I have just started using ACE Studio to replace my own voice and first impressions are good.

You kind of have to nail your own vocal performance in terms of timing, pitch and enunciation, because anything less and it'll come up with it's own interpretation of what you're singing.

It also does MIDI+text to voice, but I haven't messed with that yet.

This is gonna make life a lot easier (and cheaper) for me. I sing well and precise enough, but my voice just isn't that pleasant to listen to. Replacing it with an AI voice is a matter of seconds and first results are solid!

Musically is very funny but:
AI let you produce your music without the need of other human beings and their bad humour. The things is that someone else wont need you for his album too and...
Albums wont be heard too, They will become uneeded too because anyone would be able to put it out andlisteners will not be enough to hear them. Oversupply anything and prices wiill drop to near zero.

So you standing theory is that listeners choose from everything? Truth is rather than most music can't compete as it is, because it's homogeneous copycat crap. And you think more of that would change what?

What says a piece of music must be like other music? If you can't be unique, maybe not make music to begin with, no matter what you use.

When you are unique, you're not competing on the same terms. Many problems are the same, yes, but many will hate you for not sounding like other music. That's great, because that's how you stand out. To do that and be great is what it's all about.

yep i put some rap in my new Jungle track, im impressed.

You may want to link your thread to the other already-going Ace user threads.

Well, you just killed the entire daw industry, guitar makers, keyboard manufacturers, harmonica factories, outboard gear makers, probably even Behringer....and Gearspace!!

Gee dude.....way to go.

AI seems to empower the individual solitary producer while weakining collective works. Why? well, in a world where cutting costs is the way to survive collective woks are doomed against this.
Myself i consider the best of us is when we collaborate.

Sure, there is no substitute being in the room together and make music in real-time!

But that’s a different discussion. I’m happy that this option exists as it helps me achieve things I could not before.

Like I said in the OP, you still have to nail the vocal in terms of pitch, time and enunciation, because anything less and it will go off the rails quickly.

But we use compression to control the dynamics of our voice, edit out unwanted noises (lips smacking etc.), EQ to change the tone, and Melodyne/VariAudio/AutoTune to improve or fix pitch.

Going one step further to “enhance/augment/change” the sound, is it really that much different? I mean, to my ears it still sounds like me, just better.

Depends on what the artist wants as a result. Overdubbing each track, one at a time, by one guy, is often the stellar approach.

I might chart something out and basic track it with 5 guys present. Or I may play everything myself as one at a time overdubs.

Artist decision.

Leaving aside the effects on music economy market, i believe that this opens a whole new world unthought before widens a lot.
You can, in a near future, even sing in live situation with different voice skins, just like a guitar pedal does or like a synthesizer. Od course latency wont be a problem in a near future.

That sounds like a pipe dream.

Ultimately, how is this any different than using a sample based piano instrument? Or a SWAM modeled instrument? I don't own a piano but I can seem like I do. I don't own a cello, but I can seem like I do. I don't have a great voice, but I can seem like I do.

For me, the proof of the pudding is in the eating. The technology behind how you compose your music is irrelevant to the end user. All they care about is if it's good. I can't sing as well as I'd like. No amount of practice or lessons can extend my range to what I'd like. This type of technology is perfect for me, as my real talent is in song writing.

Now, if you're letting the software write the song and do the parts... well, no harm done. Most likely you don't care enough to make it anyway. You'll get bored quickly and lost in a sea of similarly garbage AI generated music.

Very Impressive ! It's a Hit !

Its gets philosophical and it is the way it should be.
The problem is in how far technology can fake. THe distance between you singing raw, and you singing with eq, compression and reverb... and lately pitch correctioin is big. Ok. Now you simply change your voice. Thats not you anymore. Ok, but why should it be you? Well thats a limit we consider as being born in the 1900s. Is easy to think that in 20 years the singers will change their voice skin for different songs. Not bad isnt it? it opens the door to a whole new unthought world.
Imagine a live performance in which you sing as you but the sound audience hear is the one of a bass baritone, and then you turn into a mezzosoprano. It will require skill too to make a good performance.
Interesting.
Anyway the distance between acoustic reality and computer calculations will be more far each day that passes... since this technology, as you stated, lets you satisfy your desires without taking lessons or practicing.
This whole trip started not today, started the day amplifiers began, when soft voices could be heard, yes... then you go for electric guitars replacing full orchestras, then synthesizers and drum machine replacing drummers.
THE RESULT SO FAR: to me is not that good, kids nowadays cant put out good songs like before. Dont tell me that in any year people complained about kids being stupid: they are indeed not prepared enough to put out music of the quality than the previous eras.
I know someone will come with some exemptions, there must be of course, but it doesnt change the general fact.

I see A.I. becoming the greatest QC turd polishing tool for the Industry to service the millions of songs cranked out each day by the enabled masses. Stream ON ! Who wants to sound like shit when you can sound foondastik !

No demo or monthly plans to try it out? $200 up front? Nah.

This thing, as much as the infamous Ghibli style drawings, do nothing but empowering the unskilled and defeating the ones that invested time studying and getting better.

The dark side of it is also that while everyone will put out his self percieved **piece of art rendered by an AI** no one will care cause there will be more songs that people willing to hear the noise. People still believe that uploading and album makes them possibly famous, that drives the whole dream and activity for which someone hires a mixer or a session arranger or whatever. Without that dream i doubt that people will put out albums as before.

I don't get this. Has someone told you this or is it that you think that your voice isn't pleasant to listen to? If you're on pitch, then I can't grasp any of this, it doesn't make sense.

Most people I know can't stand their voice, even amazing singers I know struggle with listening back to their own voice, it's just something you have to eventually get over.

If it's the case that you need artificial aids like AI voice replacement and the likes, the only person you're fooling is yourself, it's not you. Find a teacher and learn to sing and enjoy the experience, you'll feel better for it and you're being honest about your singing voice which should seep into your musical works.

Be you, not some fake idealistic representation of you.

Sorry to be *that guy*

I plan on signing up/trying the Ace thing.

I have a couple of hours of studio tracks of isolated vocals of a friend of mine who passed away a couple of decades ago.

Anyone have tips as to best-use/procedures of Ace training steps? I haven't visited the actual site yet.

Is it generally a two-step process?....step 1 to upload the voice to be modeled for training...and step 2 to then reconfigure a set of melody+vocal (sung by me) to then be substituted with his now-modeled voice? Which I then download to insert into my track?

I would imagine that to be the gist of the process.

I imagine some people have encountered best approaches for best results.

Yeah, that is a lot of Captain Obvious.

Horses for courses, etc.

You haven't answered any of my questions that I posted, I'm intrigued to know what the answers are.

There was only one question, why I don’t like my voice or if people tell me that. Yes, it’s both. It’s just not a great sound.

I’m competent at layering BGV’s, nobody hates those.

I had one EDM song and the AI rendering of my lead vocal really worked.

Now, you seem very opinionated on AI vocals, so I assume you have tried it yourself? What did not work for you?

Post up an example of your voice so we can be the judges of your voice.

I've no idea what BGV's are. I'll take a guess and say background vocals?

But you said that you have to sing in tune for AI to work which other than changing your sung lines with different tonal qualities making the voice sound like a different person, I really am struggling to grasp the need to use AI if the user can sing in tune. Makes zero sense to me.

I can sing so I've got no interest in using AI. I've used Melodyne some time back and think it's cheating and won't use it again, it also makes stuff sound sterile and dead, nothing wrong with mistakes and mishaps. Sure, you can use AI and the rest of it, but when it comes to playing stuff live and you sound like arse, you'll fall flat on your face.

Like I said, go and get singing lessons and stop using AI or any other such nonsense as a crutch.

Just to add:

I'm not trying to be mean, harsh or flippant, just getting to the crux of the matter without sugar coating it. There's really no spitefulness meant on my part

Mmmm, sugar-coated crux matter. yum yum

You can be the judge? Of my voice? Anonymously on a public forum? If only more people knew!

I am surrounded by talented people, and while this town is known for sugarcoating, I think I have a pretty realistic idea of my abilities.

But if I do decide to release that song, I'll come back and post a link to it here. Maybe even with OG voice for comparison.

Yes, AFAIK it is a fairly standard acronym.

It doesn't make sense to you because of your assumptions.

You step in here with your mind already made up, offering uninformed opinions about something you have no experience with.

Why not try it out first, and then come back here to tell us why it sucks.

So you have no interest in AI vocals, but you already know it is no good because you have used Melodyne in the past? Got it.

Again you assume things. I play keyboards live all the time. It is my main income. I'll be fine. If I fall flat on my face, it won't be the first time, and it won't be because I used AI vocals on a track.

As an aside, there is whole generation of autotune rappers that have made careers out of it, performing live. It's not really for me, but many of them make way more money with it than I do.

Again, assumptions. Why do you think I never took singing lessons?

You are entitled to your opinion, but if you haven't used and do not plan to use AI voices, if you are here only to convince me (and others) that using AI vocals is wrong, I have no interest in discussing that further.

I simply posted about my experience with replacing my own vocal with an AI model and that I was pleasantly surprised by the result. That is as far as I have gotten with it.

That's an interesting thought.

TC Helicon already makes live voice processors with real-time Autotune built-in.

Of course, different AI voice skins would require a LOT more processing power. And as of now, it won't be able to do that in real-time, forget about low-latency.

But I think that is also because developers aren't particularly concerned about real-time or low latency.

Same way everybody in Silicon Valley just knew you need $20B worth of server farms to power LLM's, until DeepSeek showed you can do it for a fraction of the cost and now they are all scrambling to adjust.

Likewise, It's only a matter of time before someone builds a Vocal AI that can do all that in real-time.

People who is in favour of AI voice singing to be coherent should accept and rejoice when they meet someone on a dating app and he/she used heavy filters to take away 15 years from their face and 15kg from their weight.

(:

Will the woke police censor my message?

If you're singing in pitch but don't like your voice as a whole, why not work with your voice and try to iron out what it is that's not so great. Like I've already mentioned, a teacher would be of great benefit, failing that, maybe read some books on it or learn from YouTube and elsewhere.

Why not keep unsolicited misguided advice to yourself? A teacher would be of great benefit, failing that, maybe read some books on it or learn from YouTube and elsewhere