicated.
[30:54.170 --> 30:57.390]  And most of my builds, I like to release the files and, you know, make it open-source.
[30:57.530 --> 31:01.290]  But it gets tricky because I don't own the creative rights to that robot.
[31:01.510 --> 31:05.790]  You know, the writer, the director, the production company, they own the creative rights to that robot.
[31:06.010 --> 31:10.050]  And so for me, I'm left as essentially a guardian, if you will.
[31:10.990 --> 31:13.910]  But I, you know, not that I would want to merchandise anything.
[31:14.050 --> 31:22.250]  But the idea, again, being that the intellectual property is owned by another entity besides the person who actually did the CAD work and the build work and all that stuff like that.
[31:22.350 --> 31:23.430]  But yeah, very, very good question.
[31:23.850 --> 31:24.070]  Yeah.
[31:24.730 --> 31:25.530]  Yes, question back.
[31:35.920 --> 31:37.040]  Yeah, yeah, absolutely.
[31:37.220 --> 31:47.100]  And again, that's a really great question in terms of, for anyone who didn't hear, evil robot, if that character had been evil as opposed to being a good, cute little psychic robot.
[31:47.240 --> 31:48.440]  And the answer is absolutely.
[31:49.080 --> 32:04.360]  And there's a reason that, you know, I had the example earlier of these, you know, the Star Wars example, which is a classic of, you know, not only is that in shape and in color, you know, looking more sinister, but in action-wise, yes, a lot of that is thinking about,
[32:04.560 --> 32:06.700]  you know, sharp, quick movements.
[32:06.740 --> 32:07.680]  It's not slow.
[32:07.680 --> 32:10.600]  So, you know, he's, are you calculating in where you're going?
[32:10.720 --> 32:14.000]  Kind of like what I mentioned, meandering from place to place, this little robot.
[32:14.200 --> 32:23.300]  Those little things that, again, we project onto these inanimate objects and think, oh, well that's, you know, my toddler wanders when he goes places, or my dog kind of gets distracted by smells.
[32:23.420 --> 32:29.340]  Those things that we really latch onto, mimicking that and then doing the reverse of that.
[32:29.480 --> 32:31.660]  How do we remove that out of the equation?
[32:31.740 --> 32:33.200]  How do we make it more robotic?
[32:33.200 --> 32:41.060]  Because I guess the answer to your question is, is, you know, building that foundation so that you know that you can then remove those ingredients bit by bit by bit.
[32:41.720 --> 32:42.080]  Yeah.
[32:42.580 --> 32:43.200]  Other question?
[32:44.040 --> 32:44.400]  Yes.
[32:45.680 --> 32:47.120]  I came to the point, sorry.
[32:47.620 --> 32:49.940]  I'm just having a curiosity about what was the movie about?
[32:50.520 --> 32:51.120]  Oh, yes.
[32:51.360 --> 32:53.540]  So, I forget it because I've seen it, but yeah.
[32:54.300 --> 32:56.560]  It was essentially, I mean, again, it's a kid's movie.
[32:56.560 --> 32:57.660]  So, it's kind of that E.T.
[32:57.780 --> 33:06.260]  style story of, you know, you have the young heroes going up against the big bad, you know, corporation, government entity, all that sort of thing.
[33:06.380 --> 33:07.920]  And they have their plucky little sidekick.
[33:08.120 --> 33:14.420]  In this case, he didn't have any magical powers, but Newt playing a similar role where he's, you know, essentially like the number two character.
[33:14.820 --> 33:21.340]  And he's there to be the comedic relief, the shoulder to cry on, like those sort of things.
[33:21.580 --> 33:33.660]  But yeah, the character is essentially going through a, you know, post-apocalyptic, or not post-apocalyptic, but just a distant future where things don't look great and trying to solve the problems of the world and taking his ragtag team along with him,
[33:33.820 --> 33:36.160]  which is a storyline that has never been done before.
[33:36.320 --> 33:42.180]  But the point being, when you execute it in a new and interesting way, that's what's, you know, kind of the fun of it, I would say.
[33:42.700 --> 33:43.960]  But yeah, great question.
[33:44.860 --> 33:45.540]  Another question?
[33:49.410 --> 33:49.810]  Cool.
[33:50.630 --> 33:52.810]  Well, I don't want to take up too much of your time,
[33:56.010 --> 33:56.990]  but anything else?
[33:58.350 --> 33:59.410]  Otherwise, should I keep going?
[34:04.260 --> 34:05.400]  I do, yes, fortunately.
[34:05.540 --> 34:07.240]  And I actually just got a CNC machine as well.
[34:07.440 --> 34:11.980]  So now I can finally make a robot that if you drop them, it'll hurt your foot and not the other way around.
[34:12.120 --> 34:15.460]  So that's the goal is to hopefully kind of keep doing more builds like that.
[34:16.720 --> 34:18.620]  It's a very small niche, as you can imagine.
[34:18.760 --> 34:22.440]  There's not that many people that, you know, get called to do a build like this.
[34:23.380 --> 34:26.560]  So they're few and far between and we'll see.
[34:26.660 --> 34:32.240]  I think there's kind of been a bit of a resurgence around, you know, coming back, walking back from CGI in that sense.
[34:32.380 --> 34:36.640]  And I mean, you look at the most recent Star Wars movies, they're like, all right, what happens if we make BB-8 real?
[34:36.720 --> 34:38.720]  And what happens if we make some of these other characters real?
[34:38.860 --> 34:45.640]  And I think that, and I don't want to speak for anyone else, but I think that that had a tangible impact on the quality of regardless of what you think of the films themselves.
[34:45.640 --> 34:48.560]  I think the execution was the fun part.
[34:48.560 --> 34:53.680]  And I think that seeing that kind of get reinvented and put back into the mainstream was fun to see, I would argue.
[34:54.860 --> 34:55.220]  Yes.
[34:55.520 --> 34:56.960]  I was going to say, there's a rule
[35:01.490 --> 35:03.990]  if they could do it in practical class
[35:07.090 --> 35:07.590]  for short.
[35:07.710 --> 35:08.010]  Yeah.
[35:08.130 --> 35:10.850]  So like a lot of these kinds of things like they had a hair done.
[35:12.710 --> 35:13.070]  Yeah.
[35:14.450 --> 35:14.810]  How
[35:19.440 --> 35:20.220]  would that be used?
[35:20.360 --> 35:20.720]  Exactly.
[35:21.020 --> 35:28.340]  And like I said, that was such a big part of it because not only was this movie for kids, but it was a movie with child actors being a big part of it.
[35:28.340 --> 35:33.560]  And one of the days that we were filming, we had a kid who was only on, you know, set for maybe two days.
[35:33.720 --> 35:35.440]  And so he was a little new to the situation.
[35:35.760 --> 35:38.860]  And he, you know, kind of tugs on my shirt and he goes, is he real?
[35:39.260 --> 35:44.780]  I was like, yeah, little man, that robot is very, I mean, to you, like for all intents and purposes, like, yeah, that's a real robot.
[35:45.440 --> 35:46.480]  But it was the same idea.
[35:46.480 --> 35:49.720]  And there were some elements like the projector in his head.
[35:49.760 --> 35:52.260]  It's a holographic projector in the film.
[35:52.380 --> 35:55.260]  Obviously, you know, that technology doesn't exist for the purpose of what we're doing.
[35:55.260 --> 36:02.240]  But my goal there was to basically just take a regular projector, gut it, and then keep the Fresnel lens in the original light.
[36:02.560 --> 36:11.420]  And so that way you can get your nice cone of light so that the VFX team can go in and, you know, pretty easily rotoscope that because you have the shape and the, you know, color temperature correct.
[36:12.040 --> 36:15.260]  But you're just overlaying whatever graphics would be on top of that in that sense.
[36:15.420 --> 36:19.460]  So yeah, using CGI kind of as an assistive tool as opposed to a replacement.
[36:20.160 --> 36:24.460]  And again, there's places where it is, you know, cheaper and better to do full replacement.
[36:24.660 --> 36:36.140]  But in an instance like this, whenever we have to go back and, you know, if we have to shoot pickups, it's a lot cheaper to use the robot you already paid for as a production than it is to re-render an entire scene, bring the animation team back in.
[36:36.280 --> 36:41.000]  So there's, you know, some cost benefit as well to that, I would say.
[36:41.660 --> 36:42.060]  Yes.
[37:03.770 --> 37:04.690]  That's a great question.
[37:04.830 --> 37:04.890]  Yeah.
[37:05.010 --> 37:11.290]  Question being for tackling impossible, seemingly impossible projects and the aha moments that go with them.
[37:11.470 --> 37:20.870]  And I would say, I was having a conversation with Ben Eadie who built the thing, the hand, you know, in the most recent episode, Wednesday, I guess is the show.
[37:21.430 --> 37:25.410]  And he was just talking about how, what is the point in which you...
[37:25.410 --> 37:28.710]  The question was, he was asked, what was the point in which you abandon a project?
[37:28.710 --> 37:34.190]  And what we were talking about is essentially, if you've hit that point, what it really means is it's time to pivot.
[37:35.050 --> 37:43.170]  And it's, you know, if you're giving up a project, it's really just that hitting, you know, the first bumps in the road, those first failures that you have to get out of the way.
[37:43.350 --> 37:48.690]  And we argued that it's, you know, pushing all the way up until you can't anymore and then pivoting.
[37:48.870 --> 38:01.810]  And so a lot of that is also to, you know, the ethos, the producer and director demands that, you know, them seem very real, but as an engineer, we have to be like, okay, well, that is actually impossible.
[38:01.950 --> 38:04.210]  And so, yes, that's the other side of it, I would say as well.
[38:04.370 --> 38:14.190]  And so that's kind of where the bad cop comes in, where I have to be like, okay, knowing what we know about physics or anything like that, these are the options that we have.
[38:15.070 --> 38:20.810]  But always structuring it in the way of, you know, how do we continue to serve the purpose of the story, as opposed to like, no, we can't do that.
[38:20.970 --> 38:21.430]  You know what I mean?
[38:21.430 --> 38:28.390]  It's okay, let's find a new way that the story can evolve, a new way that the script can work with the limitations that we have.
[38:28.870 --> 38:31.290]  And fortunately, I would say filmmakers are very used to that.
[38:32.030 --> 38:36.490]  You know, most of us come from the short film world where there's not a lot of money to do the thing you want to do.
[38:36.630 --> 38:41.530]  And sometimes you end up with a really cool end result unintentionally, I would say, in many cases.
[38:41.530 --> 38:57.570]  But for the aha moments, it's usually... and again, I'm very lucky to have come from an interesting background of doing all these weird problem-solvings, but it is usually looking at it from a perspective, and I know everybody says this, but looking at it from a perspective that we wouldn't have
[38:57.570 --> 38:58.250]  approached otherwise.
[38:58.550 --> 39:04.070]  And to go back to MythBusters, that was what was so great about that team, is that none of our skills really overlapped.
[39:04.410 --> 39:05.370]  And that was perfect.
[39:05.730 --> 39:08.970]  Because none of us were trying to solve the problem in the exact same way.
[39:08.970 --> 39:16.590]  So when we all were working together, now you had a completely new angle to every single problem, just by the people that were sitting around the problem.
[39:17.070 --> 39:19.890]  So you didn't have to do all the legwork in that sense as well.
[39:20.310 --> 39:22.990]  So yeah, I would say teamwork is another big part of that.
[39:23.830 --> 39:24.270]  Yes?
[39:24.790 --> 39:26.950]  Is Backhaul still in your studio?
[39:27.070 --> 39:27.630]  Yes, yeah.
[39:27.790 --> 39:30.770]  All of the special effects stuff that I do, yeah, is all through Backhaul.
[39:31.530 --> 39:32.590]  But, yeah.
[39:34.510 --> 39:35.450]  Any other questions?
[39:37.090 --> 39:37.530]  Yeah?
[39:45.530 --> 39:49.610]  It's, as my friend once put it, a bigger budget just affords you bigger problems.
[39:50.910 --> 39:52.890]  And not in the, like, more money, more problems way.
[39:52.990 --> 39:56.770]  But more of like, okay, if you have 100 people, now you need bigger tents.
[39:56.950 --> 39:57.670]  Now you need bigger trucks.
[39:57.810 --> 39:59.730]  Like, those are different scale problems.
[39:59.950 --> 40:02.550]  They're not necessarily, you know, bigger or worse.
[40:02.710 --> 40:04.130]  But they're just different problems.
[40:04.730 --> 40:11.130]  And so I would say there's kind of a fortunate side of working with a lower budget, which is the expectation has been set.
[40:11.390 --> 40:15.930]  So if you can over-deliver for that budget range, then you're doing a really good job.
[40:16.030 --> 40:17.750]  Whereas if the budget's really high, it's like, well, great.
[40:17.830 --> 40:22.450]  Now you're competing with, you know, ILM, you know, the Star Wars and the industrial light and magics of the world.
[40:22.450 --> 40:25.210]  And that's a very different bar that you've set for yourself.
[40:25.430 --> 40:37.710]  But yeah, overall, I would say that there's that sweet spot of knowing that we can push the envelope, we can get to the point where we're going to give you enough resources to get the job done, but we're hiring you because we know you can go further than that.
[40:37.850 --> 40:43.630]  We know that you're the one who can bring us a little bit extra, you know, again, are you the chef or are you the cook?
[40:43.690 --> 40:51.190]  Are you going to look at what's already been done or are you going to take those ingredients and say, okay, what if we rearrange this in a way that is, you know, goes beyond what's already been accomplished?
[40:54.830 --> 40:55.670]  Any other questions?
[41:02.690 --> 41:03.930]  Thank you very much.
[41:04.270 --> 41:07.690]  I finished a little early, so I appreciate you for the extra questions.
[41:08.410 --> 41:09.670]  But thank you so much.