Corridor Digital is democratizing high-end VFX by prioritizing open-source accessibility over proprietary gatekeeping. This project proves that community-driven AI can disrupt industry monopolies more effectively than traditional corporate software.
Installez notre extension pour rechercher instantanément dans n'importe quelle vidéo
We Made a SPLASH at NAB Vegas 2026 and Tested Out VFX Artists React as a LIVE Show | EP#246Ajouté :
What's up everybody?
>> Hello. Hello. Welcome back to the corridor cast.
>> There we are. That's our >> Nico and I are joined by Austin today.
Austin Reeves. Welcome to the couch.
>> Nico and you are joined by Austin.
>> What about me and Nico are joined by Ren.
>> Austin joined by Ren today on the couch.
>> And Ren and I have been joined by Nico.
>> I think that covers all the >> and also by Faceless Me sitting just off screen.
>> Yeah, you got to bring him in again.
This is This is how we get demonetized immediately. Demonized.
>> Demonized. I mean, that's the demon >> and he's gone.
>> It's how you demonize.
>> Demonitized.
>> He's in and out.
>> Uh, so yeah, Ren just informed me that my quarter key presentation from NAB, which was last week, by the way, uh, has 250,000 views, which is a little frightening because I prepared for that as one would if they're going to be doing a presentation to a room full of 100 people informally, not to a quarter million people worldwide.
So some context here. Last week was a trade show called NAB for the National Association of Broadcasting. It's a big sort of film making, media, hardware, software type convention. Takes place in Vegas.
>> Uh it's huge. I think there's something like 100,000 people go to this thing.
It's massive. Um >> yeah, every camera company, software, broadcasting companies, all there.
>> Yeah. Anybody who's Anybody. and uh Pu Systems who makes all of our computers uh were uh we've been partners with them for like eight years maybe more and >> they had us basically give some talks. I get I gave a talk about my true scale videos called like true scale of true scale. Um, we did a live VFXRS React episode.
>> That was fun.
>> And we did a live debunking episode, but also Nico dove deep into talking about corridor key for an entire hour.
>> And they uploaded all that.
>> And by they, I mean Puget, it was like a a crossost between three different YouTube channels.
>> Puget, I forget the other one, and Nvidia.
>> So, the Nvidia YouTube channel. Yeah, it's at 300,000 views. Oh, man.
it more.
>> We're the third YouTube channel.
>> Oh, it's us. Okay.
>> Whoever those guys are, >> he's just someone in Nvidia. Oh, it's >> actually all of the talks are on replay on Pug channel if you missed anything and you want to watch it.
>> Including Rens.
>> Yeah, Ren is up there.
>> Oh, nice.
>> What's his view at?
>> But no, for real, like uh Corridor Key was the talk of the town while we're at NAB.
>> Yeah, it was really cool. People kept like congratulating me and telling me how good of a job I did and I'm like thank you.
>> It was all me.
>> Hey, that one key you pulled later on was really good.
>> Yeah. All of them. Boom. Even the followup video.
>> Even me and Nick had to step in and do >> Oh, my videos got 2,000 views.
>> Me and Nick had to Me and Nick had to step in for Jon >> at a modest 200 >> 33K for Bx Artist React. And then uh >> Wait, where's the quarter key one?
>> Nvidia. Wait, did that get uploaded on Nvidia's channel?
>> Yeah, Nvidia took that. That was >> Yeah. Uh, had I known, um, I would have prepped more.
>> Wait, I think it was an an Nvidia adjacent channel. It was like Nvidia something.
>> Oh, video studio.
>> Uh, yeah, that's crazy. Um, >> they only have 110. Oh, they only have one.
>> And now it's their most popular video of all time.
>> Okay, that's cool. That's close. No, no, not quite.
>> Yeah. No, I was I was just joking.
>> Right there, though. It's in the top three, four rows.
>> Top three or four rows. It's >> truly I was like on the airplane from North Carolina back to Vegas. I was like, I will I think I'll talk about this and I can talk about this for like 10 minutes and I can talk about this for a while and I got some pictures for this and then I'll talk about this. Okay, that's my >> Well, dude, that screen tripped me out for a moment. I thought you were in the studio, >> dude. It was cool. It was cool to see it like that.
>> But yeah, this is this is what I was talking about with the Mo. So, there's an LED wall that's 20 feet wide and 10 feet tall behind the couch there behind Nico. And they actually got a scan of our studio Gaussian splat. They're actually uh projecting on the screen in real time. So, it's not just a photo. We can actually navigate around it, but you can kind of see a little bit of like cross-hatching on the screen. And that's a that's a phenomenon called Mo.
>> That's M O I R E. I think it's a French word. Uh, but it's if you've ever taken a phone out or like a camera and aimed it at your computer screen or something like that and you start seeing these weird patterns, it's because your camera is able to actually pick up the individual pixels and when you get enough of them kind of like next to each other, they start kind of cascading into this weird pattern.
>> Mhm.
>> You get around it by getting a higher density screen, a higher resolution screen. So, >> or by defocusing your >> or by defocusing a little bit.
>> How did it look in person?
>> It looked great in person.
>> It looked great in person. Yeah, >> NB was crazy though. Like it's really cool meeting so many cool filmmakers and like engineers and you know all the people that work on the stuff. Um but yeah, you're right. Like everybody was mentioning corridor key when they'd come up to us and >> I once again I had no expectation that this is what would happen when I made that video. Um and it's really cool but to be fair a lot of people come up like quarter key is super cool. Like oh that's really awesome. Have you tried it? Like no not yet.
>> The power of storytelling. Everyone loves the idea.
>> Even if you can try it because it's easy.
>> Yeah. Um but thankfully that momentum has rolled into a lot of people joining the Discord and we've had a couple like actual machine learning students, graduates, professionals >> uh kind of jump in now and start like being like, "Oh, this is your model, huh? This is how you trained it, huh?
This is your data set. Let me see what I can do.
And uh you know it's it's going and people are working on it and they're kind of free time here and the stuff that they are putting together is crazy.
>> Uh cuz basically imagine corridor key with the intelligent you know green screen unmixing thing going on. But also imagine running at native 4K where it's truly pixel perfect. Imagine it with a additional AI that reduces flicker >> so you don't get any chatter in the edges even if you have like chroma noise and things like that. Um, and that's, you know, that's the new version. So, you know, fully 4K native flicker-free.
>> When you say that's the new version, what do you mean? Like, that's corridor key 2.
>> Well, not the new version. Sorry. That's like that's what's people are working on and trying to, you know, evolve the model into. I don't know if it'll be Corridor Key 2 or what. Like, you know, it's all >> It's not It's not like out yet.
>> It's not out yet. It's 100% experimental.
>> Has plenty of problems >> beyond that stuff too, but like I am very very hopeful. See, like the the grandmaster plan is that I just don't want to have to do green screens anymore.
>> Um, and at this rate, if if people like stay passionate about it and I can like keep fueling that passion from our end somehow, like truly green screens might not be a thing to deal with by the end of the year in the sense that like, you know, you'll still shoot with one, but like the software to use it will be, you know, instant and easy.
>> We'll see. We'll see. Um, but there's some very very smart people that have like volunteered some of their time. Um, and it is really really cool to see and I'm learning a ton in the process. Like I need to sit down like every time they leave a comment I need to like pull it into like Claude and be like, "Can you please help explain this to me?"
>> Yeah, >> I need some help understanding this comment.
>> Yeah, 100%. It truly takes me like half an hour to read through like >> a few pages of comments.
>> Dang.
>> Um, and then by the time I'm finished, I need to do something else. When I come back, I have another half hour of reading before I can even leave a comment.
>> My god. Now, we've got a blue screen version coming out soon. Is that going to be rolled out into the same corridor key one or is it straight up just quarter key?
>> So, it's it's the exact same model architecture just trained on different data for blue screens and I'll probably upload that today.
>> But is it is it a separate model?
>> Yes.
>> So, >> so you just swap you just swap the models. Um, >> is there a way to just merge both of the models together so there is no swapping?
No, I mean you can put them both in the folder. Maybe like a a you know a drop down menu, but >> yeah.
>> Well, select whether or not you're doing green or blue.
>> No, you can't. Right now, you can't mush them together because like the model's either trained to remove green or it's trained to remove blue. And we could train it to remove green and blue.
>> But the whole point is like you get to focus a little bit as a filmmaker.
>> Yeah, that makes sense.
>> Yeah, that makes sense.
>> Yeah, Ron, it's cool to hear you're making your own uh tools. I saw your comment there. Um but yeah, NB was nuts.
It was uh we did a live react episode and we've kind of talked about doing a live thing here in LA at some point.
>> Mhm.
>> And this is kind of like a soft attempt at that.
>> Soft attempt. Yeah.
>> Um it was actually really fun.
>> What did you learn >> from doing it?
>> You can't geek out too hard on VFXRs React cuz you need to like listen to like the laughs per minute and like once it's been like one to two minutes and you're like going in like Yeah. So I think with that they did like rotocoping on this like but you haven't heard any laughs in about two minutes like uh oh >> like got to got to move on. Got to drop it. editing. Yeah, got to play more clips.
>> So, truly, you are self-editing as opposed to like trying to be specific and geek out. And you actually go through clips a lot faster.
>> Yeah, it does seem to be faster paced.
>> I brought that freaking bulletproof monk clip and we looked at it for like 3 minutes and you're like, "All right, what's next?" I was like, "Okay."
>> Um, yeah. No, it's interesting though because I did get some feedback from people saying that they really liked uh when we would kind of go into detail.
So, like the moments you're talking about where we weren't laughing, like they liked that part. M but we're also like guessing and I feel like if we did a live show, we'd have to like be dialed in with like our >> guesses. Yeah, cuz there's definitely like a vibe difference of like everyone just doing their conjecture, but it's like once you you got to hit that moment of like here's how it really happens, right? Exactly. That's when you're like, "Oh, then you got to like somehow preserve the natural reactions and stuff."
>> Yeah. Well, it all seems like, you know, well, once again, we like we didn't we didn't prep for this like we were prepping for our first ever actual live show. This is >> we prepped for this like how we would with >> film a session. Yeah.
>> Yeah. And it's like the person who brings the clip has a resource be the other two people get the natty react which is how we typically do it.
>> Yeah.
>> Also they have a green sofo over here.
>> How's that possible?
>> Facebook marketplace.
>> Yeah.
>> They just found it. They happened to find it. They found it in Vegas for us and they did the whole thing.
>> Is it Is it the exact same one?
>> Nope.
>> Okay.
>> Nope.
>> It's quite different but it looks very similar. At first glance you're like, "Oh, it's the same couch." And then you really get to look and you're like actually it's very different.
>> But you know green leather.
>> We also did a live debunk episode in AI shopping scams.
>> That was really fun. I think that worked really well as a live thing.
>> Yeah, 100%.
>> Cuz you're like you're hitting stuff way faster. Everybody's laughing at it with you.
>> Yeah. If we got some gnarly products next time, it could really work as a >> Wait, how many views is this one at?
>> Only 36,000.
See, that was the amount of views that I was kind of expecting out of all these.
I was like, "All right, it'll probably get somewhere between like 1 and 10,000 views." And then I saw your video pop up in my feed the other night at 250,000. I was like, "Oh, okay.
>> Well, Nvidia's happy, I guess."
>> Yeah, that's cool. It's always nice to have Nvidia be happy.
>> Yeah, >> the GPUs must flow.
Um, yeah, but the something about like showing an AI shopping ad and then pulling the box out on stage and opening with everybody like >> it's a really good moment.
>> Yeah, exactly.
>> Emotional support biggie.
>> Part of me is a little bummed we burned two products on the live presentation that we're not going to be able to put in a future episode. But the other part of me is like it was worth it.
>> I think that makes good products to burn too. Yeah, >> like they weren't the most outrageous.
Like, dang, that would have been >> The problem was that they were actually pretty close to the the picture.
>> Yeah, they're actually like pretty accurate to the >> were a little accurate. So, I think it was fine to do for live and, you know, sacrifice a video for it. I guess >> you got more coming.
>> Like if the other like live another addition of doing this live is also like getting a close-up cam on it so for everyone to see like I guess this one has a close-up cam here too. Yeah. Yeah, we uh we should object to everybody that's watching.
>> Yeah, we shot it just like React. So, Oh, no. No, that isn't that isn't for the crowd, though. The crowd say the crowd >> the crowd does not see that close-up shot >> that's so small. Like someone in the back is going to have a hard time seeing that.
>> Yeah.
>> But that also makes it good like if you're doing if you're doing scam like shopping scams live.
>> Mhm.
>> I think it to audience at the end.
That's what we did. That's so fun. Yeah.
At the end of it, we uh gave these products away.
>> And I think in a you know, in a planned live show, we would have this feed on the screen. probably the two camera feeds.
>> So, the live show that we've been kicking around for literally like 2 or 3 years now is the idea of like doing some sort of like 2 or three hour show here in LA somewhere. Actually rent out an entire like theater. Um >> kind of as like a proto attempt to like see what touring could be like. Um >> I don't know if we could actually do a touring show because that would take a lot of time, but it'd be cool to do a single >> tour either because we're in LA. You know, we could probably fill up a theater here once in a while and just have that thing. But the idea is that we'd have multiple different things we do. One of them would probably be like a live react show. I'd probably give some sort of talk. Maybe someone else would give a talk as well. Just a quick thing.
I like the idea of doing an improv show wearing mocap suits connected into Unreal Engine. And so you get suggestions from the audience of like what characters you are, what location you're at, and then Sam just really quickly like links up those characters and the location, loads it up, and then we're seeing it on the screen, and we just the two people just act out the stuff.
>> I mean, it's like a corridor variety show.
>> Yeah. Yeah. Exactly. It'd be a variety show.
>> Yeah.
>> Um >> a little live like a live render challenge.
>> How many of y'all would come to that >> sort of thing, you know?
>> Yeah.
>> I've also heard that like shows like this are outside of like seeing the the people, it's also people go to these kind of shows because they get to meet other people that are into the same thing they are.
>> Like that's a big aspect of that. So, you know, we could try to also like tailor some of the show to that where it's like not necessarily all about come and watch us for like two and a half hours or whatever, but like >> you before and after event too that >> like make the community come together a bit.
>> Yeah.
>> Yeah.
>> Like-minded people. A >> little cocktail.
>> Oh, yeah. We'd have to do some sort of trick shot competition. That'd be fun.
>> Oh, yeah.
>> But then we're starting to get into the dude perfect territory cuz they actually do.
>> They actually do that.
>> I got to ask them recently about how they do their trick shots.
>> Mhm. I was like, "Is it, you know, are you guys just really good? Is it editing?"
>> Actually, they said, "It's just," they don't feel like they're really good.
They said one, one of their guys is really good. They're like, "Takes forever."
>> But, uh, they're like, >> "We're not particularly good.
>> It's just we got friends who are there to throw the ball back to you until you make it."
>> Yeah.
>> They're not good, they're just patient, >> you know. So, we've been friends with the the behind thes scenes crew to Dude Perfect for I want to say >> 10 years now.
>> We saw him in NAB.
>> Yeah. Yeah, I know. Uh and we've seen them at NAB multiple times.
>> Yeah.
>> Um >> but yeah, they're very behind the scenes. They don't like being in front of the camera, but yeah, so it's it's very interesting kind of hearing about that whole empire and everything that they do and and the just the smooth system that they have going. Blows me away. They have an entire like I don't Well, first off, >> Dude Perfect is such a huge channel.
They they make bank. So they're able to buy and build warehouses and and I I also Where are they based?
>> Uh Dallas >> outside of Dallas to buy warehouses.
>> Yeah. Their own area outside of Dallas.
>> Yeah.
>> Yeah. It it does help for that. That's >> way cheaper to do live shows not in LA.
>> Well, no. They tour they go around the country.
>> They tour it. So >> they spent a couple months just like and they've done that I want to say like two or three times already.
>> Yeah.
>> But yeah, >> it's probably a big show. But the thing is is that all their trick shots are real.
>> Yeah.
>> I do have I have a crazy story from NAB.
Yeah.
>> That uh so I was sitting with my cousin so I could a lift from the show back to my cousin's house and guy picks me up >> and he's like you you know we start talking. He's like so you know what's the what's the event going on? I'm like oh it's it's NAB. Oh your microphone just fell apart.
>> Oh my god.
>> Are you kidding me?
>> The stand literally just broke.
>> You weren't even touching it. This thing came out like broke broke or is >> Yeah, it's broken. The plastic is bent and broken piece off.
>> You're ruining my story.
>> I'll just try not to really again.
>> Oh, what the freaking >> It's fine. It's fine. We'll worry about it later.
>> Just Ren can't talk for the rest of the episode. Maybe try not to wrench on it next time.
>> He was freaking wrenching on it this whole time.
>> Okay. Well, it's okay.
>> We We'll fix it.
Um, so he, "Oh, it's NAB." That's cool.
And I was like, "You know what? What brought you to Vegas?" And he's like this older guy, um, probably retired or something like that. He's like, "Well, >> I actually uh used to be with EA Sports.
Um, I opened up all their Asian studios like EA Japan, EA Beijing, EA Australia." Oh, that's really cool.
Like, in what capacity? Like a engineer or like Oh, no. CEO.
>> And this is your lift. This is my lift driver.
>> What? Um, and this is uh 1995 and earlier that he was with EA.
>> Well, EA Sports is in the game.
>> EA Sports is in the game. That's uh that's him.
>> He's out of the game now.
>> He's out of the game now. Uh, yeah. He said he retired in '95. So, um, >> but I was like, "Oh, okay." And tell me more. So, he starts getting into it and like telling me all these different stories of like the things they worked on. I was like, "Well, is this, you know, is there any like game that you're particularly proud of?" He's like, "Well, actually, yeah. So, this isn't this wasn't for for EA." So, I had a a friend. He was uh they're trying to, you know, they're European. They're trying to sell games in the US at this point.
Um his name was like Eve Gilmo, which is the Ubisoft owner.
>> And he's like, you know, uh I'll I'll leave this guy I'll leave this guy anonymous. Uh who is my driver. Uh but he was like, you know, he came to he's like, hey, you know, the problem is like our games aren't selling United States.
Like, what can you have any advice?
Like, yeah, you know, the thing is you guys are making games for like teenagers, like 13-year-olds, 15 year olds, etc. And the United States is an older gaming audience. And what you guys need is you guys need like sports and you need military realism. That's what's selling right now. He's like, "Okay, military realism." He's like, "You know, >> I actually have an idea.
>> There's an author I know >> who, you know, people don't people really know him, so you can probably get him for cheap. You guys can go sign up with him, take his stuff. I think it'd be a perfect pairing for you guys to then launch like a line of video games."
So he proceeds to go to Boston and negotiate with Tom CL. I was about to say some guy no one knows. Are you kidding me? Tom Clancy has been famous for decades.
>> He's been famous for decades.
>> He's been famous after he died.
>> Some dumbass like, you know, he's like, you know, he's got a niche audience, but like you could probably pull something off. He's not like a full household name yet. Even though he made, you know, all those movies about his books, too. But was that >> pre was that when did those movies come out? Was that also in the >> 80s?
>> Okay. Okay. Well, either way, um, so he goes to Boston and helps negotiate the deal with Tom Clansancy. And then they make Rainbow Six.
>> Wow. Wow.
>> And he's like that. He's like, "Rainbow Six and Splinter Cell." He's like, "Those are the two I'm really proud of, >> dude. Splinter Cell. Oh my god."
>> Yeah. So, >> dang.
>> Yeah. He told me all these other stories. And by the end, he's like, "I've been talking this whole time. I usually never tell people about this stuff." Uh, he's like, "How do you really manage to get that out of like, "Well, actually, I interview people professionally for my for my living."
Oh, okay.
>> Yeah. And that was that was crazy. Um, >> yeah. What a wild driver.
>> Yeah. You know, he's like retired and he's been retired for like, you know, >> but you said he retired in ' 95. That was before Splinter Cell came out.
>> Yeah. Yeah.
>> But I guess he just set the groundwork for it, >> right?
>> Dang, man. Dang, man.
>> Yep.
>> That's wild. David M just said, "I love seeing the full range of facial hair on this week's podcast."
>> We go from from Ren to a full Nico.
>> Austin and I will do the fusion dance and we'll turn on Nico.
>> We get way blonder. We fuse for some reason.
>> So, I don't know if you guys saw the react episode with Stephen Spielberg's movie Always that we covered the forest fire plane flying uh video.
>> Yeah. Um, but Sam had asked a question in it. How do they hide the wires for all these model airplanes?
>> And we speculated and postulated and whatnot. Um, but I actually went to an event last night, the uh the technical and scientific achievement awards at uh the Oscars that the Academy gives out.
Um, and I met the producer who happens to also be an old family friend of Sam's.
>> How does that keep happening? What the heck?
>> Cuz Sam's dad was apparently a baller back in the day and knew the community out in San Francisco. Um >> Sam Sam's the secret Nepo baby.
>> Yeah, Sam's the secret NEPO baby.
>> Secret ultimate Nepo baby.
>> Um >> without cashing in.
>> Without ever cashing in. Yeah.
>> But uh he told me that here's how you hide the wires. You put them on a little spinning disc. So the wires are motion blurred >> when you film.
>> Yeah.
>> But it's not spinning enough to make the thing it's holding up move.
>> Right. Exactly. It's spinning at the top. But the point down is >> Yeah. It's just like a little a little spin like that. Whoa, that's a really >> And apparently that was a technique they used back in like the 1930s.
>> That's a really cool.
>> So they use it for this movie.
>> So So that's for if we do a VFX artist react redo re follow-up episode.
>> Y >> answer that specific question.
>> Yeah, >> that's really cool.
>> So the like they're just on like little mini motors.
>> Mhm.
>> Like that dog from the AI shopping scam.
>> Exactly. the same mini motors that are in the toy dog.
>> Also, this plane shot here. So, this is all miniatures um next to the ground.
And what they did here is the plane in the foreground is a bigature and the plane in the background is a smaller scale airplane. So, they're actually right next to each other.
>> Oh, wow.
>> And it's just a force perspective trick with different size >> miniatures. Whoa. So, is that a fake guy in that cockpit there in the front?
>> Yep.
>> How big is that bigature then? I think it's a 1/16th in the in the foreground and like 127th or something like that in the background. Um exactly how the clouds, how are they going through the clouds?
>> So the clouds, they got two Navy fogers from World War II that they would use for naval battles and it uses uh like mineral oils.
>> Oh, whoa.
>> And that's what they're using to create the clouds here.
>> Is that to like fog up the the water so that the boats can't see each other?
>> Yeah.
>> Huh.
>> Whoa.
Huh.
>> We need to do a follow-up episode pronto.
>> This is crazy stuff.
>> Yeah. So, that was really cool to like just get all these facts dropped on me last night. Um, >> was So, was this an actual like awards show? People are going up on stage and receiving awards. How long was this show?
>> About hour and a half to two hours.
>> Oh, that's pretty quick for an award show.
>> Yeah. And it's it's pretty much like the technical people. So, the people that won, uh, there's three different pyrochnic companies that figured out how to make squibs that don't have lead in them.
>> Oh.
>> Uh, cuz squibs up to this point would have lead for the explosive. And of course, you can't really breathe and it's getting that lead dust everywhere.
Um, so they all kind of three simultaneously invented non-LED based squibs, which is cool. So, they all got an award. Uh the guy that made DX Revival, that um audio processing tool, got an award because truly you can give it like the garbage, the most garbage of audio and he makes it sound good. Um and then Paul Debec uh won an award for his HDR image based lighting invention >> from like 30 years ago.
>> Yeah, it took him a little while.
>> What?
>> Oh my god.
>> HDR.
>> And he was he was the guy who invited you to go, right?
>> Yes. That's why I was there cuz he invited me out. Yeah.
>> He's like, "Hey, Nico, you want to come see me win an award?"
>> I was like, "Yeah, let's go.
>> Dude, Paul's awesome."
>> Yeah. Um, but you know, it's like it's all, you know, industry effectsy producer people and like once again, half of them we know cuz we've had them on the couch.
>> So, so it was really cool. Um, you Brian Brian Grill was there.
>> Yeah. Yeah. Yeah.
>> Yeah. We did just talk about I actually told him last night we were just talking about you in today's episode.
>> Um, so that was cool.
>> Yeah. When I'm 60, I'll get an award for Corder Key. Just >> another 20 years, you'll be fine.
>> Hopefully sooner. Um, I also spend a good chunk of the evening talking to uh Randall Klyner. I can't I think it's Klyner. Um, I'm bad at remembering last names, but he's the the director that stopped over and visited. He's a friend of Sam Wicker. He directed Honey Kid.
>> Yes. Yes, that's right. I remember him.
I remember being like surprised at how sharp and intelligent the dude was from being 70some. Yeah, >> it was like he was up to date on like current like tools and whatnot and he was like, "Oh, yeah. I I fired up Comfy and did some stuff." I'm like, "Excuse me, what >> Yeah, >> you've been on Medicaid for how many years?"
>> Or Medicare. Which one's >> um Let's see. I think he also directed Greece, >> I believe. So, yeah.
Yeah, because it was like he did uh Honey, I blew up the kids and >> that was a whole thing and he came by because he had seen our episode and he's friend he >> Sam uh Wicker like lived at his house.
>> Yeah.
>> And like some spare trailer on his property, >> which is crazy. Yeah. I mean, >> it's cool. Like he's so like chill. It's like you are a historic director. Like you made Greece, you know? Um, and he was there with his brother who's a visual effects uh, artist as well and producer and lead and whatnot. But yeah, it was super cool. It was like it's really cool to now go to these events, whether it's the VES awards, the scientific and technical awards.
>> Yeah.
>> And just like have people there who are like friends who we know >> as opposed to like being like, "Oh my gosh." It's like, "Oh, hey, what's up?
How's it going?" And like, you know, it just you can slip into conversation now.
It's it's great. I love it. I feel like colleagues now, even though they're all way more accomplished than we are.
>> Yeah.
>> No, we're just young.
>> Young at 40. Um, yeah. Then the other winners were the uh uh Sony Image Works um drawing system that they use for like Spiderverse and all that kind of stuff >> for like doing all the like 2D elements on top of stuff.
>> Yeah, 2D elements and like paint effects and all that kind of stuff. Uh they definitely deserved it. Um, and then there's a bunch of like audio syncing and reconforming tools that have been coming out because, you know, as people bop from one program to the next and the audio guys is working on an edit, but then the edit the director wants to change the edit, etc. Um, people have basically made intelligent systems now that are a mix of like waveform scanning and time code scanning and image matching >> that will just figure out how the edit changed and propagate it for everybody.
Like this went here and this went here and this is what's different here and >> which is also crazy and super cool and complicated. Uh and I mean cool and then all the uh multi-layered material stuff that uh Weda and ILM have built. Um they got awards as well. Those teams >> multi-layer materials.
>> Yeah. Basically having physically accurate 3D shaders that can be layered and mixed as needed.
>> Um in particular like if you think about car paint or like a wooden table, you have your wood which also technically has little like rough pores in it and then you have your lacquer on top of that and that lacquer has scratches and it also has finger grease and oil. So it's like you're actually dealing with like four different shaders.
>> Yeah.
>> For you know like that surface and like all surfaces are like that. It's not just one shader. It's multiple >> materials. And so both uh I think it's called Manuka rent like shaders or whatever for Weda for example. It's just a hyper complicated shader system that lets you simulate all of these things.
Yeah. Dang. All right. So they got awards as well. Um I think that's all of them. I could be missing some. I apologize if I did, but >> there were a lot of people on stage.
>> Yeah, it was cool. So, >> was it a fun time?
>> It was a fun time and it was like a small group, you know. It wasn't actually that crazy or that big. Okay.
It's like a very focused >> event. Um, you know, two 300 people or so.
>> His story.
>> He doesn't have it on his >> He had He had a bunch of stuff on his story today, but I don't know. Not letting me pull it up.
>> You got to log in, I guess.
>> Um, I'm not sure I follow him on Instagram.
>> That was super cool. I got to fix that.
>> Yeah, frame store was there, too. Yeah, they won an award for something. I can't remember what it was.
>> One of their systems.
>> Might have been that sinking system. I don't know.
>> Yep. Oh, and then uh Dragon Frame. The guys that make Dragon Frame. Uh also won an award >> for Dragon Frame.
>> For Dragon Frame.
>> Dragon Frame is a great stopotion tool where you can use onion skinning to look at the previous frame >> and picture you take when you're doing we used it on a stopotion video a while back. Yeah, too. I was thinking >> they're fans of the channel that watch our stuff.
>> Yeah. And Dragon Frame also will synchronize with motion control systems.
>> Yeah. Okay.
>> Which is cool. So when you advance the frame, everything advances. And if like, oh, I screwed it up. I need to go back 10 frames. Everything can rewind 10 frames as well. So it's really neat.
>> Yep. Yeah. So dragon frame, uh, matchbox. Yeah.
>> Matchbox 20.
Rob Thomas >> back. Did you guys see any good any notable tech at NAB that you're excited about?
>> I mean, it's it's always like I was less jaded about it this year than I was last year. Like last year someone asked me Yeah, a little bit. Someone asked me last year I was like, "What's your what's your favorite thing you saw at the show today?" And I'm like, >> "The door."
>> I honestly like nothing.
>> The Uber driver.
>> Because I don't know. Everyone's kind of got the same thing to some degree. You can see all these booths and they're all different companies kind of making the same thing in different areas, right?
>> Uh but it's like LED walls are cool.
>> Uh robot arms are cool, cameras are cool, and they're they're just newer this year. A little bit better.
>> Autograph Max on dropped autograph for free. I was pretty psyched about that. I was talking to somebody at the Maxon booth.
>> Yeah.
>> Sick. Yeah. Autograph is like After Effects, so now you can do compositing for free if you don't want to use Fusion.
>> Yep. Uh, I didn't realize Maxon had them though. Like I >> bought them and then uh ended up sitting on it and people are like, "What are you doing?"
>> Yeah.
>> You bought them and you're going to kill it. That's not good. And they're like, "Okay, we'll put it out."
>> I did a talk at Maxon. That was fun.
>> I I basically did a breakdown of my satisfying render, which if you guys haven't seen already, go check it out.
>> Video on YouTube right now.
>> Uh, and that was fun. I was so confident. This is It goes back time and time again. The moment you expect views is the moment you don't get them. And I know this and I'm like trying not to get excited about my Leaf render because I was pretty confident and I was like I think this is going to do pretty well.
>> It did not do well.
>> It's not repost on Instagram's official page.
>> No, no. I just mean like >> if you're watching this, give it give it a watch. Give it a like. You're >> just talking about the Instagram post.
>> Yeah, I'm talking about my Instagram post. Like the YouTube video is doing well. Sure.
>> You just got to give it time. Sam's video has how many millions of views now? And that just took a couple weeks to pop off.
>> Yeah, but that was on YouTube.
>> Yeah, which we which Ren just uploaded his to YouTube.
>> Can you sort by popular?
>> Damn.
>> Oh, Jen beat him now.
>> Jen.
>> No, but that's that's Jalen and Dean.
>> Yeah, this isn't his render.
>> Yeah, that's from this stack. That's a short from the video.
>> Oh, never mind.
>> How many Lego?
>> There you go. Watch your step. 53 million views. That's pretty impressive.
>> That's such a what a what a a breadth of our content. Yeah, >> it's like like essentially like a true scale video, a satisfying render, and then like a cool build >> video. A couple cool build videos >> and even a react episode.
>> Yeah. And a react.
>> That's like That's like every every content bucket, you know.
>> But yeah, so like my bubble wrap render last year got 350 million views on Instagram, >> which is insane. It it got so popular that Instagram themselves were like, "Hey, can we repost this?" And I was like, "Yeah, sure." And then that video got like 50 million views. Um, so I was like, I bet this will probably get a few million views in the first week, and it's sitting at like a quarter million right now.
>> Mhm.
>> So, yes, a lot of views objectively, but relatively speaking, spoiled.
>> I'm spoiled. I'm a spoiled little view brat.
>> I'm surprised you guys didn't uh pick Clint to be the winner. To be honest, >> I was not judging. So, >> honestly, I thought I thought Clint was amazing. I did not expect it. I thought it was a really, really cool sound bath.
There's a there's a Reddit post being like, "I didn't think it was satisfying." And I'm like, "I don't think you're right.
>> I don't think you're easily satisfiable."
>> That's But like uh Danielle watched it and she was like, "What did did they vote for Sam as like rage b as like engagement babe?" Like, "No, they I was actually kind of surprised that they said that in the moment, too." Uh cuz that's that's what they said.
>> Like Sam's render is really good.
>> Yeah, >> it was really good. But I thought Clint's render was very good.
>> Yeah.
>> Um >> Yeah. I don't know. I mean, it was it was Dean and Griff doing the judging.
>> Uh, >> it did it did help with some of the engagement, though, because there were a lot of comments that were like, >> "Lynn's won. What are you talking about?
>> Clint got robbed."
>> And And it's like Jordan Allen yet again. I I was complaining to him before we actually did. I was like, cuz I knew I didn't know what his thing was, but I knew it was going to be another weird render.
And like weirdly, this is a satisfying render challenge, not a weird render challenge. Why do you keep doing the weird ones? And Nick and I were both kind of like, he's just doing another weird thing. But then I realized, I'm like, no, this is just good content. At the end of the day, he's making good content for the video because it's unhinged. It's funny. It's chaotic. Uh >> he's like the wild like you need a wild card.
>> Yeah. You know?
>> Yeah, totally.
>> And it still is a really really good render. Like it's still pretty satisfying.
>> Very good quality render. It's It is satisfying I suppose. It's just weird, >> you know?
>> Definitely weird.
>> It's like it's like ASMR. Like people get like triggered by different aspects of ASMR like tapping or like a little scratch or something like that.
>> He scratches the weirdos, the weirdo freaks.
>> He scratches the weirdos.
>> You got the normie the normie satisfaction, the avantgard satisfaction, and then the weirdo freak satisfaction. Also, in that shot there, I'm wearing the mic on my on my shirt, and it keeps catching the light, so it looks like this white square, and I keep thinking I'm wearing like some sort of like father outfit, like a suit.
>> Yeah, that does. Oh my god, it's like >> I keep looking up. I'm like, wait, what am I wearing? Oh, that's just the mic.
>> Father Ren, >> I didn't notice that. Pop a like one.
Pop a like.
>> Um, >> yeah. Anyway, so that that was that was a fun video to do. I really enjoyed actually working on that, too. It's like it's one of my favorite things is when I have the chance to say like, "All right, I'm going to try to make this cool render. That's all I got to worry about.
I don't I don't have to worry about whether or not the YouTube video is going to be good."
>> You're not writing like a science paper about some, you know, scale topic. Which speaking of which, today is the start of me diving into the next one, and it's going to be, I think, a hard one, but I hope not. The whole idea is that it's going to be easy, and we can make this pretty quickly. Nick and I are going to be working on this. It's about rising sea levels, and I'm going to visualize that. But I I have a feeling I'm like I'm wandering into this being like, "Oh yeah, how hard can it be?" But then I remember back to the solar panel video where I had the same thought. I was like, "How bad could it be?" It's like, "What's so hard about this?" This ended up being like one of the hardest videos I ever did.
>> Mhm.
>> Well, don't be afraid to pivot if you need to pivot to a different idea.
>> I mean, yeah, we'll see. We'll see.
>> We can we can come up with new ideas >> with an hour of effort.
>> Well, yeah. So, uh this time around, I'm trying a few things differently. Nick and I are going to uh brainstorm on this together. I got a new app called Miro.
Have you heard of this? It's a collaborative brainstorming app. It's like an infinite Zoom type uh canvas.
Mhm.
>> Uh, and it's collaborative, so we can both like make notes, make little like uh post >> infinite whiteboard.
>> Yeah, it's like an infinite whiteboard.
You can zoom in, zoom out. Um, >> I haven't actually tried it yet, but I'm hoping that between the two of us, we can brainstorm something together.
>> Um, make it work. Cuz again, it's like it's one of those things. It's like, how do I make a fun video that communicates this without going too deep into the preachy side of things?
>> Yeah. cuz it's like a climate change video kind of at the end of the day.
It's a climate change video.
>> It's like another mood lifter like like Dean and Jordan's garbage one >> and but like I went deep into that on the solar panel video. It's like I went was like this that was very much a climate change video. Um and >> it's like okay I I want to I want to I don't know. I haven't figured out how to communicate that stuff in a fun way yet.
I've done it before. The Nuke video I think is a great example. Um, so that's my concern on that.
But from an effect st, the reason why I want to do this is because I think the effects will be easy. You just load up Google Earth inside of Unreal Engine.
You set whatever the you figure out whatever the the rising sea levels will get up to, whatever altitude that is.
Yeah.
>> You put a plane at that altitude because the Google Earth plugin, I think it's called like >> Cberus or something like Cerebellum.
>> Cberus arms.
Um, I forget the name of the the plugin, but it's literally just Google Earth, but you have Unreal Engine lighting and rendering, so it's pretty nice. I did it for the mountains video, the scale of mountains.
So, yeah.
>> H I made uh my first movie script with my daughter.
>> What?
>> Yeah. So, she's like she's been seeing all this green screen stuff and I think it that helped her understand the idea like, oh, you film a person and you put the background to be something different. It's like, "Let's make a Minecraft movie video."
>> Oh, really?
>> Yeah. And so we was driving the car.
She's like, "Okay, I want to make a video." And okay, well, >> if you want to make a video, usually the first thing you do is you make what's called a script. You know, you kind of figure out the story like, "Okay." And so then like we talked through it. And she figured it out. And you know, I helped her like, you know, where are we?
What are we specifically doing, etc. And then when it came time to write this like, "So you can either like write a story board or make a storyboard, draw a storyboard or write a script." She starts for about, you know, a minute and she's like, "I don't like doing this.
Let's just go shoot it."
Okay. Okay. Okay. It's like I can totally see how like a storyboard is a pain. So like instead, let's do this. I want you to we're going to talk through the story together. We're going to say all the lines into my phone. I'm going to record it and I'm going to take that.
I'll make the script out of that. And so we did that where she talked through the entire story and if I needed a prompter for dialogue, I did and whatnot. And then I took that and gave it to Claude and I said, "Format this now as a script for a six-year-old." and Cloud put together a really nice >> like formatted with characters and colors and colored boxes around all the dialogue and action and scene breaks.
>> Is that a PDF or are those pictures?
>> It's a PDF.
>> What?
>> And like this is such a nice >> show to camera real quick. It's got to be a way to >> I don't know if you guys can see that.
>> Um >> really squares. It's just like a really nice formatted like take my my daughter's like so then then then we go into the backyard and then there's a portal you like because she's talking through it and it's like it takes that and cleans up to like scene one kids are playing ball in the backyard a portal opens they turn around in awe you know it's like >> it just polishes it really nicely so now I need to try to figure out how I can do Minecraft effects >> like in a really casual dad way.
>> Oh here's something I'm curious about.
Are there so like I love being able to just take voice notes like during a brainstorm session.
>> Sorry. Finally a corridor short.
>> Finally. It's been years. Uh so it's like I love being able to take a like a voice memo of a brainstorming session and summarize that with AI just to kind of get a clear idea of like the main things we were excited about, the things that were like eh. It's just filler words for the for the minute. And but the problem that I've faced when doing that is that it can never differentiate voices. Can never differentiate who's talking. And I'm curious, especially from the chat, if anyone knows of any tools that can listen to an audio feed and differentiate voices >> while transcribing that feed.
>> That's a >> because that would be extremely useful for me.
>> Yeah.
>> So I think that I mean that tech exists somewhere. Like I feel like there's some sort of like subtitling like plugin or something that and like knows to separate people. So combining that with >> like I've used dcript but like >> it doesn't have the LLM side.
>> Yeah, >> I could always like use dcript to like export out the stuff, but ideally I just want something I can just upload. It generates the transcript for me with the different voices already made and I can upload that to chat or cloud or something. Yeah, I I have noticed that if I do emphasize like when I'm feeding the the transcript like this is two people talking, try to figure out where the breaks are and they're and how you know who's talking when and do your best to label it and it generally gets it right.
>> I I that's what I've tried in the past, too. And at least my attempt with it was pretty inconsistent.
>> Um Microsoft Teams does it.
>> Microsoft Teams, I guess.
>> Oh, it's Teams. Yeah. Steam. Steam's.
>> Hey, I need >> the other Seattle company.
>> If you've been in pre-production on a short for two years and you can't get motivated enough to work on it, don't worry about it. Just do something else.
>> It's totally cool.
>> Yeah.
>> Like, you know, you want to make art, just go make other art.
>> Don't let it be.
>> Even in pre-production on a short for the last two years.
>> Yeah. Just make something else if you want.
>> Yeah. I've always been a big fan of saying like quantity over quality because from quantity comes quality. It just means that move on.
It literally that's advice earlier, right? It's like I know you didn't ask for this advice, but like literally one of the best things you can do for for art is finish it and move on.
>> Like it doesn't matter if it's perfect.
I know like your soul will care. But like it's okay. Just make something and move on to the next thing. And you know what? You can that next thing could just be another of the same thing you did.
You're just doing it a different way or doing it better. You're redoing it.
That's totally fine. But it's it's it's very good to just have something tangible you can move on from. So like if you do three small ideas after that, then maybe that'll just respspark an idea for that original one and then you go back to that one like recharged and you can finish that or something. You know, >> I think it's also super important not to polish something at the beginning like to like if you're want to make a short film and it's like your first one like shoot it on your phone and just get it done. Like do whatever you need to do to get it done. If that means doing it in the most roughest, shoddy way possible, great. Because getting it done is more important than like making it polished.
>> But at the same time, I also understand the idea of like not being able to get motivated enough to work on it consistently. And I think that's just uh a symptom of not being passionate about the idea. And that's where art becomes problematic.
>> Yeah. cuz it's like I've been there so many times where it's like I have to do this thing and I literally would prefer to do anything else other than this thing that I have to do and the passion is just gone from it and you have to just kind of work through it.
>> That's hard. Uh and and and anytime I find myself getting obviously if I'm trying to finish a project I'll just go ahead and try to finish it. But like >> I think it's it's it's you know it's not a bad idea just to walk away.
>> Yeah. Well also like the work can be daunting sometimes.
>> Yeah. where it's like like a 3D thing.
Like it takes a lot of work to do that.
I mean, music's a good example where it's like music can be a thing where it's like, oh, writing the song is going to be days and days of writing and going into the editing program and tweaking things and finding sounds, but you can also just sit down and play a song on a piano and walk away and the song's done the moment you walk away. And I think that's also really important to like remember that like, you know, if that work is daunting, just don't do it.
you know, just do the thing that's the equivalent of like live theater then to your film making, you know.
>> Yeah.
>> Or or obviously like you can you can also just try breaking it down into smaller smaller more approachable bits.
Like if you're trying to make a whole like 3D uh short film, it's like that's a very intimidating thing. You're talking about multiple different shots that are all CG and that they all have different assets in them. Maybe you just uh >> start by just making a like give yourself a render challenge. It's like, all right, I have this character. I'm going to just finish modeling this character. I'm just going to make some cool beauty shots. That's my only goal.
I'm not worried about the short film.
I'm not even thinking about any of that.
I just want to make a cool little render of my little guy here.
>> Yeah, it's like a single shot like like a little motion poster or something.
>> You know, just something like that. And now that might actually, you know, resp-spark your passion for the whole thing and maybe you'll be like, "Yeah, and now I'm working on the next shot."
And it's like, great.
>> Yeah. And if it doesn't, you just move on too. You know, >> I think a lot of cartoonists do that on YouTube where they make like 5second long or 10-second long cartoons. They just put that out there cuz it's like you get the audience feedback faster.
You make something faster and you can always spend more time on a longer cartoon.
>> But yeah, Josh was mentioning that uh you know, he just had the rug pulled on a film that he was working on and like that's tough like when you have put so much time into like a project and then like it feels like it all falls apart.
>> Yeah.
>> But the thing with films is that like you just have to always push. I can't remember who said it to me, but like the most the film's only dead when no one's pushing for it anymore. And so many films just require pushing and pushing and pushing and pushing until it happens.
>> But that's also why it's really nice to like make things that don't require other people to say yes to them too, you know?
>> Like and to fill in those gaps like if you're struggling with waiting for financing or permission.
>> Yeah.
>> Do extra stuff on the side or take a break, play some video games.
Yeah, I know. Knife advice. No, actually, this is uh Griffin told me all this stuff earlier.
>> Yeah. Yeah. I'm just regurgitating what Griffin told me right before we came in here.
>> Yeah, we're just like the TAs.
>> He's off camera like coaching us through advice.
>> Yeah, guys are wearing ear wigs.
>> You might be listening right now.
>> Yeah.
>> Yeah.
>> Getting where he said to buy knives.
>> To buy some. He said that that would fix the problem is to buy knives. Okay, so >> buy knives.
>> Go buy knives, >> dude. Knife stock is about to go up.
>> Invest in knives. That's what Griffin is saying. I'm channeling Griffin.
>> Yeah.
>> Oh man.
>> Yeah. I'm kind of blanking on what else to talk about right now. A lot happened.
>> I know, right? It's like I I was gone for the entirety of last week cuz I I got back from Vegas and literally the next day flew to Oregon >> and hung out with my family and my wife's family. Yeah. And >> so it was a whole It was fun. It was super fun, but it was so much socializing within the span of a week.
Like I probably met and shook hands with over 300 people at NAB.
>> Oh yeah.
>> Uh easy. And it's just like cuz you just you meet 100 people every day. And yeah, it's it's awesome. It's really It's always awesome meeting fans.
>> I'm surprised I didn't get sick.
>> Yeah. Sometimes >> I' I've never gotten sick from any of these conventions.
>> Really? I've gotten sick from a few of them.
>> Yeah, I've noticed it. It's built different.
>> But what what about what you were doing just before NAB, >> right? Yes. I can't go into super super detail, but yeah, Sam and I were at a Mr. Beast shoot, uh, which was >> honestly we had a great time. Um, it was an awesome experience.
Uh, but the operation they have there is pretty amazing. Like there's constant productions all happening simultaneously with multiple production teams. They have the largest indoor sound stage in North America built out there. Um, which is crazy. It's just it's crazy to see that like the compound and be like this started from a YouTube channel and like it's a full-blown studio like on par with any other big studio that exists.
Um, it's not quite the size of Tyler Perry's compound yet, but you know, give it a few years and who knows. Um, we also got to meet like a bunch of the editors and like the people that work on the videos there and just the stuff that those guys do was insane. Like we I have a picture of a timeline, but it's truly like 80 layers, just all this different sound, all these different graphics, all this different video. And they said that they put an episode together in like three weeks of editing, which is >> crazy, >> bonkers.
>> And then from there, they take another 10 days to localize it. Um they have so >> Mr. Beast is a worldwide operation.
>> In other words, >> Mr. Beast is kind of Mr. Worldwide.
>> Yeah.
>> Um they have dedicated hired voice actors who play consistent characters on the channel.
>> That's cool.
>> Um and they they have this whole wall that has a picture of every voice actor for each person um from each region. Uh and that was really interesting to see.
So, it's like they kind of like keep you attached to like here's the team that's like playing these characters here and here's like the actors who are playing the characters here and here's the actors who are playing the characters here in this country. And it's just cool to see that like they've taken a picture of each person and put them there so you can like identify with your fellow teammates who are overseas who you don't necessarily get to see. Uh like like Jimmy is voiced by the uh voice actress that does Naruto in Japan. Um, and so yeah, they take 10 days to go through and like redo all the graphics in all the languages and to do all the voice acting in all the languages and dub every single project, get that all cut back in, mix the music, mix the sound, and then uploaded.
>> It's a nightmarish work.
>> It's so much work. That's because it's because it's also just so many graphics, too.
>> It is like it's regardless of how you feel about Mr. business content. You cannot deny the incredible highlevel editing and graphics work that's happening because they are turning it around so fast with such a frenetic pace like cuz you know it's like density.
It's like information.
>> Yeah. How do you get like a like three weeks to finish a video that's like I spent a 100 days on a boat, you know?
>> Yeah. It's a lot of footage to go through. Uh, and that's actually a thing they talked about like they have >> because like they need to record everything when you're doing like reality stuff because >> you can't miss a crucial moment where like if somebody says I give up or you know they cross the, you know, the line of >> Yeah.
>> defeat, you know, it's like if you miss that shot, you miss the whole video.
>> Um, >> show me a picture of a bucket of GoPros.
No, it wasn't. It was a bin. A tub a tub of GoPros.
>> A tub of GoPros. And like they're they're all stacked like naked GoPros just stacked next to each other just line after line after line and then stacked vertically.
>> So, you know, they have, you know, people are on like a cameras making sure everything's being recorded. And then they have like story editors who are constantly taking notes during the shoot so that when all the footage is ingested, they they already know which points of the footage they want to use.
>> They go straight to Yeah.
>> and go straight to it cuz there's no way somebody can comb through all that footage. And I think they go through like they get like a pabyte of footage from like each one of these like videos they do these like reality recordings >> which is crazy. It's crazy. Like that is >> that's so much >> for anybody. That's crazy.
>> Um >> how much do they store that or do they just like once the video is finished or they just like wipe? I imagine they probably wipe a good chunk of it.
But uh I was also told a story about Beast Games and they're talking to like Adobe and they're like you know so we have a problem with multicam you know we it's not working like okay well you know what's the issue like well you know we load in the thousand cameras and just it's not there's no way to like work with the feed. They're like well okay so you loaded in 100 cameras like no >> a thousand >> a thousand camera feeds. They're like oh >> well that's >> okay we'll have to work on that.
>> They're like who in their right mind would ever have more than 10 feeds? So, we'll design for a factor safety of like 30 feeds at the end. No one will ever do more than that.
>> How do you like switch with that many cameras? Like cuz like with multicam in like Premiere as you're editing, you can hit like one, two, you know, like one through zero and it'll automatically make a cut and switch to that camera in the timeline. How do you do that with a thousand?
>> Good question. Good question.
>> See how many keys are on a keyboard and you combine that with control, shift, and alt different variations of that.
Yeah, >> it's like you have all the keys, then you have shift plus all the keys, then you have shift alt plus all the keys, then shift alt control plus all the keys, then you have alt control plus all the keys, and shift control all the keys.
>> And that's just switching. Like you still need your normal, you still need your normal macros, you know, just like adding a normal cut, you know, doing some ripple edits and >> and all that.
>> Hey, Griffin's in the chat. My podcast >> and we started giving our own advice.
>> Yeah, but you know, it's like when you have when you're doing something like Beast Games where it's like we're going to have a thousand people you need to have cameras pointed at all the characters to catch stuff like yeah you're going to have a thousand camera feeds like who would have thought that people would push game shows to that point but I don't know it's really fascinating to hear about that kind of stuff like >> the production side of things yeah >> the technical techie as as techie boys it's like >> the technical side of Mr. beast is a crazy operation. What the people are doing there is bonkers and they are super high level.
>> I like your story about how the uh all the PAs were like faceless.
>> Oh yeah, it's a really cool aesthetic too when they're like when they're shooting with how they uh dress and costume up the uh the camera people and the PAs and stuff.
>> They keep the aesthetic going for the for the people that are being like in the the situation. All right. I can't talk about the shoot beyond that though.
>> Yeah. when I hear when I hear about like all the all the production stuff, I'm like >> I'm so glad we're so much smaller than that.
>> Me too, honestly. Like there's a uh >> you know, yeah, like we're not cranking out 200 million view videos, but there's also like I don't know. I'm just I've come to realize that not just from Mr. Beast, but like from all the YouTubers I talked to, like Corridor Digital has perhaps the best work life balance out of any big YouTube channel. And even just like the logistics of the videos too where it's like I'm only ever editing like two cameras tops, you know, like two cameras and a screen cap.
>> Sometimes a third camera, >> maybe a third. It's like the the worst that it gets is like when we're doing like like the satisfying renders and I just had like six mics out, you know, but those are really really quick to sync up, you know? So it's like that's not even that it just gets annoying when I have to like like put everything together, you know? It's just like takes a little bit longer to get the sync timeline going, but like six mics, two cameras, and that's it.
>> Yeah.
>> And the screen cap.
>> So, uh, with the the cost to make where they actually make profit on these videos. So, the idea, >> well, okay, so here's the thing, though.
First off, don't forget Mr. Beast videos get 50 plus million views pretty much guaranteed.
>> Guaranteed.
>> Guaranteed. So, you're looking at 50 million views worth of AdSense per video. And and these are like high view videos to begin with. So advertisers want to be on those videos. So that CPM is going to be higher. On top of that, Mr. Beast is Willy Wonka.
>> He has a chocolate factory. Everything is in service of selling chocolate to children.
>> And also all those videos are also being pumped to every other like country too.
And so it's like >> it's a huge operation. I mean they're spending a lot of money. Don't get me wrong. They're spending. I guess what my point is is that like their revenue stream goes beyond just AdSense.
Obviously, their their brand deals for Shopify or whatever that are going on the actual videos are in the multi-millions, I'm sure.
>> U and the AdSense is probably around, I don't know, maybe a million.
>> Hard to say per for a video that gets 100 million views.
>> Probably million.
>> Um but then it's like when it's selling chocolate throughout the whole video, everyone's like eating all the Fastables stuff. It's like, okay, how much are how much chocolate are they selling? And it's all Yeah, it's it's >> it's operation.
>> Yeah, he's got toys. He's got C. I mean, they got everything going on. It's crazy.
>> It's kind of similar to how like Mark Robber, you know, he's got Crunch Labs and his videos now are kind of in service of selling Crunch Lab uh packages.
>> Yeah, I think Crunch Lab far outweighs the revenue from YouTube for Mark.
Whereas, >> yeah, >> I don't know if it's quite the same situation with with uh Beast Industries.
>> And also, I'm not sure how much Beast is actually like setting aside to keep. I I get the I get the feeling that pretty much all the money they're making, they're reinvesting into their operations or whatever they're doing.
>> Yeah, maybe.
>> And I don't know what sort of outside investment they've taken either. I mean, they've got the whole like freaking Beast World thing in uh >> they went to Riad for the Beast games.
So, they've got some like royal crown prince money coming flow on in.
>> Well, not sure about that.
>> Yeah. Well, the revenue makes it uh makes itself known when you get 200 million views in a video. I did see a poster on the wall that said 26 main channel videos in 2025. And I was like, damn.
>> Wow.
>> That's a video every two weeks.
>> Yeah.
>> Oh my god, that's that's pretty nuts. I mean, for us to do 26 Sunday videos in a year is like pretty dang good, >> you know.
>> Yeah, that's tough.
>> That's quite tough.
>> That's quite tough.
>> Yeah, it's got a crazy team there. So, shout out to all the editors and producers and >> crew Beast Industries. You guys are doing incredible work.
>> And again, everything we're saying is outside of your opinion of Mr. Beast. It >> has nothing to do with the the content.
>> We're just saying, hey, the crew is working hard.
>> Yeah.
So, hey, you know, if you like Mr. Beast stuff, whatever. Enjoy what you like to watch, guys. It's all good. Don't let other people tell you what you can and can't watch.
And it's like, yeah, I I I constantly see a lot of hate online around him, but at the same time, I'm like, all right, if I'm seeing so much hate, why are his videos still getting 100 million views?
>> Well, it's cuz the hate is coming from adults and the views are coming from children.
>> The hate is coming from the inside.
>> Yeah. And and it's worldwide. That's a whole other thing.
>> Yeah. Yep. Hey, people like to watch they like to watch the stuff that makes them feel something.
>> But at least they're putting effort into the content, you know? Like I'm always jealous of of the creators who are able to like get a million views off of a video where they literally just like spoke into a camera for 5 minutes.
Didn't even edit it. Just straight upload from their phone.
>> Million views.
>> I'm like, are you >> That's always That's always hurt me.
>> Yeah, especially because like we picked the like >> So, uh I can't remember. Do you remember there's this comedy skit or this? So, uh, Aziz and, um, two other guys, I can't remember their names.
>> Aziz and Sari.
>> Yeah. And two other comedians, they're pretty well known. Um, they had a comedy troop back before, you know, Aziz kind of blew up. Um, it's like Human Giant or something like that. And there's this one where it's like, so Aziz plays a or uh, one of the other guy in the group. I Jason, is that it? Paul Shear? I think it's Paul Shear.
>> Yeah. Human giant. Um, >> so one of the guys wants to be YouTube famous. This is like really early in YouTube, right?
>> Oh, wow. This Yeah. 2007. This is the birth of the partner program YouTube.
>> Yes. So, this is super super super early. And one of the guys wants to be YouTube famous. How do you get YouTube famous? And he's like, you got to do something extreme on camera. So, he makes a video where he cuts his uh he cuts his dick off with with like garden shears and records it and becomes the most viewed video on YouTube. and he went through all this pain to do it and he's like, "You got a million views on this video that you're famous for.
>> Most viewed video on YouTube. I did it."
>> And then it's like, "Oh, Mr. Sin, we got a new most viewed video on on YouTube.
This guy makes funny faces."
>> He's like, "Oh, IT'S SO GOOD. OH MY GOD." And like just the dichotomy like a guy who had to cut his dick off to get views versus a guy who just makes funny faces and like triples the views. That's how it felt a little bit when we were doing corridor digital >> and all the shorts. It's like we worked so hard and spend so much money and sleepless nights doing effects and sound and color grading and like cinematography >> all to make a 3minut video >> that gets like 400,000 views and >> well back in the day our videos were getting a lot of views. Some of them would still only get 400,000 views. But >> yeah, yeah, some of them got a lot of views, too, of course.
>> But then somebody else like, "Here I am going grocery shopping and like or truly like playing video games and making funny sounds was the equivalent to making funny faces."
>> Yeah.
>> So that's uh Yeah, that I think about this human giant skit all the time.
>> The life of a YouTuber >> boiled down. They're ahead of the curve.
Yeah, >> that's that's funny. Yeah, because that that was so early days.
>> Yep. But then also it's like I don't get a million views when I make funny faces and so I should be happy that I do get a million views >> even if I have to work extra hard for it because I'm just happy >> that I'm getting views. Yeah.
>> So you could be there trying to break out getting 100 views per video too, you know? It's like so be thankful for what you got is the only myself.
>> Would you want to be doing funny faces for a living, you know? Well, you know, we all like to say that deep inside we're happier because we get to express ourselves as artists versus being funny faces on camera. But chances are >> I mean, yeah, definitely like it'd be nice.
>> But also like there would be a part of you that would be a little bit like >> I'm just a funny face.
>> Time to go to the funny face factory.
Another day at the funny face.
>> But also though, like to a certain degree, isn't that kind of what I already do?
>> Yeah.
>> Isn't that Isn't that what I do? I just make funny faces. The end.
>> Yeah.
You You're so much more than a funny face to me.
>> You're also a funny body.
>> You know what's crazy? My personal YouTube account turns 20 next month.
Wow. 20 years I've had that account.
>> On YouTube, >> dude. The Fatal Farms Robocop scene.
>> They probably have like >> Was that the one where every dicks get shot by gun?
>> You guys can watch it on quarter.com right now.
>> That's right. We did do a react on that.
Yep. Finally got to meet the fatal farm.
One of the fat >> Wait, that's not actually Scott Manley.
>> No.
>> Hello.
>> I don't think so. Minnesota Clay's been in here for a while.
>> I'll text him and make sure.
>> Wait, I think I met Minnesota Clay at NAB.
>> He goes, I'm Minnesota Clay.
>> Oh, really?
>> Yeah.
>> Oh.
>> Oh, NAB Tech. I'll have my Here's my one NAB tech and then we'll probably wrap up the podcast here and go to our afterparty oncorigital.com.
Uh, there was a guy that brought a rig that he made.
>> Yes. This is very cool. explicitly for like AR film making, iPhone based AR film making >> and he had 3D printed all the rigs and so he had two phones >> iPhone 17s whatever pro and one was running Jetet and the other one was just running like a good video capture and the Jetet one would give you all your motion capture. It let you preview your set, you know, on the phone. And the other one would be getting the actual video capture that you would be using for your actual editing. And it's all rigged up with batteries and a synchronized thing that gives you time code sync between the two set into a system that had a green LED ring around the lenses of these cameras. So that if you put up a retroreflective backdrop, you could just turn on the green LED ring and you'd have flawless green screen no matter where you pointed the camera without having to worry about green lighting or green bounce. and >> gonna pause the video. But >> yeah, it was all self-contained rig.
>> It was really cool.
>> It was pretty cool. He was holding up a circle. It was like the little >> We're gonna do that for Son of a Dungeon instead of Corridor Key.
>> Well, you'd use corridor key, of course, to key out your green screen eventually.
>> But yeah, it was really neat. I think Sam already kind of has a Jetet rig working right now uh for getting motion capture and then it all runs through Bible, which I believe has quarter keyed now.
>> Yeah. Worked in.
>> I think it's official. Yeah.
>> Yeah. Which is awesome. So, I I I was seeing some uh some some reactions to like the AE scripts plugin being expensive.
>> Oh.
>> And and people are like, "This is supposed to be open source and yet oh unless corridor gets a cut, >> you know, how do you feel about that type of reaction?" Like, so this is a whole open source movement where we're making, you know, quarter key, anyone can download it, anyone can mess with it and play with it, but there are paid plugins, >> but only if they work out a deal with us directly. like h how do you feel about the sort of negative reaction to there being paid plugins at all?
>> Well, I think uh first and foremost, the most important thing here is to keep the open- source free side of it going forwards. So, I don't want like any paid plugin to be at the expense of like the passion or exploiting the work people have put in. Mhm.
>> But if you're an actual film studio and you want to use a plugin and then be able to send an email to a developer when you have a problem, that's not going to work with open source where everybody's volunteering their time >> unless it's like super super super mature. But still, it's like you can't really hit somebody be like, "Hey, I'm working on this product and I need this thing right now." It's like they're just volunteering their time, right?
>> Yeah. It would kind of just go out into the ether and like maybe someone would fix it, maybe not. And if a studio on a deadline, it's like we need it fixed now.
>> Yeah. So you know to me it's like the commercialization side of quarter key has two things. One I wanted to restrict it so that a company didn't go in or really a hundred companies didn't go in and try to make variations of this which would end up just fracturing any like passion and community around this where now you have 100 different versions everybody competing with themselves.
>> I want to bring everybody into like one center point where anybody with a cool idea can bounce it off of other people with cool ideas and we can all work together to make this tool. Because also deep down inside I don't feel like there's that much revenue opportunity in like a visual effects plugin. Um I know people are like you guys gave away a million dollar idea and I I really appreciate that people think it's worth a million dollars but there's just >> there's just not that much money in visual effects. I mean maybe you can swing a deal like that if you pull off something crazy with Adobe but in reality it's tough to make money in visual effects and like my priority >> in in tools tools >> in tools in tools. I mean in general.
>> I mean in general. Yeah. But we're talking tools for this.
>> Um >> you know in general like it's what's more like and I have a day job you know.
>> So what's more important to me is like making a good green screen tool. That's like my number one priority. And the way to that is with open- source communitybuilt tools. So that's why I restricted commercialization. And I want if there is commercialization, I want it to happen after there's good support for something like I wouldn't want like a paid After Effects plugin to come out if there weren't already people working on free After Effects plugins. So there's an option for everybody else, >> you know. Um, >> and the other thing is like even though I'm in a position where I can work on this and not have to worry about necessarily making revenue off Coror Key and people are volunteering the time, like not everybody's in that same position. And if somebody has a really cool idea and a thing they want to build, but like they also need to make their living, like I don't want to necessarily ban that from happening. So it's on a case by case basis, you know?
It's like >> what are you going to offer? Is it is it this certain thing that you've built? Is it a certain level of support that you can offer? And that's really what it comes down to more than anything else is support. Like if you have a bug and you have a problem, you got to be able to turn to somebody to fix it, you know, if you're trying to actually run a studio with this tool. Um, >> and like also what's the value proposition of it being integrated in in something else like Bible for example?
We did work out a license agreement with them. So that quarter key is now built into their program which >> does a whole separate thing where it's like right it takes your footage and it renders out and generates like normal maps, roughness maps. So you can relight 2D footage in a 3D scene which is a pretty powerful thing.
>> But if it also is able to key your footage in a really good way at the same time that would be even better. Yeah.
>> So, it's like that makes sense to incorporate it into there, but they couldn't just do that without a license agreement >> because they're selling a product, >> right? Exactly. And, you know, we're taking it a step at a time, too. Uh figuring it out like, you know, which is also why it's like close the gates and open them little by little. Like gates are fully open for anybody doing anything they want for free that's sharing it. Like 100% open for anything.
You can truly make whatever you want if you're going to just release it and share it with anybody, you know?
>> It's just >> like easy quarter key.
>> Yeah. Like easy quarter key. or really anything that people have made. It's just yeah, it's when if somebody wants to put it into a commercial program or sell, you know, the stuff they're making, that's where like let's just take it a step at a time, case by case basis. And yeah, it's actually been a really cool experience. And like everybody I've worked with on this has been like really awesome and supportive and like >> yeah, >> I'm trying to be upfront with like very much like I'm learning as we're going here too. Um, and just moving a little slow to make sure I take my time and try to fix things as I go. So yeah, um it's I don't know. It's really cool. I think uh I get why somebody might be like, "Hey, there shouldn't be a commercial version of Cordor Key, but as long as there's like so many free versions out there and like public support still going and I'm very much like prioritizing all the open source stuff also >> like and so yeah, cool." Like if somebody reached out like I built this thing, I want to support it. It's like okay, sure. Let's do it. Let's figure it out, you know? So I think I think it's fair. I think it works fine. Um, and it's not at all right now coming at a detriment to like the free work people are doing. And like in the future, like let's say, you know, we're developing a new version of the model. Like say somebody comes up with like their own version of it in the Discord, like I made this and it's like, holy crap, that is insanely good. Like if we wanted to commercialize that, I would bring them in on the deal, >> you know? Yeah. It's like we'd figure it out. So yeah, I mean, I'm not looking to like, like I said, my highest priority here is making a good green screen tool.
And the only way that really happens is through mass adoption and mass use. And the fastest way to mass adoption and mass use is open source.
So that's where that's where we're going. That's the plan, everybody. Yeah.
And yeah, the current Da Vinci plugin for for Fusion. Awesome.
>> Yeah. All freaking killed it on that.
>> Also, Josh Davies made a after effects plugin as well, >> um, which is 100% worth checking out.
>> Okay.
>> Yep.
So yeah, that's my thought. We'll figure it out, but it's pretty cool.
>> Yeah, I'm curious to see where the V2 quarter key will will land like a better training.
>> Did you see my data set?
>> Your example? Yeah, >> I think you showed me.
>> Okay.
>> But like I'm I'm Yeah.
>> Okay.
>> I'm thinking like a little different.
>> Yeah. I mean, we'll see just like how intelligent can you make a model? Can you make it so it gives you shadow passes? Can you make it so it gives you >> Yeah, that'd be really nice.
>> It would be very easy to train to get uh shadow passes or at least let me rephrase. It'd be very easy to create >> the training data.
>> The training data.
>> Mhm. Yep. That's a whole aspect too. You know, there's a an interesting training data issue that somebody brought up which is so they're trying to work out flickering and remove flickering from quarter key. So how do you do that?
Well, you just look back a few frames, look forward a few frames and figure out where things are moving, right? You're basically doing optical flow and you use that to help guide how much the mask changes. So you have an object here, let's say it's a semi-transparent piece of glass, and you have another object here, let's say it's a person. And this mask is semi-transparent on the glass and the person's opaque. And then that semi-transparent piece of glass moves over the opaque person.
>> So does that semi-transparent mask then move over the opaque person or does the opaque person shine through it? Of course, we know the person shouldn't be transparent, but we don't have any training data in the training set right now of transparent things that then move over opaque things.
>> So, when the program is training on itself and going, okay, if I look back a few frames, this transparent mask is here and it's moving here. I guess I'll keep moving that transparent mask. And then suddenly you have a hole in the person's face because it never learned that transparent things can move over opaque things in one shot.
>> Wa. Yeah. Okay. And that Okay, that's a good hold. Yeah. Well, not the current cuz the current quarter key doesn't have any flicker reduction. That's a problem that we discovered as we started training newer models that have flicker reduction.
>> And that's just a good example of like, oh, there's a an actual gap in the data set right there that we missed. You know, also I uh I read or understood from another person in the discord that when you change the opacity of something, like let's say you're filming a a piece of glass that you can change the tint on, for example, and let's say you're eyroppering that value as you turn the opacity up and down.
Apparently, that's not a linear thing.
Apparently, opacity, if you turn the opacity up and down on a material or like the transmission is actually a curve. And so even though the model's been trained to do linear subtraction and mixing at the edges of semi-transparent things, when you have something like a piece of glass or something that's changing transparency, that response is actually a curve.
>> And so if you want to get your transparency looking correct in glass versus correct in just simple like semi-transparent things, >> dude, you got to subtract the curve.
>> Yeah. You have to do more training data of that glass at different opacities so it learns how to predict what the actual correct opacity should be on glass versus >> train the curve in. Yeah, this is all very vague training on the way training.
Yeah, >> this is my top level understanding which may be way off, but these are like the weird intricate things that you start learning about when it's like, okay, want to really get good with this? This is the stuff to figure out.
>> Wow.
Yeah.
>> Well, if you're using a blue screen, you're already not using quarter key correctly.
>> Except for when we drop the blue screen version uh today.
>> Is it today?
>> I'll put it up today. I just want to post it on the Discord and we'll we'll go from there. people can start working with it.
>> So yeah, so if people want to, they'll see a link for it in the dis in the quarter creates discord at some point.
Yeah.
>> Yeah. Yeah. And then if people are like, "Oh, this is working. It's cool." Then like then we'll, you know, make it more public.
>> Nice.
>> Uh is depth info something that could help with these scenarios? Honestly, I don't think so. I don't think depth info would help at all with this kind of stuff. Um it's just nobody's ever going to like have depth info when they're like, "Here's my green screen footage."
It's not like, "Here's my green screen footage and my depth info."
>> Um and then for training, like we're just we're not trying to train anything that involves depth, >> you know? We're training mats and we're training unmixing. So, and frankly, mats are already like kind of a solved problem. It's just the unmixing that's not.
>> Yeah.
>> So, I think it's also really important not to try to reinvent things or solve things that other very highly talented and skilled like graduate students have solved like 10 times over.
>> Yeah. So, >> which is why you're not bothering trying to solve the matting issue, the segmentation issue. It's like exactly >> people are already doing that really well.
>> Yeah. So, uh, Ron Albert Ramos, you've asked a couple times you want me to make a video learning autograph and comparing it to After Effects, and I'm just gonna not do that. I'm sorry.
I have no desire to make that video.
>> I also have no desire to learn autograph to be honest.
>> I'm a node node guy now.
>> Huh?
>> I'm a node guy now. I'm a fusion boy.
>> That's my life.
>> How about we uh do some questions in the after party?
>> We're going to wrap up the real podcast.
We're going to go to the afterparty podcast, a secret one that only cool people at corridor digital.com can join us at.
>> So, uh, hey, come consider joining us.
Come on, come come on over.
But also, I hope you guys had a good time watching this on YouTube and or listening on whatever platform you listen on.
>> If you come to the afterparty, you can talk to this guy.
>> Oh my god.
>> So, subscribe now.
>> Yeah.
>> All right. Afterparty begins in three, two, one. Uh-huh.
Vidéos Similaires
OpenHuman VS Hermes AI: Who Wins?
JulianGoldieSEO
285 views•2026-05-29
Long-Running Agents — Build an Agent That Never Forgets with Google ADK
suryakunju
142 views•2026-05-30
This computer is made from real human brain cells. And you can buy it.
Talktmsmedia
3K views•2026-05-28
BREAKING: Microsoft’s New Image Generating Model Beat Out GPT 1.5 and Nano Banana 2
aimmediahouse
122 views•2026-06-03
I Made the Same Anime Fight Scene in Every AI Video Generator
NobleGooseAnime
295 views•2026-05-30
Nvidia Bets Big On AI PCs | New Chip To Power Windows Laptops | Technology | AI Updates | N18S
cnnnews18
3K views•2026-06-01
I Tested NEW Opus 4.8 on Four Projects (Updated LLM Leaderboard)
AICodingDaily
298 views•2026-05-29
3D Platformer Update - NO CAPES
SolarLune
294 views•2026-05-30
