While AI singing or just using AI as voice over is great for in studio recording and listening, it is not good for live performances. We human actually like to hear real human singing and playing musical instruments in opinion. We human wants some errors in our work. It's our nature way to be flawed and accept some of those flaws. Not make everything as perfection in my eyes. Lol.
@@mhamusic. I struggle with this one too ngl. I hate my voice, but everyone else seems to think it's ok. I think it's just something you have to learn to get past. Your voice is great! Use it and it'll get even better.
Thanks for this video. Quick & varied BGVs, gang vocals, and lead vocal performances for sync songs come to mind immediately as use cases. Brilliant and crazy too!
@@FedorLosev i think it there is a legislative limbo in this space but I would not be so sure if faking a voice of another pro singer that is already in the market does not violate any laws. It should. Even Bjorn from Abba mentioned about those challenges ...
@@MP84OfficialI don't think there is a limbo. For sure it is not and can not be covered by any copyright law (as it is not a creation). But if you pretend to be somebody, make an identity with somebody or their publicity without permission, it is a completely different story. It is covered by personality rights etc. laws. So, if one produces a song in a genre and style of a known performer with a generated voice, making a clear connection to a specific person, yes, it is asking for a trouble. But in such a case it is less related to the voice or generation per se, and more to using a personality without permission. Background vocal in a different style/genre without clear connection to a personality is unlikely to constitute a case of misusing personality rights. Actually, it is not different in any way from using human voice similarity or face similarity (there are many people with a voice or a look similar to celebrities), be it generative or natural doesn't matter. "Faking" voice is irrelevant, but faking identity is. For example, you can hire a singer with a voice quite identical to a famous singer. The possibility of getting sued depends only on whether there is provable exploitation of a link to personal identity.
We need this implemented with all the other AI models into one big super AI model that does everything by itself exactly like a person but infinitely better ❤
@@ahmxrel_prod If music will be noticeably less original with AI, then many people will still crave and make music without AI. And when it finally does become as 'original' (this will certainly happen with enough time), then there is no difference as a music consumer if originality is what you want.
> Record outline vocal > Rerecord with AI so you can produce around it > Send full track to a trained singer so they can recreate the vocal, but better That seems to be the best use scenario for this, and it seems really helpful.
@@michaelsanders9147 Well, it's certainly easier to get good singers to work with you if you already have a song that sounds top notch and has vocals that represent what the finished product will sound like.
This would be great for song writers that are not great or powerful singers but want to hear how their song will sound like with good vocals. Also if you do produce a song a real song and hire a real singer or have a family member sing your song, they can already have a sample of how it will sound or how you want it to song so they can practice and imitate it better
Yeah. That's me. A songwriter with no real vocal talent or producing skills. This tool will help my pre-drmos sound much better I hope. I'm excited about the prospects
This would be a godsent for songwriters who maybe aren't great singers. Working with other people always has its issues as sometimes it's hard to articulate whet you want, but with this it'd be a great mask for composing. Vocal plugins aren't quite there yet, so this is the next best thing.
This technology is so amazing, and I think it might really help up and coming producers and songwriters the most. Like, I've always written together with two other guys. Neither of us can sing very well, and the vocal is SO important when pitching a song. So either we have had shit vocals on the demos or we had to pay a ton of money for a demo singer. This will make it so much easier to get a pleasant vocal take. I guess it might be nice for A&R's in the future as well, potentially hearing their artist on the demo already instead of someone else.
It just sounds like you passed it through melodyne and pitched everything perfectly! The Music Industry will change dramatically in the next 5 years that’s for sure!!!!
You are wrong. Any part of the music industry that likes vocals that sound like this, already has ways to obtain vocals that sound like this. Everybody else, who things the vocals in this video sound absolutely dreadful (like me), is not even remotely interested in AI. So nothing will change at all. Shit pop music will remain shit pop music, real heartfelt, human, skill and training based music will remain so as well and will not make use of AI.
@@isaacnewtech how am I wrong? This is just the beginning of AI in music, pretty soon everything will be made with AI in one way or another, the music industry will change dramatically in the next 5 years and less.
This is really brilliant thank you so much!! super helpful and also so entertaining love the humour!! btw your real voice is actually amazing! and sounds great!
@@MrSatoou That is very true, but I have also seen many softwares / websites that can create full songs in seconds. A lot is about to change in the music industry... But I am 100% excited for what resources are now available.
Impressive and good demonstration and will help writers pitch songs to popular artists. Regarding sample packs, I am not convinced the legal system will agree without a licensing model in place so artists can benefit from the usage of their voices and be compensated.
:59 the fry intro on line 2 didn't quite work and the quality seems a hair degraded overall to me. 1:01 the "t" in time there was blurred. But other than that, yeah it's a different timbre mapping and it sounds usable if you don't need a professional result.
Whoa, nice video dude. Those AI tools are awesome! Thanks for sharing! Btw, your questioning at the end of the video is basically the one that was flooding the internet on the visual art theme: "Is it okay if you're a singer and someone just grab a bunch of vocals from you/your tracks and train a model to make music/sample packs and earn money on your voice?" I don't know the exact answer... one might find it okay if they get a share of the incomings, one might disagree with that idea at all... That AI thing still in it's early stages, we might have a clearer answer in the near future, I hope.
You ABSOLUTELY have a wonderful singing voice - Clean and soft and PERFECT and very nice and calm - I would definitly buy an AI Voice made with your voice !
I think the artifact at the beginning is from the in-breath sound. The first 2 models you tried probably didn't have that in their training data while the Ed Sheeran model did.
I think it's a good concept overall, but I think where the problems arise is when people abuse this technology. I personally don't think we should be "cloning" popular singers' voices, but if there was a way to make our own "Personal" AI voice model that we can create and customize ourselves would be a better use of such technology. I think it should be used as a tool to "enhance" people's natural voices, and not trying to "emulate" people who already exist. This creates the obvious legal complications & potentially opens yourself up to litigation if you actually try to use someone else's voice in your project.
We pretty soon will connect our own individual intellectual properties to the blockhain. Can you imagine? Due to copyright, that's the current way I see the future developing, we will be charging $ to use that IP.
I´d feel completely imbecil if I would ever find myself in a situation where I am faking another artist`s original voice. Why would you do that? To break your bathroom-mirror?
Great overview, thanks! So many questions: - Did you also try Voicify? If so, how does it compare to Lalal? - Is the Pro output from Musicfy really worse than the Demo version? - Would using autotune to pitch up a male vocal b4 putting it thru a female model help with range differences? - Finally, are those Woodupp panels on the wall behind you?!
completely honest, I think you should sing on your own songs in the future. With frequent singing practice you'll also feel more comfortable. your singing is not "cringey" at all - probably everyone feels like that about his/her own voice. 😉👏
Bro this is so good yet so dangerous at the same time, at this rate some of the well known artist’s demo/id will be leaked in hq more easily, like imagine someone take advantage of the Lala AI and recreate a mimic of rewind repeat it? I mean sure I’m a fan and all but thinking about the possibilities in these AI are making me excited yet kinda terrified at the same time😅
This got me so bloody excited always struggle with vocalists and I do not feel my voice is good enough no matter how many compliments I would get...BUT then my next thought was like FARK this is crazy scary for EVERYONE involved. Producers, Engineers , Vocalists all are becoming less and less needed, or can be completely abandoned altogether 😢😢 all the hard work sweat time and effort to become better, become the next pro seem like a waste....Let's try to keep this at bay at least for another decade so I can retire as a musician 😶🌫😶🌫😶🌫😶🌫😶🌫😶🌫
@@drinkinslim AI in music has been around for a long time and is very common. It's used in everything from instrument separation tools, remixing, and even neural network denoisers and de-reverbs. The ability to reasonably make your voice sound like the opposite gender of you for vocals is literally the impossible made possible. This is what I meant by "one of those things that can slide." AI is so misunderstood and misused that it became the boogeyman of the internet. I don't see the integrity of stealing a big artist's voice and putting it under a beat generated to sound like some big producer's style for clout and money, but using it as a tool to creatively achieve a unique sound you are seeking without the help of others makes artists even more independent and retains integrity. AI is a tool, and so it should be thought of, and used like one. No tool can replace humans, as it relies on the expression and creativity of humans to be put to use, but it makes life easier for individuals trying to enjoy their hobby without getting peers and money into the equation.
it's good but in all honesty this concept has been around for 20 years and it always surprised me that it wasn't developed quicker. vocaloid used text input and what sounded like multi band vocoding, filtering and formant manipulation to create 'AI' vocals back in the mid noughties. I think the AL zeitgeist will see this concept of creating vocals us ing AI lyric generators linked to more sophisticated voice synthesisers being developed a lot in the next in the coming months.
Also, Grimes is one of the first artists to have her own AI which everyone can use. The only thing you have to do is give the AI 50% of the split if you release a song with it in it
@@kahyui2486 I think she has every right to disallow use of her "own" voice out there saying things she wouldn't want to or sining in ways she wouldn't. I've listened to a couple songs already made with her A.I. voice, and as a longtime fan of hers it was clear it wasn't really her, either because they had her singing in a range lower than she ever sings, or the voice lacks the slight lisp that is a delightful part of her sound. But after hearing her latest release "welcome to the opera" a few days ago I decided I'm really inspired to make a song with "her"!
@@kojoefante when you're already a multimillionaire with nothing to lose. It's asinine and completely out of touch with the reality of small creators, who DON'T have access to giant legal teams ready to enforce their will.
your singing in this sounds really good, it's not cringe. and I think the quality of the vocal pack from you would be high, so I don't see no problem with you selling an ai vocal pack. I guess as long as it's the generic voices and not celebrities, because I don't know the legality behind that.
This is really cool because I’m doing music since 2010 and I never found friends with the same hobby - so now I can really do my own music with vocals on it! :)
This one great new video dude!!! Thanks for showing us a nice little demo of how these AI voices can be used...I do love the idea of transforming our own voices into the other gendered type. I also agree, your voice is pretty sweet...some vocal exercises for a little while will get you singing much better. You do have a nice quality to your voice for sure, great tone as well. Stay oh so so blessed dude! :)
This is insane 🤯. But can you not get copyrighted for this if you're using a famous singer AI voice and release that track under your name?? Also, what if you have The Weeknd AI doing harmonies in one of your tracks and me singing lead vocal, would this be okay to release under your name for copyright matters?
just tested it out. For now, it's having a hard time with the cracks in my voice and other slight nuances in timbre. Kinda glad it's not there yet, but as soon as it is I'm about to drop something crazy w it
If you want to keep it safe without any fear of tracks being taken down because the law changes, better use synthesizer v, it's also good for people who can't sing and sounds just as realistic when used correctly. You can also use it to transform vocals like this, using praat and a script
A lot of hate and for some reason personal attacks as well 😂 but yeah its great having a video pop off like this i guess. Most people see the use of it! Thanks for the kind words man! 🙏❤️
I mean I get the hate, I’m scared too but you just showed us how to use it, you didn’t make the software lol😂 So I don’t really see why someone would be mad at you
Interesting, thanks for showing. Question: For a female performance, wouldn't it work better if you just transpose your own voice a couple of semitones after recording the vocal melody? I think singing in falsetto will alter the timbre and energy of your vocal track and should be avoided for maximum results.
Sounds like it's more than good enough for backing vocals, back in the mix or even not so far back. Impressive. And it must be getting better fast. Makes me dizzy.
In 5-10 years people will just be able to tell AI to make them "an incredible electronic indie song with female vocals like Daughter in "Youth" " and it'll just make the best thing you've ever heard. As someone that's always wanted to make music, but wasn't born lucky with a good voice, I've always wanted this particular ability to be able to create a voice. But AI will soon surpass our singing, writing and creative abilities and I have to wonder where that leaves artists and us as humans when the works created by our best artists, which could take years, can be dwarfed in a fraction of a second. Does humanity begin to develop a "why even bother" attitude towards creating our own art?
Maybe. As a drummer though, you can get high quality drum samples that sound like real drums, you can buy drum packs that come with a bunch of beats. And you could use them live and they’d sound great, it’d save getting a drummer and would make a band simpler. But people still get drummers even though technically you can reproduce their sound digitally. I think in the future you will be able to customise any movie or song you want, but watching someone play an instrument live really well is a different type of experience.
thats not really true. i think people don't really have a good conception of what makes something art. AI will be a great tool for producers and also will be able to to generate muzak, like 'chill lofi beats' and whatever bland background music / stock music, and maybe a pop / hip hop song. but anything that is truly art will never be able to be created with AI except with an artist using it as a tool. the idea it will just crank out art is just a false concept. AI is just a tool much like using logic pro or anything else. the reason why AI can't make art is because real art involves many subtle details, or a high degree of information. Like the best movies for example the director will be extremely meticulous with all the details on screen. Studies show the unconscous mind is aware of these details that the conscious mind isn't aware of. It takes in information an order of magnitude more than conscious mind, which is why subliminals work. in other words, there is a reason why we are able to tell when something is great even we don't know why. So this is why AI can't just crank out art like that, because it is the tiny details that only a master can do that makes something art and we CAN tell the difference. But yes AI will be able to make bland art. which actually could replace a lot of stuff as you suggest
OMG this is a life changer!!!! I work with K-Pop and I love to write my toplines but I'm not a good singer and the publishers are VERY picky, also de the labels A&Rs. Sometimes they don't want my Demos even when it's good because my voice it's not professional enough. And when I hired other vocalists, they change the way I sang the song and lost the Soul I put on it. The A&R literallly sing it like my recording BUT WITH A PROFESSIONAL SINGER VOICE. I gonna make my job as I always did but now I don't need to worry about my vocals and make sure to make the best DEMOS!!!
Its crazy to think about how far AI is going haha, but at the same time its harmful for vocalists (i personally work as a singer-songwriter) and thinking what i do could be replaced with AI is completely scary :(
Like pretty much all AI, it seems to be threatening mid tier and below content creators. If you are a top quality vocalist, AI isn't going to replace you any time soon. Just like how sample libraries haven't really replaced live performance. Maybe over the next few decades but certainly not in the near future.
well you just have become producer or join a band. You just no longer get commissioned to sing someone else's music unless it large label that can afford it. AI synth vocal very human sounding. vocal synth can came a long way since 2004 with release of vocaloid 1. How ever vocal synth vocals are still underground thing. iF vocal synth song get 1,000 views it hit for genre.
AI music is not scary for people with talent, but scary for all those wannabe producers / artist out there with zero talent making generic beats, AI is about to take your jobs 😉finally we get all the mediocre people out of the industry , bye bye
Your generation is so afraid of being “cringe”… You should just drop it entirely. Why care about whatever anyone thinks/says about you?? Just be you. Do you. You are magnificent the way you are. We all are. No matter what anyone thinks.
Way to go man.. You doing it for everyone when your voice is cold and you're humble about it. It shows that you're authentic and that is a rare thing on you tube thus I'm interested in you're channel and the various things you present.. Thanks for being real!!
I tried a few using my own shxt vocals thru a MacBook Pro mic into Ableton and bounced the stems at like -20dB lol and it was still like 10X better…and def beyond just usable…exciting times 🤪😎🍭. Thanks G
You know.... I've been using a 10 year old iMac with the Garageband from that old edition, because it was the last version of Garageband that had a "vocal effect" built in that alters male vocals flawless into female vocals. Somehow they left it out in later upgrades. Very pitty and very annoying. But....... It wasn't until now, with AI and the possibility to alter vocals, that there are great new and HQ tools that do that same job, but: much better and with a much wider range of possibilities. I'm so exited!
Capture harmonic tracks with various celeb voices and mix them together with the melody track for some all-star duets, trios, and choir pieces. Put celebs to work singing backup to YOU. The sky's the limit.
Great video and after trying this for harmony parts (especially if you're like me and cannot sing harmonies well), is add something like iZotope harmony generator, or Waves or Antares (I have the Antares one) which you can turn the main vocal off and just export the harmony (which usually sounds really bad and digital) and import that into one of these and then re-import the newly created human sounding harmony, sounds really amazing... So thanks for sharing this, realistic harmony parts have always been an issue for me making them sound real with generators, but now they absolutely do pass as real in the mix.
nice video, I know its a bit old now, but wow is the ai a game changer. From my point of view within the next 5 years any producer will be substituted.and singer. I like it and even with this quality you can use it in commercial music and no one will know. I it will come down to the skill of marketing that is it as writing hits will become easier with every step of the way.
![Lp. Сердце Вселенной #60 РОЖДЕНИЕ ЛОЛОЛОШКИ [Финал] • Майнкрафт](http://i.ytimg.com/vi/YoR0pAV9FVQ/mqdefault.jpg)
Check out my new vocal pack here if you're interested! ⬇
www.mhamusic.com/
Your voice is better than the AI if you ask me...
love this producer
nice
What AI did you use to change your vocals?
While AI singing or just using AI as voice over is great for in studio recording and listening, it is not good for live performances.
We human actually like to hear real human singing and playing musical instruments in opinion. We human wants some errors in our work.
It's our nature way to be flawed and accept some of those flaws. Not make everything as perfection in my eyes. Lol.
your real singing voice is amazing, you literally have the infinite vocal glitch lmaoo
Hahaha thank you man, that actually means a lot. With this AI tech you really dont need your voice to be special at all which is so great
@@mhamusic. no joke there’s nothing to cringe about , your voice is really good . Make your own tracks with your voice 😮💨
@@mhamusic. I struggle with this one too ngl. I hate my voice, but everyone else seems to think it's ok. I think it's just something you have to learn to get past. Your voice is great! Use it and it'll get even better.
Concur, it's actually not bad. Rule #1...be Believable no matter what you sing.
This could be a great tool to make a choir out of just one vocals track.
Stuff like this is what this technology is great for!
Well there is already that with the kontakt things by native instruments check the dominus choir or the eastwest Hollywood choirs
@@salvab_music yea but it feels different if you control it with your own voice
Tho this workflow sucks for now
Agreed!
@@salvab_musiclmao this is nothing like that
Another use case: If you work with a singer regularly, you can create a model of that singer and produce demos for them already using their voice.
Mind blown.
Really now ?🤔
Dumb
True 👌🏻🔥
Wait what? Really?
Thanks for this video. Quick & varied BGVs, gang vocals, and lead vocal performances for sync songs come to mind immediately as use cases. Brilliant and crazy too!
Exactly that! Thanks for watching
Randy Marsh was way ahead of his time
Best comment
😂😂😂😂
I am Lorde, yeah yeah yeah!
Lorde lorde yeah yeah
Feeling good on a Wednesday!
Ideal for backing vocals. Just run a few variations through different models and layer them up so it doesn't sound like 'just' an AI.
Right, ad 4 famous singers mixed as one voice and it will be almost impossible so sue
@@bubblanbIt is impossible to sue in any case, voice is not a subject to copyright
@@FedorLosev i think it there is a legislative limbo in this space but I would not be so sure if faking a voice of another pro singer that is already in the market does not violate any laws. It should. Even Bjorn from Abba mentioned about those challenges ...
@@MP84OfficialI don't think there is a limbo. For sure it is not and can not be covered by any copyright law (as it is not a creation). But if you pretend to be somebody, make an identity with somebody or their publicity without permission, it is a completely different story. It is covered by personality rights etc. laws. So, if one produces a song in a genre and style of a known performer with a generated voice, making a clear connection to a specific person, yes, it is asking for a trouble. But in such a case it is less related to the voice or generation per se, and more to using a personality without permission. Background vocal in a different style/genre without clear connection to a personality is unlikely to constitute a case of misusing personality rights. Actually, it is not different in any way from using human voice similarity or face similarity (there are many people with a voice or a look similar to celebrities), be it generative or natural doesn't matter. "Faking" voice is irrelevant, but faking identity is. For example, you can hire a singer with a voice quite identical to a famous singer. The possibility of getting sued depends only on whether there is provable exploitation of a link to personal identity.
@@FedorLosev yes it makes sense what you have written
0:26 Oh brooo! 😆 Nice pictures btw 🤔 And song! 💛
We need this technology implemented into a plugin
Working on it.
@@jeremywright7832 Seriously? You mean that you are? I would love to be kept updated on that.
We need this implemented with all the other AI models into one big super AI model that does everything by itself exactly like a person but infinitely better ❤
I really don't like this whole ai thing. Music would really lose originality
@@ahmxrel_prod If music will be noticeably less original with AI, then many people will still crave and make music without AI. And when it finally does become as 'original' (this will certainly happen with enough time), then there is no difference as a music consumer if originality is what you want.
i would surely buy a vocal pack. Your packs are usually at a very high quality so i personally think it would be a great idea
Good to hear that!! And thanks so much
Wow, thanks! Now I can use my vocal for lead and overdub female vox and also change my own overdubs to a complementary different voice. Super helpful!
> Record outline vocal
> Rerecord with AI so you can produce around it
> Send full track to a trained singer so they can recreate the vocal, but better
That seems to be the best use scenario for this, and it seems really helpful.
I hope this will be the standard workflow for producers.
So where do you find trained singers laying around with amazing voices?
@@michaelsanders9147 Well, it's certainly easier to get good singers to work with you if you already have a song that sounds top notch and has vocals that represent what the finished product will sound like.
I can see how this would help also!
This tech is perfect for songwriters in terms of creating song demos, pitching purposes, etc.
Great video, I might want to add that you could fix a few things with Melodtne .. not pitch fix but lower, og set some tones higher before training :)
This would be great for song writers that are not great or powerful singers but want to hear how their song will sound like with good vocals.
Also if you do produce a song a real song and hire a real singer or have a family member sing your song, they can already have a sample of how it will sound or how you want it to song so they can practice and imitate it better
Yamaha: or you know,you can just buy our vocaloid
Yeah. That's me. A songwriter with no real vocal talent or producing skills.
This tool will help my pre-drmos sound much better I hope.
I'm excited about the prospects
Vocaloid exist for a reason,you might want to look into it.
Damn, having so much fun with this! Thank you for this, man!
This would be a godsent for songwriters who maybe aren't great singers. Working with other people always has its issues as sometimes it's hard to articulate whet you want, but with this it'd be a great mask for composing. Vocal plugins aren't quite there yet, so this is the next best thing.
Yes it is. As a Songwriter this video has given me life 😊
This technology is so amazing, and I think it might really help up and coming producers and songwriters the most. Like, I've always written together with two other guys. Neither of us can sing very well, and the vocal is SO important when pitching a song. So either we have had shit vocals on the demos or we had to pay a ton of money for a demo singer. This will make it so much easier to get a pleasant vocal take. I guess it might be nice for A&R's in the future as well, potentially hearing their artist on the demo already instead of someone else.
It just sounds like you passed it through melodyne and pitched everything perfectly! The Music Industry will change dramatically in the next 5 years that’s for sure!!!!
I was gonna mention the ability to do this in melodyne
It does sound like melodyne but it also doesn’t sound like a real voice
@justinstewart3248 sounds pretty good too me
You are wrong. Any part of the music industry that likes vocals that sound like this, already has ways to obtain vocals that sound like this. Everybody else, who things the vocals in this video sound absolutely dreadful (like me), is not even remotely interested in AI. So nothing will change at all. Shit pop music will remain shit pop music, real heartfelt, human, skill and training based music will remain so as well and will not make use of AI.
@@isaacnewtech how am I wrong? This is just the beginning of AI in music, pretty soon everything will be made with AI in one way or another, the music industry will change dramatically in the next 5 years and less.
This is really brilliant thank you so much!! super helpful and also so entertaining love the humour!! btw your real voice is actually amazing! and sounds great!
I've been waiting for this for SOOOOO long!
Kinda insane where this is all going, seems pretty scary for producers. But I am excited for what is coming.
Scary for singers and recording technicians! Heaven for producers.
@@MrSatoou That is very true, but I have also seen many softwares / websites that can create full songs in seconds. A lot is about to change in the music industry... But I am 100% excited for what resources are now available.
@@TrapVanGoat I think a good chunk of producers can realistically be replaced with tools like Soundraw and Google's Music LM
@@MrSatoou haha was just gonna say this. Producers (me) dream 😂
Not scary for producers, they dont need singers anymore in the studio en make own full songs.
Impressive and good demonstration and will help writers pitch songs to popular artists. Regarding sample packs, I am not convinced the legal system will agree without a licensing model in place so artists can benefit from the usage of their voices and be compensated.
Imagine putting Rick Astleys vocals on your bass house track....
:59 the fry intro on line 2 didn't quite work and the quality seems a hair degraded overall to me. 1:01 the "t" in time there was blurred. But other than that, yeah it's a different timbre mapping and it sounds usable if you don't need a professional result.
Whoa, nice video dude. Those AI tools are awesome! Thanks for sharing!
Btw, your questioning at the end of the video is basically the one that was flooding the internet on the visual art theme:
"Is it okay if you're a singer and someone just grab a bunch of vocals from you/your tracks and train a model to make music/sample packs and earn money on your voice?"
I don't know the exact answer... one might find it okay if they get a share of the incomings, one might disagree with that idea at all...
That AI thing still in it's early stages, we might have a clearer answer in the near future, I hope.
Yup, still early stages. Its the wild West right now! Completely agree with you
dude thats badass,you get your props
You ABSOLUTELY have a wonderful singing voice - Clean and soft and PERFECT and very nice and calm - I would definitly buy an AI Voice made with your voice !
I Second This 🙋♂️
Thanks guys!
Thanks for featuring our service 💛 much appreciated!
I think the artifact at the beginning is from the in-breath sound. The first 2 models you tried probably didn't have that in their training data while the Ed Sheeran model did.
This sounded crazy good! Great tutorial 💯
Although I already get a pretty good result with pitch shifting for female vocals, but it's insane how much AI has evolved recently...
This is insane! Dope vid!🤙 btw 4:07 had me dyin 😂
I think it's a good concept overall, but I think where the problems arise is when people abuse this technology. I personally don't think we should be "cloning" popular singers' voices, but if there was a way to make our own "Personal" AI voice model that we can create and customize ourselves would be a better use of such technology. I think it should be used as a tool to "enhance" people's natural voices, and not trying to "emulate" people who already exist. This creates the obvious legal complications & potentially opens yourself up to litigation if you actually try to use someone else's voice in your project.
Damn, I thought one of these tools would be able to make your own model. I know some people have done that but it was way more advanced.
This the best idea I have heard. AI enhancing the user's voice to sound as good as the pros....or better!
On the first link there was a option to train your own ai voice
We pretty soon will connect our own individual intellectual properties to the blockhain. Can you imagine? Due to copyright, that's the current way I see the future developing, we will be charging $ to use that IP.
I´d feel completely imbecil if I would ever find myself in a situation where I am faking another artist`s original voice. Why would you do that? To break your bathroom-mirror?
That "we are the fighters" would be such a good hook in a progressive house song. It actually makes me want to learn a little about lyric writing
Can't wait to add MHA vocals to my songs.
❤️
Great overview, thanks! So many questions:
- Did you also try Voicify? If so, how does it compare to Lalal?
- Is the Pro output from Musicfy really worse than the Demo version?
- Would using autotune to pitch up a male vocal b4 putting it thru a female model help with range differences?
- Finally, are those Woodupp panels on the wall behind you?!
Did you choose one? If so which did you prefer?
Ed Sheeran: "Just because they are the same chords doesn't mean I stole the song!"
MHA: *"Hey Ed, is this your voice?"*
Lmao
We just need someone to release an Ed Sheeran song, sung by Marvin Gaye 😂
You have a great voice, the tech is sick, and I'd love to see more great videos from you. TY
completely honest, I think you should sing on your own songs in the future.
With frequent singing practice you'll also feel more comfortable. your singing is not "cringey" at all - probably everyone feels like that about his/her own voice. 😉👏
Thank you!
I was too lazy to comment before I saw yours, but yeah, agreed. The singing is pretty good actually. Not cringy at all.
@@heroforesthis songs are good too. I was like "Who is this guy? New subby: 😊
your voice is amazing man
Bro this is so good yet so dangerous at the same time, at this rate some of the well known artist’s demo/id will be leaked in hq more easily, like imagine someone take advantage of the Lala AI and recreate a mimic of rewind repeat it? I mean sure I’m a fan and all but thinking about the possibilities in these AI are making me excited yet kinda terrified at the same time😅
It's already happening search Michael Jackson n stuff
I thought vocaliod pro vst or somthing but now i know how they done it. Thanks for sharing. CRAZY
This got me so bloody excited always struggle with vocalists and I do not feel my voice is good enough no matter how many compliments I would get...BUT then my next thought was like FARK this is crazy scary for EVERYONE involved. Producers, Engineers , Vocalists all are becoming less and less needed, or can be completely abandoned altogether 😢😢 all the hard work sweat time and effort to become better, become the next pro seem like a waste....Let's try to keep this at bay at least for another decade so I can retire as a musician 😶🌫😶🌫😶🌫😶🌫😶🌫😶🌫
For real .People with Real talent will be replaced by computer nerds
This is insane ......
Thanks for the upload
So good!!! Cheers, but I really think you should try making a song with many artists singing on it with the AI voices
Wow! This is massively impressive. And of course, just the start. ✨
This is one of those AI things in music that actually can slide. This is epic
Can slide? As in, go away?
@@drinkinslim AI in music has been around for a long time and is very common. It's used in everything from instrument separation tools, remixing, and even neural network denoisers and de-reverbs. The ability to reasonably make your voice sound like the opposite gender of you for vocals is literally the impossible made possible. This is what I meant by "one of those things that can slide." AI is so misunderstood and misused that it became the boogeyman of the internet. I don't see the integrity of stealing a big artist's voice and putting it under a beat generated to sound like some big producer's style for clout and money, but using it as a tool to creatively achieve a unique sound you are seeking without the help of others makes artists even more independent and retains integrity. AI is a tool, and so it should be thought of, and used like one. No tool can replace humans, as it relies on the expression and creativity of humans to be put to use, but it makes life easier for individuals trying to enjoy their hobby without getting peers and money into the equation.
ok this was ❤🔥❤🔥❤🔥 but lets talk about that intro bass and synth... i know its simple but sounds sick!!
This is mind blowing 🤯 The music industry should absolutely take AI so serious…. This reminds me of the MP3 sharing era 💿 ⬇️
I expect the music industry to attempt to use AI to cut out artists entirely
I was at a conference about Computer-assisted music in France, they (the music industry) were talking about this very topic ... in 1999.
they should take it serious? hahaha who do you think is behind all this?
npc comment, you got to be insane to say that, there's no single though behind your comment.
💯💯💯
Wtf dude ur voice tops it holy! U got such a great voice
it's good but in all honesty this concept has been around for 20 years and it always surprised me that it wasn't developed quicker. vocaloid used text input and what sounded like multi band vocoding, filtering and formant manipulation to create 'AI' vocals back in the mid noughties. I think the AL zeitgeist will see this concept of creating vocals us ing AI lyric generators linked to more sophisticated voice synthesisers being developed a lot in the next in the coming months.
Dude, your song is a masterpiece. It's stuck in my head. Great tune. Great singer, great guide.
Also, Grimes is one of the first artists to have her own AI which everyone can use. The only thing you have to do is give the AI 50% of the split if you release a song with it in it
If she don't like it she will take it down tho 😂
@@kahyui2486 I think she has every right to disallow use of her "own" voice out there saying things she wouldn't want to or sining in ways she wouldn't. I've listened to a couple songs already made with her A.I. voice, and as a longtime fan of hers it was clear it wasn't really her, either because they had her singing in a range lower than she ever sings, or the voice lacks the slight lisp that is a delightful part of her sound. But after hearing her latest release "welcome to the opera" a few days ago I decided I'm really inspired to make a song with "her"!
But why would you do that as a creative person?
@@razoraz no shit sherlock
@@kojoefante when you're already a multimillionaire with nothing to lose. It's asinine and completely out of touch with the reality of small creators, who DON'T have access to giant legal teams ready to enforce their will.
Wow just came back to this vid, holy is it doing well!! So deserved
Really appreciate that mate 🙏❤️
Imagine using the ai voices and have them speak in different languages, applying them in the songs as a dub for other countries. 😳😳
thats genius
Keep making sample packs lol these are fire
dude why is your real voice so good. you also nail nearly every note dude dont make yourself worse than you are
Haha thank you!
Aaaaaaaaaaaayeeeee ❤️ Thanks for the Tutorial 🎉
Glad you enjoyed it!
your singing in this sounds really good, it's not cringe. and I think the quality of the vocal pack from you would be high, so I don't see no problem with you selling an ai vocal pack. I guess as long as it's the generic voices and not celebrities, because I don't know the legality behind that.
Thank you so much!
This is really cool because I’m doing music since 2010 and I never found friends with the same hobby - so now I can really do my own music with vocals on it! :)
Thats exactly what i see this being used for! Glad you found it helpful :)
This one great new video dude!!! Thanks for showing us a nice little demo of how these AI voices can be used...I do love the idea of transforming our own voices into the other gendered type. I also agree, your voice is pretty sweet...some vocal exercises for a little while will get you singing much better. You do have a nice quality to your voice for sure, great tone as well. Stay oh so so blessed dude! :)
Thanks so so much!
Love it. Such great fun, currently working on a dance model. God, where has the time gone.
I just understood by hearing that it's Tate McRae. It's insane how accurate these AI tools are.
Really crazy
This is a great thing to be covering right now! You made me realize I should have thought about this earlier lmao
This is insane 🤯. But can you not get copyrighted for this if you're using a famous singer AI voice and release that track under your name?? Also, what if you have The Weeknd AI doing harmonies in one of your tracks and me singing lead vocal, would this be okay to release under your name for copyright matters?
Don’t use there name ;)
just tested it out. For now, it's having a hard time with the cracks in my voice and other slight nuances in timbre. Kinda glad it's not there yet, but as soon as it is I'm about to drop something crazy w it
It does have its limitations and different models will suit different singers better. Cant wait to see where this will be in the future
heyy mha quick question how do you record the guitars that well? and if you can could make a detailed video about the whole process?
Its just through my sm7b and doing multiple takes and compression, eq etc
Your natural voice is one of the nicest voice I’ve ever heard :) 👍
If you want to keep it safe without any fear of tracks being taken down because the law changes, better use synthesizer v, it's also good for people who can't sing and sounds just as realistic when used correctly. You can also use it to transform vocals like this, using praat and a script
YOU ARE SINGING SO GOOD OMG!!!
Haha thank you
Great to see your video blowing up man!🔥❤️
Your vocals are great btw
A lot of hate and for some reason personal attacks as well 😂 but yeah its great having a video pop off like this i guess. Most people see the use of it! Thanks for the kind words man! 🙏❤️
I mean I get the hate, I’m scared too but you just showed us how to use it, you didn’t make the software lol😂
So I don’t really see why someone would be mad at you
Interesting, thanks for showing. Question: For a female performance, wouldn't it work better if you just transpose your own voice a couple of semitones after recording the vocal melody? I think singing in falsetto will alter the timbre and energy of your vocal track and should be avoided for maximum results.
It could! I think different voice models work best in different ways, depending on your voice and the model of course
Sounds like it's more than good enough for backing vocals, back in the mix or even not so far back. Impressive. And it must be getting better fast. Makes me dizzy.
In 5-10 years people will just be able to tell AI to make them "an incredible electronic indie song with female vocals like Daughter in "Youth" " and it'll just make the best thing you've ever heard. As someone that's always wanted to make music, but wasn't born lucky with a good voice, I've always wanted this particular ability to be able to create a voice. But AI will soon surpass our singing, writing and creative abilities and I have to wonder where that leaves artists and us as humans when the works created by our best artists, which could take years, can be dwarfed in a fraction of a second. Does humanity begin to develop a "why even bother" attitude towards creating our own art?
It's a very good question. Interesting to think of a future, where all art you see and hear, is AI generated.
Maybe. As a drummer though, you can get high quality drum samples that sound like real drums, you can buy drum packs that come with a bunch of beats. And you could use them live and they’d sound great, it’d save getting a drummer and would make a band simpler. But people still get drummers even though technically you can reproduce their sound digitally. I think in the future you will be able to customise any movie or song you want, but watching someone play an instrument live really well is a different type of experience.
Probably 5-10 months, not years.
That already exists
thats not really true. i think people don't really have a good conception of what makes something art. AI will be a great tool for producers and also will be able to to generate muzak, like 'chill lofi beats' and whatever bland background music / stock music, and maybe a pop / hip hop song. but anything that is truly art will never be able to be created with AI except with an artist using it as a tool. the idea it will just crank out art is just a false concept. AI is just a tool much like using logic pro or anything else. the reason why AI can't make art is because real art involves many subtle details, or a high degree of information. Like the best movies for example the director will be extremely meticulous with all the details on screen. Studies show the unconscous mind is aware of these details that the conscious mind isn't aware of. It takes in information an order of magnitude more than conscious mind, which is why subliminals work. in other words, there is a reason why we are able to tell when something is great even we don't know why. So this is why AI can't just crank out art like that, because it is the tiny details that only a master can do that makes something art and we CAN tell the difference. But yes AI will be able to make bland art. which actually could replace a lot of stuff as you suggest
OMG this is a life changer!!!! I work with K-Pop and I love to write my toplines but I'm not a good singer and the publishers are VERY picky, also de the labels A&Rs. Sometimes they don't want my Demos even when it's good because my voice it's not professional enough. And when I hired other vocalists, they change the way I sang the song and lost the Soul I put on it. The A&R literallly sing it like my recording BUT WITH A PROFESSIONAL SINGER VOICE. I gonna make my job as I always did but now I don't need to worry about my vocals and make sure to make the best DEMOS!!!
Its crazy to think about how far AI is going haha, but at the same time its harmful for vocalists (i personally work as a singer-songwriter) and thinking what i do could be replaced with AI is completely scary :(
Like pretty much all AI, it seems to be threatening mid tier and below content creators. If you are a top quality vocalist, AI isn't going to replace you any time soon. Just like how sample libraries haven't really replaced live performance. Maybe over the next few decades but certainly not in the near future.
well you just have become producer or join a band. You just no longer get commissioned to sing someone else's music unless it large label that can afford it. AI synth vocal very human sounding. vocal synth can came a long way since 2004 with release of vocaloid 1. How ever vocal synth vocals are still underground thing. iF vocal synth song get 1,000 views it hit for genre.
Quality video fella and funny as f*** love the real talk bro!! Also great tutorial
Thanks a lot! Appreciate that! 🙏
AI music is not scary for people with talent, but scary for all those wannabe producers / artist out there with zero talent making generic beats, AI is about to take your jobs 😉finally we get all the mediocre people out of the industry , bye bye
That's absolutely insane ❤
Your generation is so afraid of being “cringe”… You should just drop it entirely. Why care about whatever anyone thinks/says about you?? Just be you. Do you. You are magnificent the way you are. We all are. No matter what anyone thinks.
Calling someone magnificent is cringe
Really appreciate your efforts keep up the honest work ✅
How do people not understand that AI is not here to "help" producers, musicians, or anyone else? It's here to REPLACE us...
This blew my mind for real, i dont even know how it all works but it sounds awesome :P
Way to go man.. You doing it for everyone when your voice is cold and you're humble about it. It shows that you're authentic and that is a rare thing on you tube thus I'm interested in you're channel and the various things you present.. Thanks for being real!!
Thanks for the kind words! I really appreciate that ❤️
Very Prophetic song if you are Not Saved!! "Believe on The Lord Jesus Christ and thou Shalt be Saved" or Don't!!!
I tried a few using my own shxt vocals thru a MacBook Pro mic into Ableton and bounced the stems at like -20dB lol and it was still like 10X better…and def beyond just usable…exciting times 🤪😎🍭. Thanks G
Wow!! Just WOW! Great video!!
You know.... I've been using a 10 year old iMac with the Garageband from that old edition, because it was the last version of Garageband that had a "vocal effect" built in that alters male vocals flawless into female vocals. Somehow they left it out in later upgrades. Very pitty and very annoying. But.......
It wasn't until now, with AI and the possibility to alter vocals, that there are great new and HQ tools that do that same job, but: much better and with a much wider range of possibilities.
I'm so exited!
Capture harmonic tracks with various celeb voices and mix them together with the melody track for some all-star duets, trios, and choir pieces. Put celebs to work singing backup to YOU. The sky's the limit.
Wow that is insane! We have great vids too
It is insane, man! 🔥
Great video dude! You react the same way I do when something just works and, the AI is a total revelation for me! Thank you!
Great video and after trying this for harmony parts (especially if you're like me and cannot sing harmonies well), is add something like iZotope harmony generator, or Waves or Antares (I have the Antares one) which you can turn the main vocal off and just export the harmony (which usually sounds really bad and digital) and import that into one of these and then re-import the newly created human sounding harmony, sounds really amazing... So thanks for sharing this, realistic harmony parts have always been an issue for me making them sound real with generators, but now they absolutely do pass as real in the mix.
Thats a perfect example of how being creative with this technology opens up new opportunities for everyone! Thanks for sharing :)
Yooo, that beat in the beginning is soooo fucking hardddd! I need thatttt!
This tech is perfect for songwriters in terms of creating song demos, pitching purposes, etc.
Riktigt bra video som vanligt ❤
nice video, I know its a bit old now, but wow is the ai a game changer. From my point of view within the next 5 years any producer will be substituted.and singer. I like it and even with this quality you can use it in commercial music and no one will know. I it will come down to the skill of marketing that is it as writing hits will become easier with every step of the way.
Can we release a music which uses ai vocals on every platform (spotify,apple music etc)????
Good vid! No bs, just being yourself. 👍🏼
Some crazy shit. I'm gonna try this out later...cheers !!