This video is amazing! I was trying to understand how the Heart on My Sleeves song was generated and I have watched a few videos, but none of them explained it as well as yours. Thanks for making it. Would like to see more AI + music videos like this!
Great video Sam, thanks for taking the time to put it together. Very clear explanation and as always, love the Smyers magic on those backing tracks. Your talent can never be reduced to an algorithm.
@@dncviorel I agree. With a slider modifier you could have ultimate control over your voice profile. Artists could lease out their voice likeness, adding yet another way for them to capitalize on their artistry.
great vid bro! I definitely think the more you can get it to sound like the artist your tryna convert the audio to you will get a better result, even listening to the bad romance one the results wasn't that bad especially knowing that it was a guy singing😅
I am not getting my demo version once I convert, everything is working great but I am not getting the final file with the artist model, any explanations?
The Setup 1, it tries to run the process for a few minutes and then I get an error and it keeps asking me to Restart the Session? Any idea how I can get past STEP 1 😅??
YES. We were just thinking this. I have a whole album full of finished stuff. One i would like a voice with 75% Cornell and 25% Bowie. How can I do this!!!I think you are looking for a locksmith that can do it inexpensively, is that right?
I am getting an error message when I try to run the "(Optional) Extract .zip Downloads - Step o.2." It gives me an error says "AttributeError: 'NoneType' object has no attribute 'filename'." Anyone got the same issue? :(
All seems too technical for me, is there no program that you can just drag your audio file into and it can make it sound in the style of a chosen singer?
I don’t want to pretend that I am someone else, I just want to be able to sing a song and have it come out sounding good so I can do my own references without having to wait on songwriters and reference singers. I can’t sing at all tho. Have any suggestions?
Just imagine spending your life recording beautiful music (eg. Brian Wilson), and then in the last week of your life you discover this for the first time...knowing that everything you've ever tried so carefully to protect, a legacy...will forever be tainted, the voice you once had no longer yours and now 'owned' by someone else :(
i don’t think anyone’s legacy’s were tainted by this. those artists have talent that is irreplaceable and a random tiktok user with a very unrealistic ai rendering on their voice will not be able to make the same level art the original artists can
the good thing about the ai is when replacing rappers that are know for their writing they can’t make the same quality because they aren’t top teir song writers like those artists and ai has a hard time replicating the best singers because their voices are so nuanced and powerful. so unless it’s an artist that uses a lot of autotune and makes music that is very easy to replicate and the writing isn’t that good they won’t be able to make a convincing ai song pretending to be them. and the ones that do all of that aren’t making good enough music to the point where it matters if ai users are making music just like it because it isn’t valuable art anyways and is already commonly copied
@@bloodclaat i'm assuming they mean 'i can't sing for s**t, can i sing and it also fix my bad singing', the best results of the AI ones are based off an already good vocal albeit it by someone else...but still a good singer for the input.
Anybody else follow the instructions at 6:48 but it doesn't output a file and instead just says True load model(s) from hubert/checkpoint_best_legacy_500.pt and nothing else... any idea how to get an output?
Not sure if I am stupid or something here. But what am I supposed to do on the collab? I run the stuff, I press rerun, get errors. Then stops. So somewhere in the process I supply my own WAV I guess, but whats all these steps really doing and needed for?
drake watching this right now punching the screen. who would have thunk it that we would be here with ai and he's this easily duplicated.. He better get ahead of this because it's moving without him.
He is ahead of it, he’s rich lol , there will be ppl who will never make money off their music cause of AI, at least he’s made his already… everyone will be using ai to make music all around the world, we will have millions of hits songs a week and no one but the person who typed in what they wanted and pressed a generate button will be needed. Which then eventually ppl will appreciate a human creating something cause it’ll be genuine and natural. So maybe I’m wrong but i have a feeling ppl will start calling themselves musicians just for typing in a description of what they want to hear and ai creating something good. They’ll say “listen to my song” and thinking they are a creator
2 weeks ago, I saw a video tutorial on Vimeo on how to use Google colab to make an AI cover song and it was the perfect video talked about it, even the link was exist there , contained tons of the artists and there was an option to change a certain artist from a menu, why no one share this Google colab link which contains all of those artists. I can't find this video anymore 😭😭😭😭😭😭😭😭. Please help , why you guys all your Google Collab links contains 4 artists only 😭😭😭. I'm confused and stressed because of that , I might commit suicide, I had a plan to make advantage of this, can you please help
hey man, i understand your stress on all this ai stuff, i also was in a time a place where i wanted to take advantage of this entire ai program before people would figure it out, but it dosen't mean you have to commit suicide! I promise!
Imagine you using a clone voice, then it became a sucess and then the drakeswitftgrande monster comes and eat all your cake? No. don't do it, we need a tool that makes a beautiful talented Jane Doe voice using the skills, not the tone.
Can I use this to create a Dolly Parton voice model from some acapellas? Or does this site only let you convert your acapellas to existing ai models, Drake etc... Thanks in advance
I've been playing around with making song covers BUT by taking mps of songs and using vocals removers to extract acapella singing. But I've fond it's pretty rare for the AI to sound good all the way through the song, the voice always seems to crack and I'm guessing its because extracted vocals are never 100% clean. Anyone got any tips?
im on the very last step after uploading the audio and i get this error " ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject" I have no clue what to do
I think like you are showing this to the people to make them learn how to do it, one day you also will be stolen & you will not find one dollar on your pocket to stay alive & eat, so my advise to you take care because what we do to some people one day it turns & return to you again, I am sure you know that, there is nothing better that you make effort on the things you create by yourself but trying to leave on the what other do this very bad, you as Musiian too AI will make you stay home too not finding work don’t worry not very far, as you learn people to steal you will be stolen don’t tell yourself I am showing people technology no you are learning them to steal
![[Live สด] การออกรางวัลสลากกินแบ่งรัฐบาล งวดวันที่ 16 ธันวาคม 2567](http://i.ytimg.com/vi/dNTiBBthKgQ/mqdefault.jpg)
![[LIVE] : ONE ลุมพินี 91 | คู่เอก "คมอาวุธ vs อเล็กเซย์"](http://i.ytimg.com/vi/V0RkEfYFsgM/mqdefault.jpg)
Bro this so fire !
Gonna drop Utopia before Travis will with this
Ong 😂😂😂
@@moeminebrahim1132fuckin rights
Dat song was hard u killed it bro
This video is amazing! I was trying to understand how the Heart on My Sleeves song was generated and I have watched a few videos, but none of them explained it as well as yours. Thanks for making it. Would like to see more AI + music videos like this!
Great video Sam, thanks for taking the time to put it together. Very clear explanation and as always, love the Smyers magic on those backing tracks. Your talent can never be reduced to an algorithm.
Sam have you tried Vocs AI for original AI singers? Game changer bro
hey sam! I followed all the steps you showed but when I press convert nothing seems to happen?
Same here
@@screensecretss Same. Any of you fixed it ?
Same
It definitely helps to sound like the original artist first to succeed and stand out amongst others. Everything else is a straightforward process.
This will be the death of authentic art
Nope. It will be the rise of it. 😉
@@dncviorel wrong. ;)
@@dncviorel wrong. ;)
@@dncviorel I agree. With a slider modifier you could have ultimate control over your voice profile. Artists could lease out their voice likeness, adding yet another way for them to capitalize on their artistry.
Tech is cool. The only viable use case is for fun or songwriters and producers to use it to pitch songs to artists and labels to be bought.
That is exactly what I've been thinking. I would love to have Taylor Swift's voice & Carrie Underwood's voice, to write songs for them both.
But they don't need the artist at all now
After 100 years AI will share this video to the people who want to go beyond AI.
how come there are no videos other than this one on my feed about this that aren’t shorts
great vid bro! I definitely think the more you can get it to sound like the artist your tryna convert the audio to you will get a better result, even listening to the bad romance one the results wasn't that bad especially knowing that it was a guy singing😅
I am not getting my demo version once I convert, everything is working great but I am not getting the final file with the artist model, any explanations?
The Setup 1, it tries to run the process for a few minutes and then I get an error and it keeps asking me to Restart the Session? Any idea how I can get past STEP 1 😅??
Same
okay so i have to the pont where you click convert..... but then nothing happens
I have an option to download the files but the .zip is empty.
@@collinbreckner5819 I found it. There is a folder in the svc repo that says something like final result
@@devonroyer2577 where can I find it?
Is there a link to this Discord community?
the singing absolutley sounds like drake he does that type of thing all the time in his songs, other than that great video!!
That shit is incredible... best drake song ever
This is awesome.. is there a link to the discord community ?
How to combine a few different voices into some new original one?
YES. We were just thinking this. I have a whole album full of finished stuff. One i would like a voice with 75% Cornell and 25% Bowie. How can I do this!!!I think you are looking for a locksmith that can do it inexpensively, is that right?
Guetta used audimee? James hype too.
I'm glad you explained too many people weirdly gatekeeped all this information
I am getting an error message when I try to run the "(Optional) Extract .zip Downloads - Step o.2." It gives me an error says "AttributeError: 'NoneType' object has no attribute 'filename'."
Anyone got the same issue? :(
Yeah same here bro. Idk how to fix it welp
All seems too technical for me, is there no program that you can just drag your audio file into and it can make it sound in the style of a chosen singer?
I don’t want to pretend that I am someone else, I just want to be able to sing a song and have it come out sounding good so I can do my own references without having to wait on songwriters and reference singers. I can’t sing at all tho. Have any suggestions?
ok so how do we use it? theres a million files and folders.. what do you click ?
whare is the discord server link
where can we get a weeknd voice model?.. anyone? :)
Cant get it to convert?.. hmmm
which website generates vocals: simulating female and male singer by just typing lyrics text, and specifying melody as well as pitch and key, and bpm?
Hello sir can you tell me how i can train ai model of other singer voice ?
How do you create a model with vocals of Whitney Houston for instance would you need to have Acapella Vocals?
I cant seem to get it to convert
I keep receiving a module not found error. Everything works until the phase where I drop my file into the Raw so-vits folder. Any guidance??
Hi is there a discord link to the AI World?
did you get it?
Just imagine spending your life recording beautiful music (eg. Brian Wilson), and then in the last week of your life you discover this for the first time...knowing that everything you've ever tried so carefully to protect, a legacy...will forever be tainted, the voice you once had no longer yours and now 'owned' by someone else :(
New realities are weird af af
i don’t think anyone’s legacy’s were tainted by this. those artists have talent that is irreplaceable and a random tiktok user with a very unrealistic ai rendering on their voice will not be able to make the same level art the original artists can
the good thing about the ai is when replacing rappers that are know for their writing they can’t make the same quality because they aren’t top teir song writers like those artists and ai has a hard time replicating the best singers because their voices are so nuanced and powerful. so unless it’s an artist that uses a lot of autotune and makes music that is very easy to replicate and the writing isn’t that good they won’t be able to make a convincing ai song pretending to be them. and the ones that do all of that aren’t making good enough music to the point where it matters if ai users are making music just like it because it isn’t valuable art anyways and is already commonly copied
@@thatguy1488 oh absolutely...just not yet.
what the hell are you talking about lol
Im getting multiple errors when trying to use the google colab links. Any tips or updates I need to try?
Are there tools that will basically have A.I. sort of autotune your own vocal? Or clone your voice and have it sing like another vocalist?
What the hell did he just say? they dont have apps where you can just download the audio file and it translates it?
Great Video Sam. My question is how do i use my own voice to sing like Taylor Swift? Is this possible? Can you please do a tutorial?
You’re watching the exact tutorial lmfao
😂😂
@@bloodclaat i'm assuming they mean 'i can't sing for s**t, can i sing and it also fix my bad singing', the best results of the AI ones are based off an already good vocal albeit it by someone else...but still a good singer for the input.
I cant find the AI Bot Channels on the Discord for Ai World - help please.
Anybody else follow the instructions at 6:48 but it doesn't output a file and instead just says
True
load model(s) from hubert/checkpoint_best_legacy_500.pt
and nothing else... any idea how to get an output?
no latest release after sovits SVC is outadeted... now coming
great - thank you sam. For songwriters this is gold.
Would like to see you remake the lead sound from turn on the lights knock2 remix.
Not sure if I am stupid or something here. But what am I supposed to do on the collab? I run the stuff, I press rerun, get errors. Then stops. So somewhere in the process I supply my own WAV I guess, but whats all these steps really doing and needed for?
Can you upload samples of another person's song (not a popular main stream artist) and make your voice sound like that?
I need Tom Araya, Nergal and the like please :)
Not working on my side
How do you train a new voice model?
Error in setup 1
Pyword is not founded
it keeps giving me errors with mega links
they are 4.0 svc
it stopped working! error message says no clustering model found
drake watching this right now punching the screen. who would have thunk it that we would be here with ai and he's this easily duplicated.. He better get ahead of this because it's moving without him.
He is ahead of it, he’s rich lol , there will be ppl who will never make money off their music cause of AI, at least he’s made his already… everyone will be using ai to make music all around the world, we will have millions of hits songs a week and no one but the person who typed in what they wanted and pressed a generate button will be needed. Which then eventually ppl will appreciate a human creating something cause it’ll be genuine and natural. So maybe I’m wrong but i have a feeling ppl will start calling themselves musicians just for typing in a description of what they want to hear and ai creating something good. They’ll say “listen to my song” and thinking they are a creator
the era of cathasthropes of music industry...
i cant find the restart runtime button bro what am i doin wrong
You didn't show tutorials how to download the codes
I can't find any Discord channel with that name
hello, how do I "train" this to sound like me?
Eyes On the money by drake sounds good ngl 😂
Insane 🔥 do this with some Jay-Z Vocals
Be careful running code from a google notebook, you may be giving access to your google files, contacts, email etc.
Couldn't find your discord channel
2 weeks ago, I saw a video tutorial on Vimeo on how to use Google colab to make an AI cover song and it was the perfect video talked about it, even the link was exist there , contained tons of the artists and there was an option to change a certain artist from a menu, why no one share this Google colab link which contains all of those artists. I can't find this video anymore 😭😭😭😭😭😭😭😭.
Please help , why you guys all your Google Collab links contains 4 artists only 😭😭😭.
I'm confused and stressed because of that , I might commit suicide, I had a plan to make advantage of this, can you please help
hey man, i understand your stress on all this ai stuff, i also was in a time a place where i wanted to take advantage of this entire ai program before people would figure it out, but it dosen't mean you have to commit suicide! I promise!
Just Go through your you tube history you can go back as far as you want lol
discord link of ai world?
Imagine you using a clone voice, then it became a sucess and then the drakeswitftgrande monster comes and eat all your cake? No. don't do it, we need a tool that makes a beautiful talented Jane Doe voice using the skills, not the tone.
Can I get sued if I made a track that has: Michael Jackson, Sam Smith, Snoop Dogg and Eminem in it?
How do i train my own voice on here?
You rock!
Ok ok good your video but how to make the model?
that taylor bad romance so sounded like Halsey
It does lol 💀
Can I use this to create a Dolly Parton voice model from some acapellas? Or does this site only let you convert your acapellas to existing ai models, Drake etc... Thanks in advance
Why would you want to??
the ai discord bot don't work anymore
I've been playing around with making song covers BUT by taking mps of songs and using vocals removers to extract acapella singing. But I've fond it's pretty rare for the AI to sound good all the way through the song, the voice always seems to crack and I'm guessing its because extracted vocals are never 100% clean. Anyone got any tips?
Maybe go verse by verse so instead of one long audio file you have multiple short audio files
I am about to go nuts with all the dead artists
would help adding the links
does this work on non-English languages?
not seeing download option
discord?
Is this free to use?
1:48
Link?
This is fuckin amazing, and it is scared
Of who?
it is scared?
@@markmarramusic because in the future IA Will can copy any voice and it is danger
@@xebario6089 Whos "IA Will"?
@@1mobile984 ?
Eyes on the moneyyyy😩😩😩
This is insane
Does this give you viruses?
Hello bro Timmy Trumpet - Synergy lead tutorial pls
im on the very last step after uploading the audio and i get this error "
ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject" I have no clue what to do
training a model is the most important part which is not covered unfortunately...
Who has the discord server link
I setup a SVC locally just in case the sites all get shutdown
how easy is it to do this?
i couldnt get to work
If your fans can't tell the difference between you and an AI bot, you're in the wrong line of work.
Ya is crazy really ducking crazy 😮😅😊
Sheesh that's alot of steps
12:04
what a champ
I just want to take popular songs, put my parodies under the same voice. This is all way too technical for me
I think like you are showing this to the people to make them learn how to do it, one day you also will be stolen & you will not find one dollar on your pocket to stay alive & eat, so my advise to you take care because what we do to some people one day it turns & return to you again, I am sure you know that, there is nothing better that you make effort on the things you create by yourself but trying to leave on the what other do this very bad, you as Musiian too AI will make you stay home too not finding work don’t worry not very far, as you learn people to steal you will be stolen don’t tell yourself I am showing people technology no you are learning them to steal