Thanks so much for checking it out! Yeah, I think there’s some really amazing stuff right on the horizon! The next few months are going to be pretty crazy for AI!
Regarding the Gaussian splatting and NeRF as tools for "post re-composing", the latest video from 2min papers about RealFill looks like it could work in tandem with these Splatting or NeRF tools, in theory that is. RealFill's function is basically reconstructing or completing an image by using several other images of the said object or scene as source images, even from different angles, cameras, lenses and lighting conditions. So in theory for example the opening of The Shining shot, if a "RealFill + Gaussian Splatting/NeRF" like tool got additional references to the opening shot of the hotell, like shots or images from the backside of the hotell and other angles, it could reconstruct the entire scene from all angles without the fuzzy blurriness. Things that are missing or partially missing from these images in different angles would be filled in as good as it possibly could by guessing with image generation like approach, like stable diffusion, midjourney and alike, guessing as good as it can with what it has basically. For now RealFill seems to be for still images but hopefully we might get to see a tool combining its function for films with the splatting/NeRF tools. (yes yes "don't mistake a clear view for a short distance", this might take a while before we get there)
That is...brilliant. I was meaning to do a video about RealFill-- but I hadn't made that connection! I'll bring it up next week (and for sure credit you on the idea!)...I mean...wow, that's crazy that we're one tool bash away from that!
Excited to check out this video out!!! I have learned so much about AI video from you already and I’m so excited to keep going! I feel like the technology is still in its infancy and we are only going up from here! 😊
I know, I constantly forget that Luma is RIGHT ON MY STUPID PHONE! I will say, the one down side is that it requires a connection. When I was out in Costa Rica this summer, I wanted to capture a bunch of locations, but the cell service there sucks (beautiful country, terrible internet!) and Luma won't capture locally to upload later. I can't get too mad at Costa Rica, BTW...I did go there to unplug after all!
@@TheoreticallyMedia actually there’s 3 options when creating a new scene and the first one is Upload my Ai bro, so you can phone film offline w your cam and then when have WiFi upload the vid and it will then process it (I just figured that out, great they have that option)
100% I'm always super excited when I run across these techniques-- mostly, since it very clearly showcases that we can ALL do these amazing scenes and films now! Granted, there are a bunch of services you might have to subscribe to for a bit-- but, man-- compared to the traditional method of doing it? It is peanuts! And the next version of these tools is even crazier!
Awesome workflow to overcome the static talking issue. Your videos are highly informative and have helped me develop my content to higher and higher standards. Thank you!
@@TheoreticallyMedia You’re welcome! Your channel is how I keep up to date with generative AI filmmaking advances. It can get a bit overwhelming, given how quickly things move these days, so your channel provides me with an invaluable service. At the moment I’m dead broke, but the moment I have extra cash I promise you I will become a patron. You can hold me to it! Thank you for all you do. It’s so immensely helpful!
I envision the development of an AI tool that can assist in automating the process of generating in between frames in 2D animation. The AI would have the capability to learn and replicate an individual's unique art style, as well as automatically apply appropriate colors to animated characters.
I agree. “Easy” keyframes should be coming soon. Prompt based actions like: “make character A walk to the door and open it” will likely be something we see in the next year. It won’t work great at first, and everyone will say: “this sucks!” And then 5 months later, it will work…ahhh, the circle of AI!
Great overview and glimpse into what's already possible but also where the limitations are! I've been side-tracked a little lately, so it's really nice to get an objective opinion on these recent developments. 🙌
That workflow is pretty cool, gonna try it out. Personally I use a mixture of green screen, stock footage and some AI tools to make it a bit more dynamic. I’m almost done with a 90 minute long animated sci-fi movie (really), and all I know is a little bit of video editing. Good times and it’s only going to get better.
That's awesome! The one thing I haven't seen yet is an AI Video Editor...like, one that will take raw footage and cut it together. I'll admit, I'm a little...unsure, of how good a job it will do. But-- we're also far enough into this that I can never say never about AI!
@@TheoreticallyMedia there are ai video editors that can assemble a movie from various clips. Don’t know how well they work though because I just love editing.
@@TheoreticallyMedia an AI video editor should be quite easy (says me). Most NLE’s use/support XML to define aspects of a project and its timeline (aka edits). Analysing footage is now child’s play thanks to Meta so integrating the output of an AI video analysis with an XML template to produce a bespoke edit really should be mostly an assembly task for any talented and motivated developer.
@@sitedev given how quickly video can be transcribed, I think for “talking head” stuff like my channel, it should be right around the corner. To be honest, that would be a godsend for me- the A-Roll is the worst part, since I have to edit my stupid face and all my stupid Ums. I will gladly hand that task over to an AI overlord!
@@TheoreticallyMedia Vrew does um and other filler word editing in it's software. Their videos are all super short so you would have to run everything as short videos and reassemble them, but it's out there. I recently saw another platform that does that but I didn't make a note of it (sorry).
Wait, you’ve gotten the hang of things?!! Haha- I’ve just learned not to say that. It’s almost like a jinx! The minute you do: 5d Quantum Midjouney will release!
@@TheoreticallyMedia ahhh yes LOL..how true. Maybe I should revise it to say "now that I'm not afraid to dive in and experiment." 🤣 Just a tad more accurate there.
Oh man; thank you so much! You guys are incredible! Been blasting through your videos and was hoping to do you solid! Seriously, your channel is killing it!
Miniatures is the answer here. There are about a million tutorials on here on how to create realistic props and environments. If you're creative you can buy doll props and other things.
I do love that handcrafted method as well. Spent some time as a kid doing model kitbashing as well. The amount of modeling glue I probably inadvertently inhaled might explain a lot about me…haha
@@TheoreticallyMedia , no thank you for putting in the research, testing and tinkering workflows (experiments w/ explanations), and overall creative curiosities that you share to your viewers. Sounds great and looking forward to watching them.
DAAMMMMMIT! Where were you yesterday!!! That is SOOOO Right! Gah, I was wracking my head trying to come up with shots that would work-- I should kick myself for forgetting that one!
Oh thanks so much!! Yeah, I’m always trying to dial that color/contrast in. The room has a stupid skylight which is the bane of my existence. I love it, but the light consistency is always changing, so I sometimes go a little overboard in color correction. I’ll nail it one of these days!
Runway's Greenscreen is the only reason I subscribe. It is just so good. One tip though, I stopped using the green screen output. I export the matte instead. I then layer it over the original footage in after effects, set it as a luma matte and it gives you the cut out, without any of the green edging you often can get from green screen output. It is just that little bit cleaner to my eye.
Oh, that's a hot tip for sure! And agreed-- I'm not over the moon with Gen-2 (although I do like Gen-1)-- but I think the real gem of Runway are the various "hidden" tools, like Interpolate and Motion Tracking. There's some real gems in there if you go digging!
@@TheoreticallyMediathey can’t get out of their own way. They keep shitcanning their bots imaging to add front end bells and whistles without saving previous versions. I don’t think their dev team actually understands what it is people in this space want cause their too busy chasing money…and whatever pika comes up with. Its sad, their bot used to be so good.
Haha, it's been a minute! I planned to have this up on Tuesday...and then...didn't...and then Wednesday I shot it and was furiously editing-- but didn't quite make it in time. I mean, I could have posted it at like 2am, but what's the point? So, it was nice to take some time on the edit today. Got a couple of fancy transitions in there! Ooooh, crossfades and zooms! haha
HA! You pointed it at your screen! Hadn't thought to that! Probably will be new versions of films called "Splats" - prediction! I can imagine indoor VR parks where you can participate in a horror movie - maybe a use for real estate that is abandoned now.
Sometimes the stupidest things are the ideal lo-tech solution! And yeah-- the whole VR/AR thing is on the verge of exploding. We've tried a bunch of times (ahem, Google Glass...) and nothing has really stuck. But we're right on the cusp now.... I'm here for it! I'm Player One, and I've been Ready for YEARS now!
imagine if in the future (assuming cctv arent potatos anymore) investigators can use gaussian splat techniques on cctv footage of suspects or even victims last moments to find details they would other wise have missed, scifi i know but that immediately popped into my head
Very Philip K Dick for sure! I think the game Cyberpunk 2077 had a narrative mini-game like this, where you had to solve a robbery. Of course, in true Cyberpunk style, the footage was streamed right into your character’s eyeballs…so we aren’t there just quite yet!
Back to normal. A lot of the channels that were hit with it seemed to stabilize around the 28th. Like, A LOT of channels. Meaning, something was fixed, although YT still denies it was a bug. Sadly, there are still channels from the Nov 22 glitch that have never recovered. All in all, I do consider it a really valuable lesson learned. Adsense is temperamental and shouldn’t be relied upon. At the same time, I’m not keen on selling VPNs to you via a constant drip of sponsorships, so. I’m working on some ideas to keep the revenue up on the channel, while not being beholden to… I’ll finish this response in a a second, but first a word from today’s Sponsor, Skillshare! …ok, what was I saying? Oh yeah, so I don’t have to be beholden to reading a bunch of ad copy. (Fun fact: Skillshare actually reached out to me a few months ago, and I was like, sure! And then they totally ghosted me…haha.)
@@TheoreticallyMedia LOL...that's good that you don't kill us with the sponsors. But people do what they got to do so I don't like to complain. That's good your channel is back in order. I don't even know what that whole bug thing was about, but man did it sound crazy. UA-cam needs to get it together.
I love the BD guys - they’re like the Milli Vanilli of UA-cam. I can’t wait for the big reveal when they’re exposed as the over enthusiastic AI generated avatars that they clearly are!
Hey my friend, another animation problem that you could speak of (please) is that the AI generated avatars don't seem to breathe, we only see their mouths, and heads move, but not signs of breathing, can this be handled anyways ? thanks :)
Oh, you know- that’s a really good point. Nothing I’ve seen yet, but I’ll keep an eye out! I mean, just thinking about it, you CAN manually keyframe a breathing cycle in- but the whole point of AI is to automate those tedious tasks!
Haha, not taking shots, and not naming names- but I’ll never do a 30 minute video where I’m just reading stuff off a website, …ok, maybe ONE shot fired. Haha
Haha. I’ll go full Werner Herzog and spend 3 months in a jungle documenting the filming of an epic AI film. 3 cast members will die during production and my lead actor will lose his leg in an alligator attack. I will return to the US, only to discover no footage survived because there is no WiFi in the Rainforest.
So, I am not an expert in this, but from a layman’s perspective: Photogrammerty is point based capture, whereas the breakthrough with Splats is that , well- it splats. It is “blobs” of captures that interlock. The really interesting thing is that you can get things like rotational surface reflection with a splat, whereas point based is “you get what you get” Also, because they are blobs, computationally, it takes less processing power.
Haha-- Dying to-- but keeping up with all the info for you all is practically full time!! (Although, I am thinking about sneaking a quick short in soon!)
We really, really, are. I wish we could hop in a Tardis and show this tech to some Disney Animators back in the 40s...Well, then maybe we could all go on a time travel trip 90 years in the future and see where we are!
Ha! It’s a bit involved, and this wasn’t one of my tutorial type videos, more “here’s what’s going on” I think the Bad Decision guys have a walkthrough up, I’ll pop a link in the description. It’s a bit involved to get splats running, which is why I didn’t dive into it
If you want to try out the AI "Secret Sauce" (spoilers, it's Kaiber) this link kicks a few cents back to me, at no cost to you: kaiber.ai/?via=tim
👋
@@LouisGedo hi Louis!!
Fascinating stuff. Thanks for the awesome videos and for sharing the possibilities.
Thanks so much for checking it out! Yeah, I think there’s some really amazing stuff right on the horizon! The next few months are going to be pretty crazy for AI!
Regarding the Gaussian splatting and NeRF as tools for "post re-composing", the latest video from 2min papers about RealFill looks like it could work in tandem with these Splatting or NeRF tools, in theory that is.
RealFill's function is basically reconstructing or completing an image by using several other images of the said object or scene as source images, even from different angles, cameras, lenses and lighting conditions.
So in theory for example the opening of The Shining shot, if a "RealFill + Gaussian Splatting/NeRF" like tool got additional references to the opening shot of the hotell, like shots or images from the backside of the hotell and other angles, it could reconstruct the entire scene from all angles without the fuzzy blurriness.
Things that are missing or partially missing from these images in different angles would be filled in as good as it possibly could by guessing with image generation like approach, like stable diffusion, midjourney and alike, guessing as good as it can with what it has basically.
For now RealFill seems to be for still images but hopefully we might get to see a tool combining its function for films with the splatting/NeRF tools.
(yes yes "don't mistake a clear view for a short distance", this might take a while before we get there)
That is...brilliant. I was meaning to do a video about RealFill-- but I hadn't made that connection! I'll bring it up next week (and for sure credit you on the idea!)...I mean...wow, that's crazy that we're one tool bash away from that!
Excited to check out this video out!!! I have learned so much about AI video from you already and I’m so excited to keep going! I feel like the technology is still in its infancy and we are only going up from here! 😊
I mean...next year? It's...my head is going to explode! Things are SO crazy right now, I can't even imagine where we'll be by next summer!
Awesome highlights and thanks Tim for reminder about the Luma app, downloaded a while back and playing with now. So cool indeed!
I know, I constantly forget that Luma is RIGHT ON MY STUPID PHONE! I will say, the one down side is that it requires a connection. When I was out in Costa Rica this summer, I wanted to capture a bunch of locations, but the cell service there sucks (beautiful country, terrible internet!) and Luma won't capture locally to upload later.
I can't get too mad at Costa Rica, BTW...I did go there to unplug after all!
@@TheoreticallyMedia actually there’s 3 options when creating a new scene and the first one is Upload my Ai bro, so you can phone film offline w your cam and then when have WiFi upload the vid and it will then process it (I just figured that out, great they have that option)
Thanks for putting in the work and bribing this to our attention!! This is going to change the future of art for the better !!
100% I'm always super excited when I run across these techniques-- mostly, since it very clearly showcases that we can ALL do these amazing scenes and films now! Granted, there are a bunch of services you might have to subscribe to for a bit-- but, man-- compared to the traditional method of doing it? It is peanuts!
And the next version of these tools is even crazier!
Awesome workflow to overcome the static talking issue. Your videos are highly informative and have helped me develop my content to higher and higher standards. Thank you!
Thank you so much! It is comments like yours that keep pushing me to learn more!
@@TheoreticallyMedia You’re welcome! Your channel is how I keep up to date with generative AI filmmaking advances. It can get a bit overwhelming, given how quickly things move these days, so your channel provides me with an invaluable service. At the moment I’m dead broke, but the moment I have extra cash I promise you I will become a patron. You can hold me to it! Thank you for all you do. It’s so immensely helpful!
I envision the development of an AI tool that can assist in automating the process of generating in between frames in 2D animation. The AI would have the capability to learn and replicate an individual's unique art style, as well as automatically apply appropriate colors to animated characters.
I agree. “Easy” keyframes should be coming soon. Prompt based actions like: “make character A walk to the door and open it” will likely be something we see in the next year.
It won’t work great at first, and everyone will say: “this sucks!” And then 5 months later, it will work…ahhh, the circle of AI!
Great overview and glimpse into what's already possible but also where the limitations are! I've been side-tracked a little lately, so it's really nice to get an objective opinion on these recent developments. 🙌
Thanks so much Christian! I’ve got your Showdown video queued up for the weekend, looking forward to it!
I use Wav2lip studio sd extension , works well on video for lip-syncing without having to use a mask. It's basically Sad Talker for video
I’m taking a look at a new one called Retalker that supposedly will follow character movement,
Haha, if so- my whole workflow is null and void!
I really appreciate your efforts to make this type of content, may God bless you 🙏🌼
Thank you so much! It really is a pleasure putting these hybrid news/tutorial videos together for you all!
That workflow is pretty cool, gonna try it out. Personally I use a mixture of green screen, stock footage and some AI tools to make it a bit more dynamic. I’m almost done with a 90 minute long animated sci-fi movie (really), and all I know is a little bit of video editing. Good times and it’s only going to get better.
That's awesome! The one thing I haven't seen yet is an AI Video Editor...like, one that will take raw footage and cut it together. I'll admit, I'm a little...unsure, of how good a job it will do. But-- we're also far enough into this that I can never say never about AI!
@@TheoreticallyMedia there are ai video editors that can assemble a movie from various clips. Don’t know how well they work though because I just love editing.
@@TheoreticallyMedia an AI video editor should be quite easy (says me). Most NLE’s use/support XML to define aspects of a project and its timeline (aka edits). Analysing footage is now child’s play thanks to Meta so integrating the output of an AI video analysis with an XML template to produce a bespoke edit really should be mostly an assembly task for any talented and motivated developer.
@@sitedev given how quickly video can be transcribed, I think for “talking head” stuff like my channel, it should be right around the corner.
To be honest, that would be a godsend for me- the A-Roll is the worst part, since I have to edit my stupid face and all my stupid Ums. I will gladly hand that task over to an AI overlord!
@@TheoreticallyMedia Vrew does um and other filler word editing in it's software. Their videos are all super short so you would have to run everything as short videos and reassemble them, but it's out there. I recently saw another platform that does that but I didn't make a note of it (sorry).
Top shelf as always sir. This is why your channel (and patreon) is my go-to reference now that ive gotten the hang if things.
Wait, you’ve gotten the hang of things?!! Haha- I’ve just learned not to say that. It’s almost like a jinx! The minute you do: 5d Quantum Midjouney will release!
@@TheoreticallyMedia ahhh yes LOL..how true. Maybe I should revise it to say "now that I'm not afraid to dive in and experiment." 🤣 Just a tad more accurate there.
What a fantastic breakdown!! Dude loved it!
Oh man; thank you so much! You guys are incredible! Been blasting through your videos and was hoping to do you solid!
Seriously, your channel is killing it!
Thank you brother means a lot! Love your stuff as well. Glad we connected :)@@TheoreticallyMedia
Miniatures is the answer here. There are about a million tutorials on here on how to create realistic props and environments. If you're creative you can buy doll props and other things.
I do love that handcrafted method as well. Spent some time as a kid doing model kitbashing as well. The amount of modeling glue I probably inadvertently inhaled might explain a lot about me…haha
Great video Tim and thanks for sharing!
Thank you so much!! Lots of cool stuff coming up next week as well!
@@TheoreticallyMedia , no thank you for putting in the research, testing and tinkering workflows (experiments w/ explanations), and overall creative curiosities that you share to your viewers. Sounds great and looking forward to watching them.
I used this to make my last video about automating my UA-cam channel.
Wait, which on? Moonvalley? Or Sad talker?
Do the final scene of The Good The Bad and the Ugly it has long shots where they all stand still and the camera pans around
DAAMMMMMIT! Where were you yesterday!!! That is SOOOO Right! Gah, I was wracking my head trying to come up with shots that would work-- I should kick myself for forgetting that one!
Great video! Only comment I'll make is the contrast on you seems a bit high. But dude I love your videos!
Oh thanks so much!! Yeah, I’m always trying to dial that color/contrast in. The room has a stupid skylight which is the bane of my existence. I love it, but the light consistency is always changing, so I sometimes go a little overboard in color correction. I’ll nail it one of these days!
I'm here because of your post on youtube, subscribed let's gooooo🔥
Runway's Greenscreen is the only reason I subscribe. It is just so good. One tip though, I stopped using the green screen output. I export the matte instead. I then layer it over the original footage in after effects, set it as a luma matte and it gives you the cut out, without any of the green edging you often can get from green screen output. It is just that little bit cleaner to my eye.
Oh, that's a hot tip for sure! And agreed-- I'm not over the moon with Gen-2 (although I do like Gen-1)-- but I think the real gem of Runway are the various "hidden" tools, like Interpolate and Motion Tracking. There's some real gems in there if you go digging!
@@TheoreticallyMedia I used interpolate on a music lyric video. Very trippy.
@@TheoreticallyMediathey can’t get out of their own way. They keep shitcanning their bots imaging to add front end bells and whistles without saving previous versions. I don’t think their dev team actually understands what it is people in this space want cause their too busy chasing money…and whatever pika comes up with. Its sad, their bot used to be so good.
We missed you Man :)
Haha, it's been a minute! I planned to have this up on Tuesday...and then...didn't...and then Wednesday I shot it and was furiously editing-- but didn't quite make it in time. I mean, I could have posted it at like 2am, but what's the point?
So, it was nice to take some time on the edit today. Got a couple of fancy transitions in there! Ooooh, crossfades and zooms! haha
@@TheoreticallyMedia your work is top level. All the best :)
HA! You pointed it at your screen! Hadn't thought to that! Probably will be new versions of films called "Splats" - prediction! I can imagine indoor VR parks where you can participate in a horror movie - maybe a use for real estate that is abandoned now.
Sometimes the stupidest things are the ideal lo-tech solution! And yeah-- the whole VR/AR thing is on the verge of exploding. We've tried a bunch of times (ahem, Google Glass...) and nothing has really stuck. But we're right on the cusp now....
I'm here for it! I'm Player One, and I've been Ready for YEARS now!
@@TheoreticallyMedia Lol. K, Timmy Rocket.
Fantastic content as always
Much appreciated!!
Wow love that bad decisions
great stuff!
Thank you so much!!
Great vid, 👍
Thank you so much! I've got some really exciting stuff coming up next week! (I'm not going to do that YT'er thing of like and sub...I promise!)
imagine if in the future (assuming cctv arent potatos anymore) investigators can use gaussian splat techniques on cctv footage of suspects or even victims last moments to find details they would other wise have missed, scifi i know but that immediately popped into my head
Very Philip K Dick for sure! I think the game Cyberpunk 2077 had a narrative mini-game like this, where you had to solve a robbery.
Of course, in true Cyberpunk style, the footage was streamed right into your character’s eyeballs…so we aren’t there just quite yet!
Great video as usual! Are you still dealing with that traffic bug that hit your channel? Is your Adsense back to normal?
Back to normal. A lot of the channels that were hit with it seemed to stabilize around the 28th. Like, A LOT of channels.
Meaning, something was fixed, although YT still denies it was a bug.
Sadly, there are still channels from the Nov 22 glitch that have never recovered.
All in all, I do consider it a really valuable lesson learned. Adsense is temperamental and shouldn’t be relied upon.
At the same time, I’m not keen on selling VPNs to you via a constant drip of sponsorships, so. I’m working on some ideas to keep the revenue up on the channel, while not being beholden to…
I’ll finish this response in a a second, but first a word from today’s Sponsor, Skillshare!
…ok, what was I saying? Oh yeah, so I don’t have to be beholden to reading a bunch of ad copy.
(Fun fact: Skillshare actually reached out to me a few months ago, and I was like, sure! And then they totally ghosted me…haha.)
@@TheoreticallyMedia LOL...that's good that you don't kill us with the sponsors. But people do what they got to do so I don't like to complain. That's good your channel is back in order. I don't even know what that whole bug thing was about, but man did it sound crazy. UA-cam needs to get it together.
I love the BD guys - they’re like the Milli Vanilli of UA-cam. I can’t wait for the big reveal when they’re exposed as the over enthusiastic AI generated avatars that they clearly are!
Come to think of it, I’ve never seen them and Joshua from Hey Gen in the same video.
Hmmmmm….i am now sus.
Hey my friend, another animation problem that you could speak of (please) is that the AI generated avatars don't seem to breathe, we only see their mouths, and heads move, but not signs of breathing, can this be handled anyways ? thanks :)
Oh, you know- that’s a really good point. Nothing I’ve seen yet, but I’ll keep an eye out!
I mean, just thinking about it, you CAN manually keyframe a breathing cycle in- but the whole point of AI is to automate those tedious tasks!
Tim compresses an hour of education into 10 mins - wow.
Haha, not taking shots, and not naming names- but I’ll never do a 30 minute video where I’m just reading stuff off a website,
…ok, maybe ONE shot fired. Haha
I appreciate you but could you either make a video or direct me to a video explaining how to use sad talker? I'm so lost on that step
Write the book that explains AI FILM MAKING IN THE WILD.
Haha. I’ll go full Werner Herzog and spend 3 months in a jungle documenting the filming of an epic AI film. 3 cast members will die during production and my lead actor will lose his leg in an alligator attack.
I will return to the US, only to discover no footage survived because there is no WiFi in the Rainforest.
Whats the difference between this and photogrammetry?
So, I am not an expert in this, but from a layman’s perspective: Photogrammerty is point based capture, whereas the breakthrough with Splats is that , well- it splats. It is “blobs” of captures that interlock. The really interesting thing is that you can get things like rotational surface reflection with a splat, whereas point based is “you get what you get”
Also, because they are blobs, computationally, it takes less processing power.
Considering the one button solution is probably only a few weeks away, I’ll let others learn the “hard” way.
Man we need to make a film together 😂
Great video by the way
Haha-- Dying to-- but keeping up with all the info for you all is practically full time!! (Although, I am thinking about sneaking a quick short in soon!)
BTW, I LOVE the SCP stuff! Its also SO perfect for AI Imagery!
"But so is real animation." We're getting close.
We really, really, are. I wish we could hop in a Tardis and show this tech to some Disney Animators back in the 40s...Well, then maybe we could all go on a time travel trip 90 years in the future and see where we are!
Gaussian splat ? What is that ?
The software?! What is the software to download? Jesus that's the only thing that matter and you don't share it!?
Ha! It’s a bit involved, and this wasn’t one of my tutorial type videos, more “here’s what’s going on”
I think the Bad Decision guys have a walkthrough up, I’ll pop a link in the description. It’s a bit involved to get splats running, which is why I didn’t dive into it