I dont think they need to worry too much anymore than pop stars need to worry about vocaloids. Its also great for low budget indie developers who cant pay for voice acting
Actually, for the first time in their careers, they're gonna make real money lol. Remember, Sonatic plans on having a marketplace of voice models. So this is passive income for voice actors
@@tstone9151 How long before they skip that step and jumped into completely generated voices? they could easily mix all the voices and generate something else completely new in their neural network and avoid having to pay humans
There are so many indie games that can't afford to hire voice actors. This would make a lot of games way more emersive if it was easily accessible. But I'm guessing it isn't as easy as typing in lines and getting and audio file ready to plug into game maker/unity
After spending hours playing Deadfire and Outer Worlds consider my mind boggled. I have been very impressed with the quantity and quality of voice work in Deadfire but now I'm questioning reality...
WTH???? This is absolutely crazy!!!! They solved the biggest problem that AI Still had so far : monotony. They solved it big time! Two minutes paper better cover this! what a time to be alive!!!
I would use this technology in open world games for NPC's, while the main cast of characters are voice acted. This would be really really immersive for a game like GTA 6 for example.
The VA industry is already who you know not what you know, they’ll blacklist companies who use non-union actors so basically unless they let you in you aren’t working on anything unless they let you in, which they won’t and which is why the same 50 people voice everything.
@@songbirdsandsandwiches8217 this is entirely bullshit. there are so many smaller projects and groups always looking for voice actors regardless of being in any union. where the hell are you getting this from? do you have any clue how many indie projects dont use big name VAs???
3:55 -- "... and they have a really consistent passive income..." Can't be too consistent if smaller studios and independent creators can't utilize the service. What is up with that???
What I think that would improve even more this technology was emphasis numbers for nouns and verbs. The strongest words would have higher numbers on the sentence, so it would have more emphasis. For instance, "KILL" is a heavy word (n7), but "BLOOD" (n8) could be heavier on the same sentence. It would be awesome if there was some kind of priority on emphasis, so the curve would not be linear. The last words on a sentence should have priority too. Like "I'm gonna kill you and drink all of your BLOOD". (Blood itselft is heavy and its the last word, so it takes priority on a dialogue) The verb "Kill" should take weight priority over "drink" (n5) too: "I'm gonna KILL you and drink all of your BLOOD". In another example, "DRINK" would take priority: "I'm gonna DRINK all of your blood!" Depending on the style, those words would have a diferent effect. It could be louder (anger) or slower (ironic) I'm a screenwritter and I studied programming for a while, so I think that's viable. Of course you can't give a number for every single word on planet, but it should have a neutral default number, and a bunch of other strong words that you would need to store in a database with an Intensity scale, let's say, from 1 to 10. Every word could be 1, except for those you design.
@@KosmiskRevolution that's right! But instead of doing this for each phrase, it would have a mathematical equation and a database for a LOT of words and combinations
I'd love to see an overall inflection control, so you can make voices go up or down in different parts of the sentence to create different emphasises. Though if it's only for prototyping then I can see why this wouldn't be highest priority. I actually see a lot of value in this for lowering the budget for fully voice-acted content, as well as install sizes. If it can be generated on the fly then you don't need to store all those sound files, and while it'd be best to keep all the key characters of an RPG fully voice acted, the potential increase in scope you could have in using AI voices for random pig farmer #4296 could mean writers and quest designers could go nuts with the stuff they make rather than having to constrain the scope.
This could help with games/animations that are set in various locations and/or have a diverse cast of characters like Sniper Elite and Black Lagoon. Still, this type of tech is something that'll need to be kept an eye on as it only gets more and more advance.
There have been leaps made in AI voices in just these two years, this is most likely the future of video games, imagine chat gpt style interaction with fully ai voiced characters in video games like candy crush or tetris, imagine the possibilities!
It's actually not. 15ai is an algorythm which analyzes, restructures and vocodes samples (and recalibrated by a human) Sonantic and oldschool tts use voice actors to pronounce english phonology (imagine syllables) This video explains how it works: ruclips.net/video/un00hr4Si3I/видео.html
This is great I don't want to hire certain people just for a voice actor. None of us want to go to prison for stuff we didn't do and I'm not rigging the whole building with security cameras and mics just so we can be around that person without being accused. I hope this gets better though it'll help us out later.
This could be very useful for low-budget procedurally generated games with dynamic stories.
Even high budget games can use this feature, voices are quite convincing
This thing could revolutionize indie game development, like unreal 3 back then.
adding this to AI dungeon would be savage
This could be great if they can implement in different languages, we can finally have dub for much more languages than now
Modding, too! Goodbye silent characters or terrible webcam VA, hello equal playing field for small creators.
I have the feeling this is going to make a few voice actors anxious.
I dont think they need to worry too much anymore than pop stars need to worry about vocaloids. Its also great for low budget indie developers who cant pay for voice acting
@@imbored6724 For now.
@@imbored6724 Yes I agree because technology never gets better
Actually, for the first time in their careers, they're gonna make real money lol. Remember, Sonatic plans on having a marketplace of voice models. So this is passive income for voice actors
@@tstone9151 How long before they skip that step and jumped into completely generated voices? they could easily mix all the voices and generate something else completely new in their neural network and avoid having to pay humans
There are so many indie games that can't afford to hire voice actors. This would make a lot of games way more emersive if it was easily accessible.
But I'm guessing it isn't as easy as typing in lines and getting and audio file ready to plug into game maker/unity
we just need to wait for software that you select tone and emotion and you just type the sentence and boom your done.
@@haroldliriano5394 tone and emotion
There's also games where the sheer quantity of dialogue would be too much for a human to get through.
@Miles Doyle CHRISTIAN BOT DETECTED
@@haroldliriano5394 here we are today and that’s completely possible
I was looking for good-quality AI voices, but this is insane!
Now AI can use crying voice to manipulate humans
Alexa: Hello Dave, I think you should buying more stuff at Amazon... pleaseeeee....
Imagine the advertisment potential!
...scary.
"Screamers" 1996 movie, kid asking: "Can i come with you?"
After spending hours playing Deadfire and Outer Worlds consider my mind boggled. I have been very impressed with the quantity and quality of voice work in Deadfire but now I'm questioning reality...
Obsidian uses it for drafting, then have real people record the voice acting later.
WTH????
This is absolutely crazy!!!! They solved the biggest problem that AI Still had so far : monotony.
They solved it big time!
Two minutes paper better cover this! what a time to be alive!!!
Yeah, this is so amazing!
I would use this technology in open world games for NPC's, while the main cast of characters are voice acted. This would be really really immersive for a game like GTA 6 for example.
just get real actors.
This was 3 years ago, holy crap!
welp.there go my future voice actor dreams
The VA industry is already who you know not what you know, they’ll blacklist companies who use non-union actors so basically unless they let you in you aren’t working on anything unless they let you in, which they won’t and which is why the same 50 people voice everything.
@@songbirdsandsandwiches8217 this is entirely bullshit. there are so many smaller projects and groups always looking for voice actors regardless of being in any union. where the hell are you getting this from? do you have any clue how many indie projects dont use big name VAs???
3:55 -- "... and they have a really consistent passive income..." Can't be too consistent if smaller studios and independent creators can't utilize the service. What is up with that???
What I think that would improve even more this technology was emphasis numbers for nouns and verbs. The strongest words would have higher numbers on the sentence, so it would have more emphasis.
For instance, "KILL" is a heavy word (n7), but "BLOOD" (n8) could be heavier on the same sentence. It would be awesome if there was some kind of priority on emphasis, so the curve would not be linear. The last words on a sentence should have priority too.
Like "I'm gonna kill you and drink all of your BLOOD". (Blood itselft is heavy and its the last word, so it takes priority on a dialogue)
The verb "Kill" should take weight priority over "drink" (n5) too: "I'm gonna KILL you and drink all of your BLOOD".
In another example, "DRINK" would take priority: "I'm gonna DRINK all of your blood!"
Depending on the style, those words would have a diferent effect. It could be louder (anger) or slower (ironic)
I'm a screenwritter and I studied programming for a while, so I think that's viable. Of course you can't give a number for every single word on planet, but it should have a neutral default number, and a bunch of other strong words that you would need to store in a database with an Intensity scale, let's say, from 1 to 10. Every word could be 1, except for those you design.
So, what comic book writers do when they bold for emphasis in speech bubbles. I can see it!
@@KosmiskRevolution that's right! But instead of doing this for each phrase, it would have a mathematical equation and a database for a LOT of words and combinations
I'd love to see an overall inflection control, so you can make voices go up or down in different parts of the sentence to create different emphasises. Though if it's only for prototyping then I can see why this wouldn't be highest priority.
I actually see a lot of value in this for lowering the budget for fully voice-acted content, as well as install sizes. If it can be generated on the fly then you don't need to store all those sound files, and while it'd be best to keep all the key characters of an RPG fully voice acted, the potential increase in scope you could have in using AI voices for random pig farmer #4296 could mean writers and quest designers could go nuts with the stuff they make rather than having to constrain the scope.
hmmm i think those annoying scam calls needs to add more feelings
Even as a traditional writer books, I can see the potential for testing out your dialogue for realistic speech patterns.
This could help with games/animations that are set in various locations and/or have a diverse cast of characters like Sniper Elite and Black Lagoon.
Still, this type of tech is something that'll need to be kept an eye on as it only gets more and more advance.
This is incredible but sad at the same time
Only 667 subscribers? I watched your stuff months ago but I'll subscribe now since I'm revisitng?!
I am waiting for this for years
There have been leaps made in AI voices in just these two years, this is most likely the future of video games, imagine chat gpt style interaction with fully ai voiced characters in video games like candy crush or tetris, imagine the possibilities!
This is amazing!!!
3:19 AI proceeds to have existential crisis
That narrator is a human though, right?
Is any open source program that can do this out there? thx.
This seems similar to 15ai. It's cool to see this tech develop!
It's actually not.
15ai is an algorythm which analyzes, restructures and vocodes samples (and recalibrated by a human)
Sonantic and oldschool tts use voice actors to pronounce english phonology (imagine syllables)
This video explains how it works: ruclips.net/video/un00hr4Si3I/видео.html
After playing Oblivion back in the days. That's amazing!
excellent so the players character can be used instead of "chosen one"
hi :) thx for the video :) Where can i try the software with my own words please ? best regards ?!
I haven’t played games in 10 years and never will again, but this would be good for audiobooks though.
Never will again you say. May I ask what your decision to abandon them was? I am aware of obvious reasons just curious about your personal ones.
@@johannesschmitz6370 can't blame them, games these days suck so much, with very few exceptions
By any chance there might be SDK for this in the future for people to use it like Unity?
whatsoftware are you using
How do I download this?
any intel on the pricing?
What is the pricing
Why u have so little subs?
wait what's the software they use
It'll be amazing if they use this to redesign Stan Lee to the Marvel films.
That would be great!
you're weird
Plot twist. This video has been voiced by an AI.
so using only ai in a video game means almost ANYTHING is voiced right?
Voice actors: "Fuck"
How is the progress on AI generated sound effects?
Well done to both teams. Love this.
RIP Matthew Mercer and Co.
They all probably make WAAAAAAAAAAY more money with Critical Role than they do from directly doing voice work.
Troy bakers gonna have An NPD meltdown
2:51 ''Σo, what does this technology mean, for the future of studio and actors?"
For the first, inferior budget. For the second, unemployment.
So, what happens to a voice actor when they die? They keep on working!
It's still a bit robotic
Imagine mixing it with chatgpt
Obsidian uses AI voces because its cheaper than paying Voice Actors
Wait outerworlds uses ai voices?!?!
Muszę to mieć do szkoły JoJo i sąsiadów.
perfect.
Every job imaginable is being automated. Even voice actors are on the chopping block.
Why is no one pointing out how morally wrong this is?
This is terrible.
this technology eventually will take jobs from humans
This is weird
WOW
we want real actors
This is great I don't want to hire certain people just for a voice actor. None of us want to go to prison for stuff we didn't do and I'm not rigging the whole building with security cameras and mics just so we can be around that person without being accused. I hope this gets better though it'll help us out later.
what are you talking about with going to prison? dude most VA is done online
Hmmmmm I wonder
Gotta get rid of those jobs
imagine this in mods, like GTA San Andreas one's.
Now Skynet is going to sound like my waifu!?!? I'm doomed!
Solutions to problems that don't exist.
exactly
They literally explain the problem which exists that they used this to solve.
Not a gayyymer I see.
Tell that to indie devs with no money to pay voice actors...
For an indie dev like me, i can't pay a lot of voice actors LOL. This is gold
signed up to spite 15.ai
more and more shtt is coming out they are getting phased out lmao its great
Yeah, unfortunately it’s dead rn