AI Singing vs Vocal Instrument VST

[LowendG]

Crazy how fast this stuff is progressing. Here's one where you just input audio and it matches it--lyrics and melody--with a "singer" of your choice: https://www.youtube.com/@AuribusMusic/videos

 

[Pirate Captain]

kind of with the OP on this one? Where do you draw the line? I know keyboard players who are still salty about arpeggiators and step sequencers. how is using AI until you find something you like any different than hiring Max Martin or Jack Antinoff to slightly rewrite the same tunes they always do? Reznor/Ross as well for that matter? Or sampling? Or Jimmy Page just straight up stealing blues riffs and whole songs? Make music, have fun.

 

[JRA]

@SaucyJackBass, i hope you get some ideas and/or assistance for the music you're working up --- i'd love to hear it!

i think there are some nifty suggestions thus far, but the blowback against the 'AI thing' may sour your thread into something that you weren't expecting.

in any event, good luck with your project!

 

[MuseChaser]

I don't know of any easy to use AI implementations that will accept MIDI/text as inputs and spit out good sounding vocals. It's really only a matter of time until that happens though. Musicians freaking out about it, like in this thread, isn't going to change that fact. Adapt or get left behind.

The "adapt or be left behind" statement gets thrown about in many discussions, with the implication being that those who choose not to adopt the newest methodology are somehow inferior, or "behind," while all the smart cool kids are moving "forward," I guess.

Has the ease of instantaneous communication ("social media") moved us forward as a kinder caring society, or backwards? Are those who've chosen not to use it falling behind and missing out on great life experiences while those who consume it frequently lead us forward to greatness?

I don't know. "Adapt or be left behind." I can think of a lot of things folks "adapt" to that I'd choose to leave behind. Heroin. Living on fast food. Heck, living in heavily populated areas... I could adapt, but choose not to. Does that make me somehow backwards?

 

[FRoss6788]

kind of with the OP on this one? Where do you draw the line? I know keyboard players who are still salty about arpeggiators and step sequencers. how is using AI until you find something you like any different than hiring Max Martin or Jack Antinoff to slightly rewrite the same tunes they always do? Reznor/Ross as well for that matter? Or sampling? Or Jimmy Page just straight up stealing blues riffs and whole songs? Make music, have fun.

The original artists are all credited in your examples.

 

[ole Jason]

The "adapt or be left behind" statement gets thrown about in many discussions, with the implication being that those who choose not to adopt the newest methodology are somehow inferior, or "behind," while all the smart cool kids are moving "forward," I guess.

Not really. The only implication from the statement is that it's futile to be boneheaded about it and lash out angrily every time someone brings up the topic of AI. In the future, someone WILL create an implementation that will accept inputs and spit out vocals 'in the style' of whomever. It will happen and there is nothing we can do about it, whether we approve of it or not is irrelevant. THAT is the reality that you must adapt to. It doesn't matter if you decide to use AI to augment your creations.

There is no such thing as a luddite in music. You use the tools available to you to create the art you want to create. Nowhere is it required that you use the newest gear or tools.

 

[MattZilla]

@SaucyJackBass, i hope you get some ideas and/or assistance for the music you're working up --- i'd love to hear it!

i think there are some nifty suggestions thus far, but the blowback against the 'AI thing' may sour your thread into something that you weren't expecting.

in any event, good luck with your project!

Paragraphs could be written here but this is it, really:

It'll take over soon enough, no point helping it along

 

[SaucyJackBass]

I don't know of any easy to use AI implementations that will accept MIDI/text as inputs and spit out good sounding vocals. It's really only a matter of time until that happens though. Musicians freaking out about it, like in this thread, isn't going to change that fact. Adapt or get left behind.

Regardless, for your purposes check out EmVoice. It doesn't sound super realistic but should be good enough for your purposes.

Thank you. Perfect.

 

[superheavyfunk]

Ugh. People can’t help but lean ever harder into the SkyNet scenario… aren’t we all supposed to be creative thinkers as musicians? Is it really that hard to see other options for how this all plays out? Some of you really remind me of the folks who complained about the radio killing off the music profession because nEw tEchNOlogy WilL dEstROy eVerYThinG!!!!!

 

[FugaziBomb]

Ugh. People can’t help but lean ever harder into the SkyNet scenario… aren’t we all supposed to be creative thinkers as musicians? Is it really that hard to see other options for how this all plays out? Some of you really remind me of the folks who complained about the radio killing off the music profession because nEw tEchNOlogy WilL dEstROy eVerYThinG!!!!!

I definitely try to stay cognizant of this, and I can absolutely be guilty of this, such as in my preference for mic'd cabs over IRs, but there's definitely something that seems sinister about AI and it's seeming rapid ubiquity in our daily lives. Even though I may not be able to express specifically what it is that troubles me, I still find it troubling.

 

[JRA]

I definitely try to stay cognizant of this, and I can absolutely be guilty of this, such as in my preference for mic'd cabs over IRs, but there's definitely something that seems sinister about AI and it's seeming rapid ubiquity in our daily lives. Even though I may not be able to express specifically what it is that troubles me, I still find it troubling.

 

[MuseChaser]

Not really. The only implication from the statement is that it's futile to be boneheaded about it and lash out angrily every time someone brings up the topic of ....

I deliberately cut out "AI" as the last word of that sentence. Regardless of the topic being discussed, lashing out angrily and being boneheaded is never productive, and I don't believe either of those descriptors are accurate in describing my responses in this thread, or hopefully anywhere else. Having said that, voicing one's disagreement or aversion, regardless of topic, can either be seen as obstructionist by those who disagree or as cautious rationality by those who agree. There are PLENTY of topics that, futile or not, on which we should all constantly speak.

In the future, someone WILL create an implementation that will accept inputs and spit out vocals 'in the style' of whomever. It will happen and there is nothing we can do about it, whether we approve of it or not is irrelevant. THAT is the reality that you must adapt to. It doesn't matter if you decide to use AI to augment your creations.
....

It does matter. Very much. Or it doesn't matter. At all. Depends upon one's frame of reference. This is going to be a long ramble. Hopefully, a point or two will emerge... who knows.

I'm 63, and I've been making music professionally since the age of 15 and taught music in public schools and public and private universities for 35 years. In that time, I've seen a quantum shift in how music is taught, and yes, absolutely, when I retired from public school teaching I was pretty much the last hold-out to refuse to use pop music, especially current pop music, in the classroom or with my orchestras. The other general music teacher used pretty much nothing BUT pop, and her kids spent 90% of their class time staring at laptops using GarageBand to "compose" music by assembling loops. My colleagues all used a lot of current music with the bands and orchestras, because "the kids like it." I always considered my job, as a music educator, to be to educate; to expose students to music they DIDN'T know, to help them understand the complexities therein, to show them the rewards of long-term commitment, to be able to process an incredibly amount of simultaneous data in real time via several senses (sight, hearing, touching), to improve their math skills (try teaching a sixth grader how to play half note triplets in 4/4 time accurately), language skills, physics (harmonic series? acoustics? arco technique? venturi? ... it's endless), interpersonal relationships (getting along and working together with everyone in the orchestra for a common good)... I could go on and on (Ok, I just did.. but could go much longer easily).

I only taught general music for two years back in the 80s before focusing on orchestral, band, and jazz teaching full time. I taught 7th and 8th graders; the grades that most teachers dread teaching general music to. By the end of 8th grade, my students could all read standard music notation in three clefs (treble, bass, and alto), knew all the key signatures (both minor and major), could form major, minor, dominant, diminished, and half-diminished seventh chords correctly, knew basic dynamic and expression terminology, could identify most instruments by sight and by ear, and which were concert pitch, Bb, Eb, or F transposition, knew the main musical eras dating back to antiquity and could make informed guesses as to what era a given piece was from by ear, knew composers from each era and a bit about their lives, and... on... and... on. For their final project in 8th grade, each student had to write a piece of music for at least three different instruments, at least one of which had to be a transposing instrument, and the piece had to have an exposition, development, recapitulation, and coda, and they had to assemble and rehearse their ensemble (they could conduct it if they didn't play an instrument well enough although everyone in class learned the basics of piano and baroque recorder, or they could play in their ensemble) and perform their piece for the class. In their two semesters in class (one semester in 7th, one in 8th), they heard opera, string quartets, orchestras, barbershop singing, jazz from its earliest roots to avant-garde (the Max Roach/Anthony Braxton album "Rebirth" was always a big hit!), and a bit of reggae, James Brown, and various other genres.

Those kids learned about music, developed an appreciation for all kinds of styles, and also developed a critical ear; their tastes evolved over those years, and yes, I'll say it... they listened to better, higher quality, more thoughtful music than their peers. It mattered.

Fast forward to the last few years I spent in public school with the GarageBand laptop colleague. Her kids learned basically nothing about music. They learned how to assemble sounds by pressing keys on a laptop. They listened to nothing but their favorite recordings of current artists she had them bring in daily. Did their tastes change? Did their appreciation for the art evolve? It mattered.

To distill what I'm trying to say (and admittedly, quite poorly), it matters. The more we dumb down our art, the dumber our art becomes, and the dumber we all become. The less we do for ourselves, the less we'll be able to do. If we don't teach, REALLY teach, music, then no one will be left who can hear the difference, or who cares about the difference, between human art and CGM (Computer Generated Music). To some, that may not matter. To me, it does. To some, there is nothing we can do about the increasing removal of human performance from music and it's futile to resist it. To me, there IS something I can try and do about it, futile or not, so I'll try. Kind of like the ol' "starfish" story... "I made a difference to THAT one."

It matters.

 

[FRoss6788]

I deliberately cut out "AI" as the last word of that sentence. Regardless of the topic being discussed, lashing out angrily and being boneheaded is never productive, and I don't believe either of those descriptors are accurate in describing my responses in this thread, or hopefully anywhere else. Having said that, voicing one's disagreement or aversion, regardless of topic, can either be seen as obstructionist by those who disagree or as cautious rationality by those who agree. There are PLENTY of topics that, futile or not, on which we should all constantly speak.


It does matter. Very much. Or it doesn't matter. At all. Depends upon one's frame of reference. This is going to be a long ramble. Hopefully, a point or two will emerge... who knows.

I'm 63, and I've been making music professionally since the age of 15 and taught music in public schools and public and private universities for 35 years. In that time, I've seen a quantum shift in how music is taught, and yes, absolutely, when I retired from public school teaching I was pretty much the last hold-out to refuse to use pop music, especially current pop music, in the classroom or with my orchestras. The other general music teacher used pretty much nothing BUT pop, and her kids spent 90% of their class time staring at laptops using GarageBand to "compose" music by assembling loops. My colleagues all used a lot of current music with the bands and orchestras, because "the kids like it." I always considered my job, as a music educator, to be to educate; to expose students to music they DIDN'T know, to help them understand the complexities therein, to show them the rewards of long-term commitment, to be able to process an incredibly amount of simultaneous data in real time via several senses (sight, hearing, touching), to improve their math skills (try teaching a sixth grader how to play half note triplets in 4/4 time accurately), language skills, physics (harmonic series? acoustics? arco technique? venturi? ... it's endless), interpersonal relationships (getting along and working together with everyone in the orchestra for a common good)... I could go on and on (Ok, I just did.. but could go much longer easily).

I only taught general music for two years back in the 80s before focusing on orchestral, band, and jazz teaching full time. I taught 7th and 8th graders; the grades that most teachers dread teaching general music to. By the end of 8th grade, my students could all read standard music notation in three clefs (treble, bass, and alto), knew all the key signatures (both minor and major), could form major, minor, dominant, diminished, and half-diminished seventh chords correctly, knew basic dynamic and expression terminology, could identify most instruments by sight and by ear, and which were concert pitch, Bb, Eb, or F transposition, knew the main musical eras dating back to antiquity and could make informed guesses as to what era a given piece was from by ear, knew composers from each era and a bit about their lives, and... on... and... on. For their final project in 8th grade, each student had to write a piece of music for at least three different instruments, at least one of which had to be a transposing instrument, and the piece had to have an exposition, development, recapitulation, and coda, and they had to assemble and rehearse their ensemble (they could conduct it if they didn't play an instrument well enough although everyone in class learned the basics of piano and baroque recorder, or they could play in their ensemble) and perform their piece for the class. In their two semesters in class (one semester in 7th, one in 8th), they heard opera, string quartets, orchestras, barbershop singing, jazz from its earliest roots to avant-garde (the Max Roach/Anthony Braxton album "Rebirth" was always a big hit!), and a bit of reggae, James Brown, and various other genres.

Those kids learned about music, developed an appreciation for all kinds of styles, and also developed a critical ear; their tastes evolved over those years, and yes, I'll say it... they listened to better, higher quality, more thoughtful music than their peers. It mattered.

Fast forward to the last few years I spent in public school with the GarageBand laptop colleague. Her kids learned basically nothing about music. They learned how to assemble sounds by pressing keys on a laptop. They listened to nothing but their favorite recordings of current artists she had them bring in daily. Did their tastes change? Did their appreciation for the art evolve? It mattered.

To distill what I'm trying to say (and admittedly, quite poorly), it matters. The more we dumb down our art, the dumber our art becomes, and the dumber we all become. The less we do for ourselves, the less we'll be able to do. If we don't teach, REALLY teach, music, then no one will be left who can hear the difference, or who cares about the difference, between human art and CGM (Computer Generated Music). To some, that may not matter. To me, it does. To some, there is nothing we can do about the increasing removal of human performance from music and it's futile to resist it. To me, there IS something I can try and do about it, futile or not, so I'll try. Kind of like the ol' "starfish" story... "I made a difference to THAT one."

It matters.

Very well expressed.

Thank you.

 

[JRA]

here's a perfectly good, new thread for discussing the undercurrents (tangents) in this thread (about which the OP had no interest or intention):

https://www.talkbass.com/threads/ai-in-the-arts-especially-the-musical-ones.1640298/

 

[ole Jason]

I deliberately cut out "AI" as the last word of that sentence. Regardless of the topic being discussed, lashing out angrily and being boneheaded is never productive, and I don't believe either of those descriptors are accurate in describing my responses in this thread, or hopefully anywhere else. Having said that, voicing one's disagreement or aversion, regardless of topic, can either be seen as obstructionist by those who disagree or as cautious rationality by those who agree. There are PLENTY of topics that, futile or not, on which we should all constantly speak.


It does matter. Very much. Or it doesn't matter. At all. Depends upon one's frame of reference. This is going to be a long ramble. Hopefully, a point or two will emerge... who knows.

I'm 63, and I've been making music professionally since the age of 15 and taught music in public schools and public and private universities for 35 years. In that time, I've seen a quantum shift in how music is taught, and yes, absolutely, when I retired from public school teaching I was pretty much the last hold-out to refuse to use pop music, especially current pop music, in the classroom or with my orchestras. The other general music teacher used pretty much nothing BUT pop, and her kids spent 90% of their class time staring at laptops using GarageBand to "compose" music by assembling loops. My colleagues all used a lot of current music with the bands and orchestras, because "the kids like it." I always considered my job, as a music educator, to be to educate; to expose students to music they DIDN'T know, to help them understand the complexities therein, to show them the rewards of long-term commitment, to be able to process an incredibly amount of simultaneous data in real time via several senses (sight, hearing, touching), to improve their math skills (try teaching a sixth grader how to play half note triplets in 4/4 time accurately), language skills, physics (harmonic series? acoustics? arco technique? venturi? ... it's endless), interpersonal relationships (getting along and working together with everyone in the orchestra for a common good)... I could go on and on (Ok, I just did.. but could go much longer easily).

I only taught general music for two years back in the 80s before focusing on orchestral, band, and jazz teaching full time. I taught 7th and 8th graders; the grades that most teachers dread teaching general music to. By the end of 8th grade, my students could all read standard music notation in three clefs (treble, bass, and alto), knew all the key signatures (both minor and major), could form major, minor, dominant, diminished, and half-diminished seventh chords correctly, knew basic dynamic and expression terminology, could identify most instruments by sight and by ear, and which were concert pitch, Bb, Eb, or F transposition, knew the main musical eras dating back to antiquity and could make informed guesses as to what era a given piece was from by ear, knew composers from each era and a bit about their lives, and... on... and... on. For their final project in 8th grade, each student had to write a piece of music for at least three different instruments, at least one of which had to be a transposing instrument, and the piece had to have an exposition, development, recapitulation, and coda, and they had to assemble and rehearse their ensemble (they could conduct it if they didn't play an instrument well enough although everyone in class learned the basics of piano and baroque recorder, or they could play in their ensemble) and perform their piece for the class. In their two semesters in class (one semester in 7th, one in 8th), they heard opera, string quartets, orchestras, barbershop singing, jazz from its earliest roots to avant-garde (the Max Roach/Anthony Braxton album "Rebirth" was always a big hit!), and a bit of reggae, James Brown, and various other genres.

Those kids learned about music, developed an appreciation for all kinds of styles, and also developed a critical ear; their tastes evolved over those years, and yes, I'll say it... they listened to better, higher quality, more thoughtful music than their peers. It mattered.

Fast forward to the last few years I spent in public school with the GarageBand laptop colleague. Her kids learned basically nothing about music. They learned how to assemble sounds by pressing keys on a laptop. They listened to nothing but their favorite recordings of current artists she had them bring in daily. Did their tastes change? Did their appreciation for the art evolve? It mattered.

To distill what I'm trying to say (and admittedly, quite poorly), it matters. The more we dumb down our art, the dumber our art becomes, and the dumber we all become. The less we do for ourselves, the less we'll be able to do. If we don't teach, REALLY teach, music, then no one will be left who can hear the difference, or who cares about the difference, between human art and CGM (Computer Generated Music). To some, that may not matter. To me, it does. To some, there is nothing we can do about the increasing removal of human performance from music and it's futile to resist it. To me, there IS something I can try and do about it, futile or not, so I'll try. Kind of like the ol' "starfish" story... "I made a difference to THAT one."

It matters.

I recognize that is your perspective and acknowledge you're coming at this from a point of good faith but I don't find the argument all that compelling. Not in the context of what OP is asking for, which is arguably a poor use case for AI to begin with, or the overarching general possibilities of using AI to augment writing, producing, or learning music. I also don't find the argument particularly relevant to the point I was making. When I said it doesn't matter I meant it quite literally, it literally does not matter whether you or I approve of the use of AI in music, it's happening and there is nothing we can do to stop it.

I do find your points about pop music potentially problematic, and forgive me if I've misunderstood you as can sometimes happen given the brevity of online communication. I have never been a fan of the approach of only acknowledging/teaching some forms or genres of music while completely omitting or disregarding others. As an example, earlier this morning I was watching a video about applying the rhythmic approaches of J Dilla to orchestral music. If we are to ignore the music of J Dilla, which in this context would fall into pop music, we'd never have the discussion on the merits of applying it to orchestral music. It wasn't so long ago that the same sort of thing was happening in regards to jazz. Your experience with the Garageband colleague speaks more to the failure of that person to be a good teacher more than anything IMO.

 

[SaucyJackBass]

I don't know of any easy to use AI implementations that will accept MIDI/text as inputs and spit out good sounding vocals. It's really only a matter of time until that happens though. Musicians freaking out about it, like in this thread, isn't going to change that fact. Adapt or get left behind.

Regardless, for your purposes check out EmVoice. It doesn't sound super realistic but should be good enough for your purposes.

This worked out perfect. It gives (crappy) vocals to the track the allows me to really play with lyrics and phrasing and fit.

Thank you for the recommendation.

 

[callofcthulhu]

Does that make me somehow backwards?

Frankly, yes.

Not that there's anything "wrong" with that. Do your thing, own it, be happy about it. But I suspect you know you're setting yourself up for a kind of bittersweet--and potentially lonely--happiness, if you point yourself against the current. Two roads diverging in the wood and all that.

 

[MuseChaser]

Frankly, yes.

Not that there's anything "wrong" with that. Do your thing, own it, be happy about it. But I suspect you know you're setting yourself up for a kind of bittersweet--and potentially lonely--happiness, if you point yourself against the current. Two roads diverging in the wood and all that.

.

Wow.
Just for context, here's the entire paragraph containing the snippet you quoted ..

"I don't know. "Adapt or be left behind." I can think of a lot of things folks "adapt" to that I'd choose to leave behind. Heroin. Living on fast food. Heck, living in heavily populated areas... I could adapt, but choose not to. Does that make me somehow backwards?"

Soo.. living outside cities, avoiding heroin, and eating healthy makes me backwards?

Wow.

 

[callofcthulhu]

.

Wow.
Just for context, here's the entire paragraph containing the snippet you quoted ..

"I don't know. "Adapt or be left behind." I can think of a lot of things folks "adapt" to that I'd choose to leave behind. Heroin. Living on fast food. Heck, living in heavily populated areas... I could adapt, but choose not to. Does that make me somehow backwards?"

Soo.. living outside cities, avoiding heroin, and eating healthy makes me backwards?

Wow.

Honestly, it's more the fact that you equate the 3 (and felt them relevant to a conversation about music and technology) that makes you backwards.

Again, nothing wrong with being backwards. You seem to be of the view that modern culture is tantamount to vice, so go ahead and take your "backwardness" as virtue.

 

[FilterFunk]

CGI effects in movies have made pretty much ANYTHING possible, and a lot of current CGI stuff can appear startlingly real. BUT.... the achievement, at least to me, is nowhere near as mindboggling and impressive as the actual creation of huge sets, "cast of thousands" scenes, amazing stunt work, etc., when they were actually DONE by people. Plus, CGI hasn't added to the poignancy and quality of the storylines; a pretty cogent argument could be made that it's had the opposite effect. It's become the focus of so many movies now, instead of the craft of truly skilled human actors who spend their lives perfecting their art.

I think that, while CGI is certainly overused, movies like the LOTR series are greatly enhanced by CGI. As much as I love Ray Harryhausen stop-motion effects, and they were the right tool for the movies at the time, the judicious use of CGI is often the right tool for today. Having said that, I hope no one ever tries to remake any more Harryhausen movies - it's bad enough that they remade "Clash of the Titans"; they'd better not touch "Jason and the Argonauts" or any of the "Sinbad" movies!

Now I agree with you about "the actual creation of huge sets, 'cast of thousands' scenes, amazing stunt work, etc., when they were actually DONE by people." CGI has not adequately replaced those aspects of moviemaking. But if things like CGI, Pro Tools, AI, Photoshop, etc. are used to enhance creative content and not be the content itself, I'm okay with it. I also like the idea of such tools being used to help create demos for people who have good ideas but just don't have the ability or resources to do it themselves. Hiring people or learning the skills themselves are not always viable options.