This is the best VOICE CLONING program (a little scary) - Descript AI
Вставка
- Опубліковано 4 сер 2024
- TRY IT YOURSELF 😱👉: www.descript.com/?lmref=SVM3Eg
If you hate how much time it takes to record and edit videos, this program turns text into speech using YOUR VOICE (not Microsoft Sam).
All I did was feed it 30 minutes of my voice sample, and in 24 hours I had my own AI voice...
Please SUBSCRIBE and SHARE with a friend - I write better than I type, and I imagine a lot of you would prefer making videos this way.
COMMENT below - What are you going to use Descript for?
HOW TO GET 100K TIKTOK FOLLOWERS: • How to get 100k follow...
ELECTRIC UNICYCLE BEATS RAPID BUS: • 50 km/h Electric Unicy...
0:00 What we'll accomplish
0:26 My first reaction
0:55 Text to speech
1:17 HUMAN VOICE
1:39 AI VOICE
2:03 Pasting text for the AI voice to read
4:44 How to set up Descript Overdub Voice Training
7:04 Where to download Descript
8:16 Top comment of the week ft. @Klos - Наука та технологія
This is great. I'm a do this now in case I ever go mute
Stephen Hawking be like
That sounds like an interesting backup plan ;)
almost good enough for videogame voice acting, can you imagine?
for instance:
30 - 50 voice actors,
a quest engine that recombines elements to generate quests on demand,
in vr,
voice triggered dialogue options,
the possibilities are astounding.
indie game revolution incoming
@@lucien-lu I'm working on it. building up my skills, starting with an educational VR app, once I get that out of the way, I plan on using some of that money to get a VERY open world rpg started.
for anyone reading this, do you have game dev skills and extra time on your hands?
replica studio has all your video game voice lines covered. You should definitely check that out.
Speech to text, muted and play - make amazing dnd sessions
You can finally get the voice actor to say your actual character's name.
What we need to do now is have a community where we each upload our training voices to each other so that we can train our own programs on those voices, thus use different voices for the characters we make for our stories, games. Thoughts?
that's it!!
lol, uberduck.ai community is exactly that discord.gg/CXgUnNgX
@@soldierscaptain6026 sorry but not really. DeScript is on a level of their own. When it comes to voice cloning quality.
No way,I will use this for everything that needs my voice,like personal podcasts and videos,and things like that for school.I have to thank you a lot,for this video.That was amazing
Gotchu Bella! Think you'd slay it at school with this
DUDE THATS SICK
I know right??
@@lucien-lu imagine if you could use this to copy the Siri text to read voice on iPhone and iPad
Wow. Just wow. Scary and fascinating to think about what's possible in the next 5-10 years of AI development.
Great content dude!! Subbed! Love your style
thanks man!!
I got lulled into a relaxed state by an AI.
No artificial homosapien
Thank you, ill give it a try
I can tell there is a future for your channel .
wish you all the best
Bless you Anas
Subscribed and liked bro 🤟👍
Bless you bro
Cool brooo!
yeee man
Voice cloning is excellent for preserving actors' voices and stall the symptoms of presbyphonia. VAs won't have to scream/growl themselves hoarse for action sequences and the like.
hahahah yee
This is amazing subscribed
Bless you bro
I'm gonna use this to copy the kurzgesagt voice and make it my voice if I ever get to put my conciousness in a robot body.
That's the spirit
Thank bro imma use this in my class lmao
It would be so great if there was a cheaper version or like a pay as you go version for the Text to speech feature
Think you can do pay as you go...
Since it's $12/mo for 10 hours, $24 for 30 hours...
Maybe just cancel when you don't need it. It's what I do!
www.descript.com/pricing/?lmref=SVM3Eg
They need this for games where you choose your name
I know right
Very nice your video... Im subscribe in your channel!
Thanks for the sub!
Not a fan of my voice, wondering is there a library of other voices and accents to choose? Haha.
Is there any AI software or program that can generate a voice clone from recorded mp3 or wav files? My dad passed away and I'm trying to create an interactive character on unreal that will use the text and voice generated from those mp3 and wav files. If you could also point me in the direction of some tutorials on how to program a two way conversation like Google AI has it would be helpful 🙏
@Lucien Lu is there an way to set up this kind of thing to auto change UA-cam videos to translated voices of other languages . so an English video is perfectly rendered with same voice but speaking in Spanish for example ?
not this one, but i think you'll like the next one I introduce
@@lucien-lu where ?
Hey man, Thanks for the video. Great work. Do you know if I'm able to download an mp3 or any other type file onto my PC the AI voice recording with the text which I wrote and then use it in any video editing program? And also which version should I buy for just this purpose? They have one for $12 a month and one for $24. I would use just this text to speach option, would it be ok with $12 version?
Yes the free version should do allow you to try text to speech.
Convert your video to text, then take that text and put it into something like murf.ai/
This is if you don't want to pay for Overdub on Descript.
Awesome
Bless you
Is it possible to edit the playback audio sounding like the original recording and less ai like? There’s something I want to test out music wise but the ai sounds are a put off
You can train it more, I stopped at the minimum
ima feed it glados voices
Thanks for the video Lucien. How many minutes can I convert to WAV audio per month? Can it be used to produce audiobooks? I would appreciate an answer. Thanks.
10 hours per month for Creator
30 hours per month for Pro
Sorry for the late response!
L
1:01 bro I just realized, your keyboard sounds like a Mr. crabs when he is walking.
Hahahah I stopped using it, was so loud 😂
Holy moly
ayyyy
Ok, so this is how we clone Morgan Freeman's voice.
If he'll let you hahah
This, and yes we need this.
I have been told by many of my friends/family members that AI is the next big thing to invest in. and this video gave me one of those lightbulb on top of your head moments. can you imagine a voice generator being so good it can produce the same quality voice of deceased rappers/music artists? sure its gonna be dead wrong to do so at first, but i dont see how the world doesnt catch on with this and eventually says fuck it and creates new hit songs from these previous artists. just a thought. I mean you see CGI in those movies like star wars where they brought deceased actors back and they played their role again, so what would be so wrong creating new tracks for these people? i maybe on to something here.
that's it... voice IP tho
it already exists, I've seen people use juice world's voice to make new songs
The future of Terminator is getting near this day
heheeee yeee
I have a question. When performing voice cloning in descript, can all languages be used?
Think just English for now but I'm sure they're working on it
They have a suggestion box and active forum
Ok intonation is getting there Now I want to know how it would sound if Michael Jackson would sing After the love is gone EW&F.😃
Imagine the meme shit we'd make...
Very interesting and helpful bro, thanks!
thanks man1
I imagine some blind person is going to get a bit experimental and use output from his old braille n speak as the source audio, just so he could have some higher quality simulation of the votrax sc02.
Damn this is the future
I wonder if anyone else Who read this comment knew what you were talking about aside from me? I mean that in the most nostalgic sense. I remember the Braille N Speak. Had one of those back in the day and thought it was the coolest thing. Talk about a robotic voice! I chuckle whenever I hear someone saying that Siri or Google sounds robotic. They haven't heard robotic until they hear that old synthesizer voice.
@@lucien-lu Technology has definitely come along way since the days of synthesizers including decTalk and Flex Talk. Both of those were definitely more robotic, though they were much easier to speed up. (Relevant to those who use a screen reader or who relied on reading via text to speech.) this is pretty good, though if one listens very closely, one can still tell that it is mechanical. The pronunciation of many of the words are slightly off, there is still lack of inflection, and not many pauses that would occur in natural speaking.
I'm trying to synthesize Billy Mays' voice. How do I do it?
Hahaha you'll need billy mays to read 30 min of wizard of oz
@@lucien-lu This is a good question. Imagine if you could get high quality recordings of a voice. Then be able to feed the translator a script that matches the recordings. Obviously a feat of programming but sounds feasible. Then you could get Freddy Kruger or Mr. Howell or Patrick Starfish or Floyd the Barber....or Billy Mays
am i the only one who use this for roleplay purposes because i cant find a sample for my FC voice?
ahhahaa that's pretty cool
am I in the future ?
thanks
No problem
Is that like Adobe Voco? Does it work with singing acapellas?
Not yet but maybe one day!
@@lucien-lu Hekp meeeeee I work with acapellas help me change words to acapellas heeeeeeeeeeeeeeeeeeeelp Dx
_Is there any way to do this in realtime vs tts?_
I haven't seen anything faster!
Greate video guy. But i see that AI can't generate voice same your tone, right? For discussion, I think Descript just is only TTS.
Think this is only the beginning :)
Excellent video. Please let me know if you know a similar Spanish based tool
You look like PSJ 💕
bless you ♥
Is there any probability to have background noise at duplicate voice clip ?
I imagine it'd be as much background noise as when you recorded the training text
Can you use celebrities audio and change the words?
that's uberduck.ai/#mode=tts-basic
does it has a Chinese version of voice AI generator? like mandarin? thanks
no it doesn't but i think you'll like the next one i introduce
I want a program to reproduce the sound in Arabic, please tell me 🙏🏻
Great informative Lucien
Can you please tell me can i use this voice only to make UA-cam videos !!!!!
Yes you can!
@@lucien-lu and monetize them with no voice based issues?
@@deck_2681 you can use your own voice and monetize
You sound exactly like Keanu Reeves
I appreciate that
can I download the result of the AI mp3?
yes, i believe you can export as mp3 or wav as well as mp4 with an audiogram
I cant find the setting.. is it you used overdub?
yes overdub is a paid feature
i saw this program but its behind a paywall so i never got to try the voice cloning experiment for my memes :(
well worth it
@@lucien-lu yeah when i get the money lol
If you can't detect the differences between the human example and the AI sample, I recommend visiting the nearest ENT doctor.
How much training you gave to Discript ?
only 30 minutes!
@@lucien-lu 30 min! tbh - if i won't look at the video - i won't be able to tell which is real and which is AI Lucien Lu :D
@@MRPtech hahaaa I'm about to get reaaal lazy ;)
began the clone war has
The memes are coming
How long should you record your voice
30+ minutes
So believe it or not they actually use this for some radio hosts just in case they aren’t in the room or at work
now way eh
What a the minimum required PC specs?
Think it's browser and web based. You could probably run it
This is scary 😂
😂😂😂
What if i want to clone someone else voice , like a cartoon character. i don't want to record my own.
you'll have to make them read 30 minutes of wizard of oz ahhaa
@@lucien-lu welp... Shit.
uberduck.ai is exactly that
Is there a way to do this in your browser? It says I have to download the free software but I don't have a computer.
you can login with the web app, but you might want to use a computer,
Can we save the recording after entering the speech?
Even if you can't, just record your computer's sound with any audio recorder
yes you can export the WAV file
@@CraigToddOfficial thank u so much
@@camillakim8717 Anytime Camilla 😉
Is it possible to use several languages in descripts ??
help.descript.com/hc/en-us/articles/360042640331-What-languages-does-Descript-support-
It does give you a slight dutch accent, can anybody else hear it?
yeeee
Does it work with different languages ?
just english for now
Would you be able to upload audio if someone deceased and use their voice
not this program but maybe others
So they now own a version of your voice ... no bueno!
might have to livestream myself at all times
Can I do this using a mobile device?
in browser yes, but i wouldn't recommend it, works best on PC
Is it in other languages too?
not yet
i hate this cuz i cant even install it...it says my conection fail...do anyone have the app for download ? cuz thats my only problem....i need the .exe...ty
Https://Descript.com should be easy to get the exe
@@lucien-lu ty mate, the problem is , the downloader.exe they give, doesnt works cuz it says that I have a problem with my conection...gonna try from a friend's pc to save it on a usb
the idea is neat but at the very least descript isn't there yet. to many artifacts and the sound is off. Really looking forward to advances. Every book will be an audio book then ^.^
Yeah it's good for filling in not so much for replacement
I have a question can you download the voice?
Not at the moment, I think it lives in the program.. but you can delete it!
Okk
So basically, this is text-to-audio using the AI clone of your voice. I wonder what audio-to-text you were using for transcription. Because, maybe having both is awesome, I guess?
Also, there is no way I can use an already prerecorded sample to make this work, right? I mean, I can't just use any old sample or audio recording with my voice for the AI to use as baseline.
it's both! they're both features in Descript, you can try it out www.descript.com/?lmref=SVM3Eg
Do you have AI for David Attenborough?
hahahha not this one but my next video will test a new AI program where you can use other voices like Dvaid Attenborough!
@@lucien-lu sure if u could do that, we would love it😀👋👍
Dose have emotional voice?
it's decent
How can I get your voice in descript ?
new video coming out you can use other voices
Can I change the AI voice for someone else?
that's coming up in the next video :)
little scary? very scary
Dw everything's going to be alright
:x and this well speak
yee
Do you really need to read what's written while recording, or can you say anything.
You can say anything. The only time you need to read their text is when you're training the Ai program
Can you set the intonation?
no but i think you'll like the next one I'm reviewing
@@lucien-lu thanks👍👍👍
Hi my name is starts making eminem song with own voice
does this works on mobile
?
no not yet, but a similar program called otter.ai/ does
@@lucien-lu thanks a lot
Is this free or cost?
There's free, I bought the paid plan for overdub ai
😁😁😁😄😁😁
It still sounds robotic
It is starting :D half glass empty or full?
Dude. Please send me the voice you just created
bro that's my voice
if you're here from wayneradiotv: same
𝙲𝙾𝚆𝙱𝙾𝚈𝚂
Can I get u to do one for me I'll pay u
new video coming out you can use other voices
Fuckin AI 😱😲😯😯
ayyy I :)
Not good enough for what i want, and arbitrary voice sample that can take distorted game voices and voice clones with recognizeable results.
overdub feature is not for free. it costs 30$ per month. (7 days free trial only). thanks, but no. i'll better try to search some free altarnatives.
correct, there are free alternatives
You need to put a new one on your list here it's called Supertone. It was developed by hybe Corporation formally Bit Hit.
Which of course is BTS label now Hibe. Which now Ariana Grande and Justin Bieber are now part of as well.
The technology is being developed for live performance but it has all kinds of shady implications.
Hibe of course states official ethical uses and people should use it only for that. But just like deepfake it can be abused.
You could blackmail somebody with this technology. Use it to impersonate leaders or law enforcement among other things. :/
Or use it for good because humans are becoming more and more optimistic and positive??? :D
@@lucien-lu I wish I could share your optimism lol. Let's hope you are right. :)
its dream smp time bois
Anime character copying intensifies.
oh lawd
this isnt even that good I have it
Early stages
@@lucien-lu yep it still could be better
@@MrEccentricities you're right
Screw this. I wanna deepfake an audio. not a text.