MY SAMPLEPACKS & SOUNDBANKS (New Vital Soundbank)! ⬇️
💥 Samplepacks & Soundbanks: https://www.mhamusic.com/ *

Musicfy AI Vocals: https://create.musicfy.lol/?ref=mha *

Lalals AI Vocals: https://lalals.com/ref/10/?campaign=mha *

MY FAVORITE VOCAL PACKS! ⬇️
Soundle: https://transactions.sendowl.com/stores/16517/249628 *

My latest remix! Please go listen to it! :D (SKIO)⬇️
https://skiomusic.com/r/6k23 

🔥 Follow me on Instagram, Spotify & Twitch: https://linktr.ee/mhamusic

🎧 Join our discord! https://discord.gg/a5EaTYp

Join Epidemic Sound for some awesome tracks for your videos!
https://www.epidemicsound.com/referral/ct1opg/  *

* Affiliate links

You now have the possibility, if you're a guy, you can make female vocals.

And if you're a female, you can make guy vocals.

Like, is it just me or is this kid crazy?

A few of you have been asking me if I'm dead.

But I just had a lot to do in school and now I've taken 2 weeks off and now I'm back.

There is 1 thing with AI that has been grabbing my attention lately.

You've all seen the memes on TikTok with Joe Biden and Donald Trump and Obama playing Minecraft together.

It's funny as hell, but does it have any real world application?

This is a little test project that I have.

This is my voice ran through an AI algorithm thing.

Don't stop to look back this is not to fight.

We're running out of time Hide don't stop to look back.

We're running out of time Hi, don't stop to look back.

I mean we've all seen the Drake AI thing by now So we know that if you want to put Drake on a beat that you made that's entirely possible But if you want to put the weekend on your pop track, then you can do that.

It's absolutely crazy I will show you how to do it in this video.

This 1 I found on TikTok and also Lalals that actually reached out to me.

So I have affiliate links for both of these sites.

If you want to try it out for yourself, I will link them in the description.

MusicFi, they have these general artists or these general voice models.

There's also an ability to train your own AI voice.

But Lalal's, they have like celebrity voice models.

They also have like Donald Trump and Barack Obama and stuff like that if you want to do some memes.

🎵Time🎵 🎵We're running out, we're running out of time🎵 Is

🎵Time🎵 🎵We're running out, we're running out of time

Honestly, I'm very impressed with MusicFi's quality on their models.

I'm just gonna record something very quick.

We'll go into a new project in FL Studio.

We can just drag in a loop from Guitar Sessions while we, volume 2, we'll just do this 1.

And we will sing something very basic on top of this.

You take me higher than I've ever felt before.

Okay, it's also in the morning, so I will have a hard time singing this, but the lyrics are You take me higher You take me higher than I've ever felt before We are the fighters that will never give it up Whoa, oh, 0 So I'm gonna try to record something that it can sound like shit, honestly It's almost better if it sounds like shit.

Sitting down, I woke up an hour ago, my voice is like all trembling already, just speaking, So this will be incredibly hard to sing.

You take me higher than I've ever felt before

We are the finest that will never give it up

Honestly, we'll just keep it that way, because it's better if it's like a bit bad.

I'm just gonna leave it completely unprocessed.

The only thing I'll do is I'll add an auto-tune.

Take me higher than I've ever felt before We

are the fighters that will never give it up

I'm cringing so hard this it's probably not that bad, but so we'll just Export this the way it is we can upload that file to Music file first here.

I really like the Nina 1 I don't know I don't know if a guy's voice would fit better or a or a girl's voice.

So we can try to do like the maybe the Dorian 1 We can just keep it on demo quality because it's honestly better than the master 1.

We click remix It just takes a few seconds and then you have it down here.

This is what it sounds like completely raw

You take me higher than I've ever felt before We are the finest that will never give it up

you take me higher than I've ever felt before

You can hear some artifacts in there I would say in general this Dorian model that I used for this 1 has a little bit more artifacts than the Nina Model the female 1 that I really like

Like honestly if you mix that in with this guitar here you Take

And that was take 1 of my original voice like if I would actually warm up my voice Maybe really try to to make it sound as good as possible From the start and then go through the the model changing thing would probably at least sound a little bit better But it does have some artifacts it's kind of sounds like if you're if you have 1 of these mics and you have like a cloud lifter, like a gain booster, and it's like a little bit janky, you can sometimes get exactly this kind of artifacts on top.

I'm sure you can remove that like somehow.

Or at least make it sound a little bit better.

But that was the Dorian model from MusicFi.

Now let's go to Lala's and we can try, I don't know which 1 would be best, maybe The Weeknd honestly, we could just try that 1.

All right here we go, we can take a listen.

Not a great model for for that kind of singing.

I want to try something else Let's try an edge here And that's the thing like you have to sort of know what model you're gonna use to begin with and then sing in a way that you think will fit that model the best.

So it takes like quite a lot of trial and error to get it to sound good.

We can also do a Tate McRae version and we'll see how it handles my lower register when I'm singing my low register high if that makes sense and I'm not singing in falsetto I believe it's called and you sing like

this you take me higher than I've ever felt before

yeah so since I'm singing in my lower register She is also singing in her lower register And I don't think the model is trained on like her lower register if that makes sense so I would have to like sing in falsetto For for this to work like really well.

We are the finest that will never give it

You can hear when I go up almost in falsetto, it sounds a lot better over here.

And all these models will have different artifacts, you know, depending on how they were trained, on what material they were trained on.

So they will all be of different quality and they will all have like different artifacts and different pros and cons to them that you will kind of have to work around as best you can.

So this is what the Ed Sheeran 1 sounds like completely dry from La La's.

We can link it to track 3, the same 1 as this 1.

We are the fighters that will never give it up.

I have been testing and like doing some trial and error stuff over the past few weeks really and I have to say the results are very varied but once you do some testing yourself, you kind of see what works and what doesn't work and what you have to change in order to get the best results possible.

Most of all, it's really fun to play around with and you can get some insane results.

Like you can get some really crazy results.

And the biggest thing for me is I've always had a thing for like female vocals and they're so hard to come by like good female vocals Soundhole has some really good packs but like Splice there is everything is so overused you now have the possibility if you're a guy you can make female vocals and if you're a female you make guy vocals guy vocals that sounds weird but you know what I mean and maybe it's not perfect but like in the context of a whole mix, it could sound pretty good.

I've been planning on making a vocal sample pack for over a year.

I've been writing some stuff, you know, on and off and recording quite a lot myself.

But with this technology, I don't know, like, would it be weird to like sell a sample pack that has like AI voices?

It's a little bit weird, but I guess it's kind of the same as a sample pack.

You can just make the samples yourself, but you choose not to.

It's easier to buy a sample pack from someone that knows what they're doing.

Like the songwriting is the biggest part, I guess.

So let me know what you would think of that.

Would it be weird if I like sold you a sample pack with vocals that you could theoretically make yourself?

As long as the quality is good I don't really see the problem but I'd love to hear your thoughts on that.

I'd also love to hear your thoughts on this AI technology in general.

There is definitely a shift happening in the music industry.

I wish you guys a fantastic weekend and I'll probably see you again next Friday.

Have a great weekend guys, let me know what you think in the comments.


Speaker 0: You now have the possibility, if you're a guy, you can make female vocals. And if you're a female, you can make guy vocals. Like, is it just me or is this kid crazy? Hello guys! I'm finally back.

A few of you have been asking me if I'm dead. I think I'm alive. But I just had a lot to do in school and now I've taken 2 weeks off and now I'm back. There is 1 thing with AI that has been grabbing my attention lately. And that is all of these voice changers.

You've all seen the memes on TikTok with Joe Biden and Donald Trump and Obama playing Minecraft together. It's funny as hell, but does it have any real world application? Well, I'm about to blow your minds. So we're inside of a studio. This is a little test project that I have.

This is my voice. This is my voice ran through an AI algorithm thing.

Speaker 1: We're running out of time.

Speaker 0: Like is it just me or is this crazy?

Speaker 1: We're running out of time. Don't stop to look back this is not to fight.

Speaker 0: I also have this version

Speaker 2: We're running out of time Hide don't stop to look back. There's no sense of fight

Speaker 0: And we also have the weekend

Speaker 2: We're running out of time Hi, don't stop to look back. There's no sense of fight the weekend.

Speaker 0: I mean we've all seen the Drake AI thing by now So we know that if you want to put Drake on a beat that you made that's entirely possible But if you want to put the weekend on your pop track, then you can do that. It's absolutely crazy I will show you how to do it in this video. So I use 2 websites. This 1 I found on TikTok and also Lalals that actually reached out to me. So I have affiliate links for both of these sites.

If you want to try it out for yourself, I will link them in the description. I earned like a very small commission. MusicFi, they have these general artists or these general voice models. There's also an ability to train your own AI voice. But Lalal's, they have like celebrity voice models.

They have so many voices. They also have like Donald Trump and Barack Obama and stuff like that if you want to do some memes. So this voice...

Speaker 1: 🎵Time🎵 🎵We're running out, we're running out of time🎵 Is

Speaker 0: the Tate McRae model on La La's. And this 1...

Speaker 2: 🎵Time🎵 🎵We're running out, we're running out of time

Speaker 0: Is the Nina voice model from MusicFi. Honestly, I'm very impressed with MusicFi's quality on their models. Like, there's so little artifacts. It's actually Insane. Whatever, let's just get into it.

I'll show you how it's done. I'm just gonna record something very quick. We'll go into a new project in FL Studio. What do we do? We can just drag in a loop from Guitar Sessions while we, volume 2, we'll just do this 1.

And we will sing something very basic on top of this. This is my writing process, okay? I'm sorry. It's very cringe. It's hard to do this while on camera.

Let's have some lyrics, very generic.

Speaker 3: You take me higher than I've ever felt before. We are the fighters,

Speaker 0: that will never... Okay, it's also in the morning, so I will have a hard time singing this, but the lyrics are You take me higher You take me higher than I've ever felt before We are the fighters that will never give it up Whoa, oh, 0 So I'm gonna try to record something that it can sound like shit, honestly It's almost better if it sounds like shit. Sitting down, I woke up an hour ago, my voice is like all trembling already, just speaking, So this will be incredibly hard to sing.

Speaker 3: You take me higher than I've ever felt before

Speaker 4: We are the finest that will never give it up

Speaker 1: Woo!

Speaker 0: First take. Honestly, we'll just keep it that way, because it's better if it's like a bit bad. I'm just gonna leave it completely unprocessed. The only thing I'll do is I'll add an auto-tune. So it sounds like this raw.

Speaker 3: Take me higher than I've ever felt before We

Speaker 4: are the fighters that will never give it up

Speaker 0: I'm cringing so hard this it's probably not that bad, but so we'll just Export this the way it is we can upload that file to Music file first here. We have it and then we can try. I really like the Nina 1 I don't know I don't know if a guy's voice would fit better or a or a girl's voice. So we can try to do like the maybe the Dorian 1 We can just keep it on demo quality because it's honestly better than the master 1. We click remix It just takes a few seconds and then you have it down here.

We'll click download. This is what it sounds like completely raw

Speaker 5: You take me higher than I've ever felt before We are the finest that will never give it up

Speaker 0: So the vocal on itself

Speaker 5: you take me higher than I've ever felt before

Speaker 0: You can hear some artifacts in there I would say in general this Dorian model that I used for this 1 has a little bit more artifacts than the Nina Model the female 1 that I really like

Speaker 5: you take me higher than I've ever felt before

Speaker 0: Like honestly if you mix that in with this guitar here you Take

Speaker 1: me higher than I've

Speaker 0: ever felt before Like honestly, if you

Speaker 1: mix that in with this guitar here

Speaker 5: You take me higher than I've ever felt before

Speaker 0: And that was take 1 of my original voice like if I would actually warm up my voice Maybe really try to to make it sound as good as possible From the start and then go through the the model changing thing would probably at least sound a little bit better But it does have some artifacts it's kind of sounds like if you're if you have 1 of these mics and you have like a cloud lifter, like a gain booster, and it's like a little bit janky, you can sometimes get exactly this kind of artifacts on top.

Speaker 5: You take me high

Speaker 0: It's like a little hiss. I'm sure you can remove that like somehow. With some kind of plug-in. Or at least make it sound a little bit better. But that was the Dorian model from MusicFi.

Now let's go to Lala's and we can try, I don't know which 1 would be best, maybe The Weeknd honestly, we could just try that 1. All right here we go, we can take a listen. Not a great model for for that kind of singing. I want to try something else Let's try an edge here And that's the thing like you have to sort of know what model you're gonna use to begin with and then sing in a way that you think will fit that model the best. So it takes like quite a lot of trial and error to get it to sound good.

Speaker 5: You take me higher than I've ever felt before.

Speaker 0: That 1 sounds okay. We can also do a Tate McRae version and we'll see how it handles my lower register when I'm singing my low register high if that makes sense and I'm not singing in falsetto I believe it's called and you sing like

Speaker 6: this you take me higher than I've ever felt before

Speaker 0: yeah so since I'm singing in my lower register She is also singing in her lower register And I don't think the model is trained on like her lower register if that makes sense so I would have to like sing in falsetto For for this to work like really well.

Speaker 6: We are the finest that will never give it

Speaker 1: up. 000.

Speaker 0: You can hear when I go up almost in falsetto, it sounds a lot better over here. And all these models will have different artifacts, you know, depending on how they were trained, on what material they were trained on. So they will all be of different quality and they will all have like different artifacts and different pros and cons to them that you will kind of have to work around as best you can. But we have the Ed Sheeran 1. So this is what the Ed Sheeran 1 sounds like completely dry from La La's.

Speaker 5: You take me higher than I've ever felt before.

Speaker 0: Also some artifacts there on the top. We can link it to track 3, the same 1 as this 1. We'll see what happens.

Speaker 5: Before. You take me higher than I've ever felt before. We are the fighters that will never give it up.

Speaker 0: Sounds okay honestly. This is just the surface. I have been testing and like doing some trial and error stuff over the past few weeks really and I have to say the results are very varied but once you do some testing yourself, you kind of see what works and what doesn't work and what you have to change in order to get the best results possible. Most of all, it's really fun to play around with and you can get some insane results. Like you can get some really crazy results.

And the biggest thing for me is I've always had a thing for like female vocals and they're so hard to come by like good female vocals Soundhole has some really good packs but like Splice there is everything is so overused you now have the possibility if you're a guy you can make female vocals and if you're a female you make guy vocals guy vocals that sounds weird but you know what I mean and maybe it's not perfect but like in the context of a whole mix, it could sound pretty good. I've been planning on making a vocal sample pack for over a year. I've been writing some stuff, you know, on and off and recording quite a lot myself. But with this technology, I don't know, like, would it be weird to like sell a sample pack that has like AI voices? Because you could do that yourself.

It's a little bit weird, but I guess it's kind of the same as a sample pack. You can just make the samples yourself, but you choose not to. It's easier to buy a sample pack from someone that knows what they're doing. Like the songwriting is the biggest part, I guess. So let me know what you would think of that.

Would it be weird if I like sold you a sample pack with vocals that you could theoretically make yourself? As long as the quality is good I don't really see the problem but I'd love to hear your thoughts on that. I'd also love to hear your thoughts on this AI technology in general. There is definitely a shift happening in the music industry. I guess that's it for today.

I wish you guys a fantastic weekend and I'll probably see you again next Friday. Have a great weekend guys, let me know what you think in the comments. Bye bye!

Speaker 1: Get

See all MHA transcripts on Youtube

AI vocals will change EVERYTHING for producers..