Why We Made a FULL EPISODE of an Anime Even Though We Aren’t Cartoonists | CorridorCast EP
Vložit
- čas přidán 28. 06. 2024
- Watch Anime Rock Paper Scissors 2 THIS WEEKEND. Live Stream Premiere Friday 8/4, On Demand Saturday 8/5 exclusively on our website. Try a membership yourself with a 14-Day Free Trail ► corridordigital.com/
Niko joins the podcast with Jake, Dean, and Jonah to talk about our upcoming release of Anime Rock Paper Scissor 2, our new process, and what to expect.
Check out David Maxim Micic, composer of the Anime Rock Paper Scissors Original Theme Song ► open.spotify.com/artist/0wQa1...
Instagram ► / corridordigital
Merch ► corridordigital.store/
Creative Tools ►
Puget Computers: bit.ly/Puget_Systems
Aputure Lights: bit.ly/Corridor_Lights
B&H Photo: bhpho.to/3r0wEnt
ActionVFX: bit.ly/TheBest_ActionVFX
Cinema4D: bit.ly/Try_Cinema4D
Nuke: bit.ly/Nuke_Compositing
Houdini: bit.ly/HoudiniSims
Octane Render: bit.ly/Octane_Wrender
Epidemic Music: bit.ly/Corridor_Music
Chapters ►
0:00 - Anime Rock Paper Scissors is Coming Soon!
2:49 - The Process
8:00 - The Art Style / Aesthetic
10:49 - Documenting the Process
15:06 - Surprising Developments
31.28 - The Best VFX Program in the World
35:18 - Does this eliminate artists?
39:45 - We're releasing tutorials! / The Art Form of Compositing
45:28 - Sam is a MONSTER VFX Artist
46:58 - The Newest Development
51:13 - What to expect. - Zábava
CHAPTERS?! FINALLY!!! 🎉
I’m glad you made a point to address the controversy around ripping-off Vampire Hunter D. I was never upset about training your AI on it, but I do understand why people were upset. Moving forward with your own style is definitely the move
People were upset because they love being negative, there not even 1% logic behind the controversy
@@techgoggles Agreed. The outrage around that was ridiculous. People were basically advocating for all artistic works from now and indefinitely into the future be already copyrighted by the works that inspired them. Like people didn't realize AI learns in a similar way to how humans learn. Man people are dumb.
@@techgoggles are you an artist? obviously not, you don't have any art work to steal, that's why you don't care
If you don't get the Flow Latent Blend Schedule correct you could cause a Resonance Cascade.
Using your own style images to track in on a style is a good first step and I'm interested to see how that goes, but we still kinda need to see a version where the underlying AI systems isn't trained with copyright art too. Right? That'll be the big test for it.
Unbelievably pumped. I don’t think Ivy’s ever been on the podcast before, do you think she’d be down to come on and talk about her experience on RPS2? Actually a “wifecast” would be funny, get Sarah and Jenna in here too lol
Holy shit, you got David Micic to compose your intro?! Can not wait to see this!!
Personally, I would love to see this process done for SOAD 3. While I love the visuals of the show, it would all mesh better I think if you put it through a similar or the same process
I’m glad that they actually decided to get their own style, I’m not a fan of AI art being used commercially with the whole “being yoinked” thing or whatever but I think I’m actually looking forward to this because THIS TIME it’s entirely it’s own thing, sure it’s style is still inspired by VHD but it’s not VHD, they hired a consenting artist to provide and be compensated for their work, having an artist’s art and expression thats being used with permission and compensation to make an experimental project for a potential new form of story telling fuckin rocks
100% when saw the first one i would lit my touch and got my pitch forks if they didnt use their own model the first one was barely a pass as the first one was a tech demo and not finshed item.
this what AI is for it cuts out the boring parts to focus on the key parts
48:50 - yeah... i tried the SDXL new model and it can definitely output some crazy good images. And once you feed it through the SDXL refiner model, it really polishes up the output. A lot of the existing tools still has to be upgraded to the new SDXL model as a lot of that is/was geared toward SD 1.5 and 2.x. But it's just a matter of time before they get things working with SDXL.
HOLY FUUUUU- David maxim Mimic is one my fav so hyped
Will episode 2 be released on CZcams after the website release?
"400 Sharts!" ...guess that one just kind of "slipped" out, huh?!?!? 😂
SAM is just BILL STECHNER from NARCOS in disguise - Tell me im wrong...
Enigmatic e is 🔥
Seems I have to go for Corridor Digital now! Then I can watch some DnD before Baldur's Gate 3 comes out as well!
Niko's analogues are on point😮
Screw people shitting on the ip thing, as someone who has worked in a vfx cgi studio we take reference copy style all the time.
Also most of the biggest creators pooping your work were mostly creators who just react on shit for a living.
Not on topic, but how are CG shots counted and what is a CGI shot? Cuz I have a hard time understanding, if a demo episode had 6 mins of runtime and only 80 CGI shots..
A CG shot is just however long is has to be until it cuts. There isn't really a defined amount of time in between cuts, and on the really long shots that don't look like they cut, most film makers will design hidden cuts into it where they can stitch two scenes together since shots become exponentially harder to pull off the longer they are. Even when the content is full CG its pretty rare for them to just use one big CG file because typically these CG files are so large that they sometimes can't even load all the assets in a single frame. Like on some of the big disaster movies with physics and thousands of people, they often have the background or the people rendered separately and layered on top of each other (comped together). So a shots length really depends. Some action sequences might try to hide the stunt acting by cutting multiple times a second, and some dialogue scenes might go minutes without being cut because there isn't much happening visually in the scene beyond the actors performance. Typically each shot has to be redesigned from set decoration to camera placement and lighting, which is why they put an emphasis on the number of shots.
love the work gents.
I would love if you had the content in the site with Brazilian Portuguese subtitles. There is a lot of people I would like to show the channel, but they don't speak English 😢
1) Turn CC on
2) Open the cog menu
3) Select Language
4) Select auto-translate
5) Select Portuguese
when is this coming out on youtube?
It's true that AI can be a tool for artists to be more productive and efficient BUT people have been let go in huge numbers being told to their face "we don't need you anymore, we can copy your style now"!
Really weird of you to act like this is currently happening instead of saying that it is something that will happen in the future. What you seem to be overlooking is that there will always be lead artists employed, its the younger artists that won't have a job. The people who create the artstyle that the younger artists copy, will still be employed and be in charge of creating art for the AI to copy. However this isn't all doom and gloom for younger artists because this AI makes it easier for younger artists to skip years of manual labor and jump straight into creating indie art/films/games.
Its directly comparable to digital music. An artist is basically the same as an Instrument player. Both of these people get paid to operate a tool. With the rise of recordable music the number of professional instrument players have dropped, and with the ability to recreate instruments digitally that number has dropped even more. However this hasn't killed instrument playing completely. There are still professional instrument players, the only difference is that only the best of the best can make a living just playing the instrument. The industry has shifted towards valuing producers that tell a story through the creation of the music, and a lot of those producers have experience playing an instrument. That experience helps them be better producers but its not a requirement anymore because music production has been made more accessible. Many new music artists aren't even making their music in a recording studio, because they are able to make it in their bedroom, and some even make it using just their smartphone.
The focus is shifting from who can produce the prettiest thing, towards valuing story above all else, and while AI can certainly help with writing a script it is unlikely that storytelling will eventually become completely automated. People typically like that human touch, even if that person is creating a shit product like mumble rap. Its the personality behind the music that the audience resonates with.
If anything this tech should increase the bar of good storytelling dramatically because creators won't have to spend years of their life on something on to find out that the audience didn't like it a year later. They will be able to iterate faster with AI and find out if their story works sooner than later. It will also increase the total number of quality projects due to lowering the bar for entry.
Only young people don't get animation jobs?
What happens when the old lead artists die and young people haven't gone into the field because studios only needed that one person?
How does the next old artist get made if there's no point of entry?
Also about the thing with musicians operating a tool: There is a lot more going into music than playing the instrument. A musician is also called an interpreter, a piece of music has to be interpreted by a musician to be heard by audiences. Maybe AI can create interpretations of music as well, but there is something very emotional and human about music. It is like all of the arts about human expression, and there will always be something special about art that was made wth that in it, and great when it is in every fiber of it.
Thr Corridor guys are doing a lot to get the AI to make sense, they put real artistry into that production to work - Had they not used the AI it would've looked different, but the idea and soul of it would've been there. But we can all tell there is a non-human part in there (im not even talking about quality or consistency, jitter or whatever), and I believe that will stay. I think that's also the same with CG characters and actors (in reference to the new video with Wren and Jordan) - No matter how good it will get, it won't feel right without a human behind it.
And all that to say: The AI movie made with a few artists will probably be fine and be fun to watch, but the one made throughout with the human touch and love to the craft will always be more special and evoke more intense emotions and thoughts in us
You guys should make a Corridor Original Fantasy-Adventure Anime!!! You guys are talented, you could do it!! 💯💯💯
they could totally animate some of their D&D son of a dungeon live action parts
I’m super excited that’s it we’ll be on there website because probably of copyright issues and what not and hopefully 🤞🏻 it’s good Hit I wish the best for them 😊
i would love to see jake play someone like a war tactician in the next one!
Cant wait to fisherprice this stuff
epic pod
37:14 I think one thing I’d wanna note is that the severe influx of bottom of the barrel content makes finding the good stuff harder, I do understand that my thoughts on this sound pretty elitist and glass half empty and I know that’s pretty bad, my thought process was if you give the ability to produce anything to people who never had the desire to otherwise, that ratio of good projects to visual chum is severely out of balance, you’d rarely get something of value or potentially the opposite as we’ve seen with all those AI art grifters on twitter that harp about how much they want artists to “adapt, elitist gatekeepers” and “just lay down and accept their fate of being stolen from all the time”, which is just a volatile mindset fueled by a software that’s entire business model is around scraping the actual image data of other peoples stuff without any regard for their consent. good lord… looking back on this I just made myself needlessly upset
both the AI techbros and artist are both have it wrong techbros think artist can be replaced but to unlock it full power you must have some baseline art skills. on the artist side they eat the bullshit that techbros spit out and then worry about think they are gonna be replaced so they try to smash the machines.
what im worried about is some techbros and working out if they study and learn art their work will be better but the artist are refusing learn the ai skills to use it. meaning some beat artist to punch and be too late for them to compete.
Ai to something not will it when and we need to adapted or be left behide
Welcome to CZcams 😂 no need for ai, that's the thesis against CZcams
Even more than that, this disbalance is self-magnifying, because it basically pushes you towards "quantity over quality" mindset. Why spend a month doing a thing really well and pouring your soul into it, when you could've done 5 things in the same amount of time with much less effort on each one, if in the end the good thing getting buried under the piles of newly released content just as reliably as a mediocre thing, but you can put mediocre thing out more often, increasing the summary amount of eyeballs you can get.
Even Corridor Digital are already a victim of this - their videos sometimes are full of jank and errors that honestly kinda ruin the final product - like their avatar videos for example - but they're "embracing" and rationalizing it because they _need_ to put out a video each week, because in this environment spending too much time on making your product _better_ ultimately makes your situation _worse._
And no, do not start on the "as long as you have quality stuff you'll be popular", that's a BS, I know tons of people that produce good content yet have barely 300 subs.
That enormous amount of garbage already exists. There are already more hours of video on youtube produced then anyone could ever be expected to filter through. Systems exist that help filter through to find the good stuff like reviewers, recommendation algorithms, aggregator sites. The problem will never be too much content imo, but not enough good ways to filter through to the good stuff.
This is the same argument I had years ago with the topic of game development. Same situation, the rise of free game development softwares with tools that are as powerful as what the leading game dev company uses makes the barrier for entry for video games really low. This led to a couple of argument that Unreal Engine should be made more expensive and exclusive. The result to this are quite obvious, quite a lot of small to medium size game dev company are now able to produce higher quality games. While there are indeed a lot of bad games that gets released, at the end of the day they get filtered by themself.
Been feeling down lately but this pod just made me feel so happy and laughing..... it's so cool to see cutting edge tech
At least they didn't use RevAnimated like everyone else uses lmao
Can't wait for the video to drop!
Hah team 182 .... next time !
so its gonna be behind a Paywall? NOO!!
Oh mate, that inventing a and training on a workflow process is so my jam.
I'm a software and systems architect, this is my voodoo.
😍
I think the tutorial being behind a paywall was more about tonal dissonance. the video celebrated the open-source and freeness of these tools and then said well if you want to learn them you have to pay up. that being said I feel like the calls to action in the first anime rock-paper-scissors videos was top-notch. maybe some extra framing would Elevate this already prudent business decision
Hey Corridor, quick question! Do you edit Jake to be at 14fps to represent that he isn't there, or is his internet really that crap? If you're editing it that way, like Spiderverse, absolute kudos to the crew, because it's super effective! If it's just Jake's crap internet...JAKE, GET BETTER INTERNET!
Spiderverse, for sure! 😂
whens da next episode
23:36 But can you do a cat?
A _quadruped_
Like you guys should talk about the south park episodes made with ai too, that's the writing part of the equation!
Can I actually do this with my friends or do I need a $5000 PC?
In cybersecurity we've been automating processes for three to five years now with products like SOAR. We take the same approach. Increase velocity not reduce labor. Our need for automation is driven by the lack of 750,000 people in the US alone who are needed to keep up with the labor requirements of cyber.
This is my favorite artistic thing made in a really long time. An actual movement happened, and you guys are in the spear head.
Dude I'm so stoked for chapter 2. Holy shit
Shouldn't it be called ''Rock, Paper, Cross Swords'' ?
You guys got David Maxim Micic?! Cannot wait!
Duggone I love this podcast
Quickist way to train is to use Ai Astria and you can use the models in stablediffusions 😊
Did you make the second part your site's exclusive because of the all the negative reactions you had received from the first one, or it's just how thing are going to be from now on - youtube is for corridor crew's and small experiments, and all the important big stuff is paywalled on the site?
business are not your freinds
at the end of it Corridor is business and needs to make money no shareholders thank god but they gotta make paywalled somthings.
business are not your freinds
I do not think the video itself is exclusive. The detailed "making of" is exclusive like the first one. They need money to make things. If I recall correctly from a previous statement from Nico, one person subscribing for one month of their website (4 usd, i think) is vastly more valuable to them than a youtube viewer because the ad revenue per view is so small. (I may be exagerating the numbers) It is like watching all of their videos 100 times is still less valuable to them than a single month subscription on their website (which again is only 4 usd).
@@mf-- Well, they keep talking that it will be on Corridor site and that you should go there to watch it. And I understand that it is a normal decision, that you need to earn money and in order to do that you need to entice people to sub on the site.
I don't really have a problem with that, I was a subscriber for a year or so, but - and I know that's a "me" problem - my country decided o do a bad third reich cosplay, so now I cannot use any online payment options anymore, and will likely never be able to use them again. 4 usd or 14 usd, I can't pay even one cent, even when I _want_ to.
@@DarthBiomechreally? Not even prepaid cards? That sucks man.
You guys are trying to turn live footage into 2d while I’m turning my cg keyframe animation into 2d. Essentially turning Stable Diffusion into a 2d render. Animating in 2d can be done faster than cg, the time consuming part comes in making it presentable by cleaning up the drawings, color, and lighting; depending on how much detail goes into the lighting. I’m lazy and don’t want to deal with all that so cg into 2d it is.
All of this leads me back to the fact that the one thing AI won't be able to do is be a creative writer. CHAT GPT and the visual AI art programs are good at copying style, but they can't do a compelling narrative, or make even a good original joke. They don't have a lived experience, nor do they know what to do when they absorb one either. They might someday, but for now, projects are only as good as the concept and writer.
First! Can’t wait guys
AI is not going to ultimately be responsible for job loss, corporations exploiting IP will.
Unfortunately people who are staunchly anti everything AI will still hate this. Tribalism and ignorance of the technology/process will allow them to just dismiss it.
If Corridor made the assets, they will have difficulty complaining about anything. Corridor's process documentation will be very informative. When the core software is based on misappropriated assets, it is easy to call foul.
@@mf-- No - they'll find a way. You'll probably see when it comes out and people act exactly the same way as they did with the first one. They've already made their mind up.
@@mf-- The next core argument would probably be "Oh but instead of needing 5 artist to create an animation now you only need 1 artist. Firing 4 artist in the process".
35:47 It’s incredible how much I disagree with this point. Look at companies are distributing content on streaming. Look at how animators are treated without this tech applied. Do you really think CEOs aren’t going to think of downsizing first? I’d believe what Niko is saying if there were parameters in place.
Love you guys, but I can’t be just optimistic about this.
And like they have also said, there will be more opportunities for people to make MORE art who can't afford it. It's just a matter of transition. The system is not working for artists anyway, and I'm pretty sure many artists out there would like to create art for themselves and be compensated for it rather than just be wage slaves.
It seems daunting because we are on the precipice of a drastic change, which will happen, with or without the said "parameters", but I totally want to see small creator collectives, singlehandedly using their own styles and stories to make entire tv shows or movies or anime on their own, instead of through Disney or Netflix.
@@billjack1857Too many bananas banana loses value
Everyone has banana nobody buy banana from someone else
It work both ways. The AI tech allows big companies to have a smaller team of artist but at the same time who's preventing artist to create smaller animation studio to produce what they are already making in the big company but with 5 people instead of 50?
Don't get sunk-cost fallacied into thinking that huge companies are the only way for artist to survive.
@@esra013 They are, indie doesn't make money
@@alexandra_avr5055 Well, then make banana bread or banana ice cream, etc. Innovate and get gud.
Also, indie doesn't make money YET. If AI can remove the impediments for filmmaking and art that only big studios can afford, we will totally see Star Wars like media from smaller artist collectives working from a garage or something.
I am become death, destroyer of animation
Ai gain of function research like this will only lead to bigger problems than the ones it helps solve.
Sucks to see you continued down the dark path that made me lose interest in the work you guys do, sad
Sucks to see them advance technology smh
Are you going to create a new version of what you did?
I need help from someone with your guys's expertise on video files. I don't want to explain it here in the comment section. I don't know how we can contact one another or correspond outside the section or I would've emailed . GoPro tech support didn't help me. I know you guys can thank you for any response
The phrase "We made" is really doing a lot of heavy lifting here
I mean, did they…not? They film, edit, and composite it. So about 3/4 of the work was…made; of their creation and input (?)
love the work gents.