Anthony is right, besides stealing other cover artists' covers and overlaying their voices, AI did jack shit. Shinoda is more than just a sound effect for Linkin Park. Sinatra always tries to deviate from the melody of a cover. And the backing choir is essential to Queen. But AI wouldn't understand that, only a human who gets nuances and preferences would.
One of my favorite things about covers is the other artists adding their own flourishes. To me, AI just sounds like they're trying to do the song beat by beat in a different voice. Thus I feel the winner is clearly Anthony Vincent.
AI is already at that level, actually how bout yall watch xQc's video, just recently he 'made' songs with an AI tool on stream. I think you would be surprised how amazing the results were...
surprisingly enough i think the linkin park one did it pretty well. not that it matters because we all know that AI art, for instance, can look amazing but that means jackshit considering it's essentially based on plagiarism
@@Steamrick Agreed but since Ai can't "feel" it can just turn patterns into mathematical patterns there will always be something that feels close but not exact
I would like to hear full versions of your interpretations, these were amazing and sounds exacly like something that the bands would do, manly the Queen part
Honestly, the only time AI impressed me musically is when you ran your version of Black Parade in the style of Queen through that Freddie Mercury filter - and that's only because you already did all the heavy lifting by capturing Freddie's style, his cadence, and all the little variations that made the performance so good to begin with. The voice filter filled in that last final gap and made it truly eerie, but it could never have done that without the actual work you'd put into the performance first!
You absolutely won each time. Although I feel like a person would win in most human vs. AI scenarios, you especially are great and your voice is just awesome. I would love to hear a full version of the Toxicity cover.
One thing I noticed on these AI covers is that they can`t change the metrics of the singing line to fit the style properly. It is just the voice that is changed to match another artist and then the rest of the song comes around that. You really put time thinking how a different artist would aproach a song. That is why you win every battle in my opinion ( that and the fact that you do everything with your own voice).
You win on all of them hands down. There’s a warmth to your tone, plus your creativity with the music. AI is doing a “cover”, whereas you are doing the songs “in the style of”, which I prefer. That being said, we need a whole album of various metal songs in the style of Frank Sinatra. Outstanding! Never knew I would want to hear something like that, but now that I have, I want more!! ❤❤❤
Seems like you went for old Lincoln Park and the AI went for new Lincoln Park. I say you still win all around since you're doing that with your own voice and muscles and it's infinitely more impressive as your ability to mimic other singers voices is literally amazing and never ceases to impress
me at this comment for last 30 minutes trying to understand if Lincoln Park is a multiple occured typo or a some joke with a inner meaning that I cant understand due to my language level lol
The Linkin Park cover is really the first *true* AI cover I've seen, where both the instruments and vocals are fully done by an AI (with a lot of human tinkering, of course), and I do have to say, I think the AI had you beat in *that* one, for both accuracy and sound. I don't know if it's a fair comparison to make, since I, like you said, can't imagine that was done by just plugging it into a machine learning algorithm, pumping out something and sticking it on the internet; I'd imagine it was done piece by piece, perhaps even each segment was made by a person, then put through an AI to make it sound more Linkin Park'y, *then* stuck back together. That's how I'd have to imagine it was done for it to sound so good and have such unique touches. The other ones though... don't stand a chance, your Queen cover of Toxicity needs to be a full video!
Since everyone wants to be a 'fan' and say you, I'll play Devil's Advocate. You are talented in ways no one in the world is. But of course an AI song, done in the actual voice of the original artist is going to sound more like the artist than another person trying to replicate that voice. Plus the AI song had more time put into it, while you were just trying basically in real time to do the same thing. That being said, you're the reason I started loving cover songs in the first place on TH-cam. Never ever stop doing what you're doing.
That Paul Anka album is fantastic, by the way. I remember playing it in my car the year it came out and my mom saying something like "I didn't realize some of these songs were so old"
i think the main way people do AI singing is they record themself doing an impression, and then the ai tweaks it to be more like the actual singer. could be fun to see how good AI versions of your videos turn out since your impressions are already so close
I think the only one the ai kind of won at was the Linkin Park, because it doesn't feel like them to me without the rap part also being included. Every other one yours was so much better. Your way of singing in so many styles just blows me away.
Awesome video! This really showcased your talents and musicians in general. Your versions are so emotive and musical, while the AI version are just cheap copies! You win easily😊
@@saysthemanbehindthemaskI think the songs have to fit like a glove or a person has to redo the arrangement to fit other genres. It is not just "good song is auto good in any genre"
I prefer the AI version of Toxicity, but moreso I think that the AI cover creator chose to go along a later Queen approach of slower and softer ballads like "Don't Try So Hard", "Don't Try So Hard", and "I'm Going Slightly Mad", hence the Robin sample. Not sure your Night At the Opera Queen style really fits the somber tone of Toxicity. But I know that's also the sound most people associate with Queen, so solid work trying to make it fit.
For the first two songs, I think you won out over the AI. I enjoyed your versions more, but I didn't hate the AI version either. It was funny to find the Frank Sinatra one was just a slightly tuned Paul Anka version. I don't have that critical of an ear for music, so none of the AI versions bothered me. I do think that the AI cover of the third song was better, but only becuase they just changed the female singers voice slightly to make it sound like Freddie. It was hard to reconcile your version when you were doing a more growly scream... don't know that I have heard much like from him before. Great video, and I would like to echo the other people and say I want a full version of the Linkin Park and Frank Sinatra covers, lol.
I do think that the Ai got you on the Linkin Park song, but only that one. I took into consideration that the volumes were completely different which impacted the feel of the Ai version. I would also imagine that a lot of human intervention took place, but Ai is definitely going to eventually be better. For Gods sake just look at Sora or whatever that video thing is called. It’s insane!! I don’t say these things with glee as I once was a musician in a past life and I poured my heart and soul into my songs, but the Ai is just coming man, and it’s coming fast! I really enjoy your stuff though. Just found you.
Hmm... I have comments that I can give to both for the Linkin Park cover. Yours: Clean sound. Joe's signature DJ input is missing, but it suits the song better. The slower paced, rockier style better suits over the early Nu-metal/grunge style used by the AI, a great choice. Chester's voice tends to be a bit higher when he's singing head voice, but for the most part, spot on. Mike's lines are usually at least a little varied from Chester's, but it's nice that you remembered he's there, lol (I sometimes don't). AI: Appears to be blending bits of Hybrid Theory, Meteora (not particularly appropriate for this song). Voices sound a bit like they've been run through a radio static filter, and Mike's rap lines, while appropriate for the style chosen, does not fit the song. I'll give him One thing (lol), he at least remembered to give Mike something different in his lines. It doesn't really work though, later on there's like a whole verse that he tries to give to Mike, but it sounds really crap because Mike doesn't tend to do that sound-bite-style rap for whole verses. This might sound odd for describing sound, but it sounds almost like the song is wading through treacle. The style calls for something a bit higher tempo, and the lines never hit that pace, kinda like gears grinding in an engine. There's also a lot of dead air, it doesn't sound like it's been sung so much as played like an instrument, which really doesn't sound right. If you could get a band to cover this version, it would be a vast improvement, but this hodge-podge emulated Linkin Park definitely does not suit this song (that said... if a person made some real effort to try to emulate the process Linkin Park went through when they were in their early Nu-metal/grunge era, instead of getting a computer to skip all that, might have a very different, possibly more fitting, result). Speaking of... he used AI to generate the video too? I don't think the video stopped at 2003 Linkin Park, pretty sure they went beyond and just chucked in anything (notably images of Chester from well into the rock era can be seen at several points), so he can't even get that right. I think this is definitely proof that you're better than the AIs shown here, since the other two songs are just stolen covers from other people with the new artist's voice veneered over the top, while the Linkin Park song just sounds (and looks) very mushy, not at all like Linkin Park (although Mike himself has proven he's not above messing with AI himself, and yeah, his experiments have also been pretty awful).
While signing up for BetterHelp, I was asked to okay having advertisements. In other words "We'll help you with your mental health, after a word from our sponsors".
Your Queen Toxicity was amazing. Maybe I'm biased considering SOAD is my favorite band, on top of also really liking Queen. But I actually really liked that one. Maybe... maybe a full song?
I think the AI does a better job at replicating the voice of the singers but they can't capture the overall feel of the artists. I think you really excelled yourself on these, especially on Toxicity - that was so good.
This video was incredible, I recently discovered your channel and I'm loving it, we definitely want more Ai videos like this. In fact, your cover of Toxicity in the style of Queen left me very surprised and happy, it was incredible!
I'd say AI wins only in mimicry of voice, it's pretty decent in how the artist themselves would sound singing those arrangements - but you actually understanding how the artists would actually have arranged it and how they would've subsequently sang it is no contest. Plus if anything this highlighted how much we overestimated AI or the people who've "used" it by basically copy pasting a different professional arrangement then running someone else's interpretation through a filter. Essentially all AI did was change the vocal cords and nothing else ingenuitive
Honestly the AI cover of Somebody That I Used is literally the lyrics of the song put over LP's Numb. Anthony's actually mixing Linkin Park and the Somebody, not just Somebody into Numb
I must agree....it´s a bit scary to hear those songs, specially the one with Freddies voice....makes me wonder what songs we would have heard if he hadn't died so early... :-(
Most AI covers, the point is literally just 'replace the voice'. They're not actually trying to get 'what if X artist *made* Y song *in* Z year', they're just trying to get 'what if X artist *sang* Y song *as* it was written in our version of reality'. The Linkin Park/Gotye one was an oddity in the 'AI cover' scene, and the Nirvana/Anka/Sinatra one was the closest most of them ever get.
I want the full version of that Toxicity cover of yours! I've been listening to the AI one for quite some time now, and you nailed it! PS. I bet all your covers will go through the AI cover of the actual singers' voices in question, so the AI vicious cycle goes on… 😅
You captured the nuances of these styles way better of course (: AI needs a lot of human work to make anything impressive but at that point, ai is just a small feature and it’d be disingenuous to solely contribute the outcome to it. As a whole, your understanding of each genre and artist and ability to recreate it for different songs entirely is way cooler lol
The moment you started to sing, put a smile on my face! Toxicity was a strong closer. I would love to see you work with postmodern jukebox. Your vocal range is breathtaking.
The second two ai ones were barely anything; pretty much just the cover they took from. I really enjoyed your covers; if anything the ai videos are a prompt/challenge for possible projects?
Anthony you definitely took the cake over ai. There is alot of things that computer generated covers need to work on. You gave it more depth and emotion. Also you know how each of these singers differ. Been watching you for a long time so I know your techniques have grown over the years. Much love brother
For the Linkin Park one, guessing they did each instrument/vocal track separately, with a lot of refinements. Shinoda bits might even a straight text to deep-voice speech thats been quantized . And probably did actual production and mixing on it afterwards. Generally though, they sound kind of obviously fake and theres little clipping bits (or occasionally actuaally skips, like Paul Anka missing words in SLTS). Compared to the actual performance. On paper one could use the AI to test/demo arrangements but its a long way from actually sounding like something other then a meme.
I think you should try to do more stuff like this in the future, AI song covers are basically everywhere. I also like the general idea of challenging yourself to do a cover of something completely by random!
Not gonna lie, that's true. I hate AI (well, not exactly, but cba with elaborating atm), but when the opposing side has as strong of a point like Chester's unique vocals and style of singing, you can't help but feel the appeal, even if reason will shout otherwise afterwards. Unfortunate, but it is what it is.
Many AI "covers" use the acapella vocal with an AI-applied "wrapper" to make the vocal sound like the desired singer... however, they still have the styling of the original performer. The ultimate results come from someone who can give the singing the inflections of the desired artists, and then applying the "wrapper" to that performance. In your case, since you can get so close to the original artist, I would expect you would be able to make flawless AI "covers." Having said that, I truly admire and appreciate you showing off your natural ability to get so close to so many artists. There's nothing like a completely natural performance!
You absolutely win ! Somebody I used to know is just the same song with a different voice and some more guitar. YOU make it sound like a really reinterpretation. The other two were ok, but not because of the ai, only because humans rearranged the songs before. That’s creativity, and only humans can do that
Most decent "AI" covers remind me of when DAWs became accessible to everyone. They used to say DAWs made songs now. Same thing now. Some people believe a single prompt plops out an entire record...
For Teen Spirit in the style of Sinatra, as soon as you started the AI my brain went "Better arrangement but the vocals sound less like Sinatra." You crushed it my digital friend. Keep up the amazing work. I'd love to hear Rise Against - Swing Life Away in the style of Pearl Jam
Queen one had to be the hardest one you to pull off vocal vise xD (I dont mean anything negative with this comment just saying you did a really good job!)
The battle was wonderful! Really exciting spectacle! Sorry, don’t know clearly how AI works with music. I understand it like AI takes micro pieces of real sound and sticks them all together. So, technically, it goes that AI songs are real singing or real singer in some point. You know, modern singers (not you of course), sing their vocal line for million times, then cut and recut, then put effects. At the end they are getting much more artificial fake product than AI work. Anthony, how I know, you are not copysinger. You are stylish artist. And have your own unique vocal. “To sing in style” for me it’s to take bright details of singer, put it through own energy and get new masterpiece. You are much more interesting than original! So, I think, in general it doesn’t make sense to compare, cause everything for their own aims. About songs: if you wanted to do copy, I’m amazed by Frank Sinatra. So warm velvet timbre - true gentleman. My compliments!
I think you won all three because yes you may not sound identical to the singers but you mimic all the little things that make them unique like the way they range and the way they pause how they breath it just doesn't sound cheap.
Dude you are awsome and your voice is splendid as well as your singing skills. On the first song I could not have any attitude, because I am 45 years old and that kind of music is way out of my interest, but on the second one, with Frank Sinatra performance, you left me speachless, great voice and than again without worthy arangement, way to go, really good. I have to admit that your version of Queen was not so good, after all you can't make a vocal impression of any singer, it's impossible. You are very talanted at all and keep doing what you do best - make music, sing and enjoy life. Be blessed!
Yours beat all three, except as you stated, the arrangement on Anka’s Smells Like Teen Spirit. I’d LOVE to hear a full version of Toxicity in the style of Queen. That was brilliant, especially the backing vocal style. You nailed that, along with some of Freddie’s mannerisms. Keep up the great work.
AI loses the depth of vocal range and the deep rich sound of breath. Also your rendition of Freddy's voice makes me miss him so very much. How proud he would have been to know he is still so loved.
Your Toxicity In The Style Of Queen sounds really amazing man, a full version of it would be great...
Beat me to it.
1000%
Absolutely agree. You nailed Freddie’s vocalizations and Queen’s style. Love it. They should hire you 😉
Yeah I found that AI cover and loved it, but it not being a real person made it weird, now it’s come around full circle
Anthony is right, besides stealing other cover artists' covers and overlaying their voices, AI did jack shit.
Shinoda is more than just a sound effect for Linkin Park. Sinatra always tries to deviate from the melody of a cover. And the backing choir is essential to Queen. But AI wouldn't understand that, only a human who gets nuances and preferences would.
1:44 Linkin Park - Somebody that I used to know
5:16 Frank Sinatra - Smells like teen spirit
7:40 Queen - Toxicity
WE WANT A FULL VERSION OF ALL OF THEM!!!!
Haven't watched yet but linkj park with somebody that I used to know is an amazing idea
The Toxicity in the style of Queen is the best I have ever heard. Would listen to that all day long. We need a full version!
One of my favorite things about covers is the other artists adding their own flourishes. To me, AI just sounds like they're trying to do the song beat by beat in a different voice.
Thus I feel the winner is clearly Anthony Vincent.
You win because you understand the nuance on each singers signature sound. Ai can't capture the "feel" of an artist
AI can't *yet* capture the feel of an artist. That's the scary part.
Can't Yet...
AI is already at that level, actually how bout yall watch xQc's video, just recently he 'made' songs with an AI tool on stream. I think you would be surprised how amazing the results were...
surprisingly enough i think the linkin park one did it pretty well. not that it matters because we all know that AI art, for instance, can look amazing but that means jackshit considering it's essentially based on plagiarism
@@Steamrick Agreed but since Ai can't "feel" it can just turn patterns into mathematical patterns there will always be something that feels close but not exact
1:44 WE WANT A FULL VERSION PLS!!!!
🤍
We NEED. I swear I listen to this song till my last breath!
I can't like this comment enough!
This is where AI shines, you can just tell it to generate a full version if anything you want, even if it's not quite as good
Anthony, you can't do a Queen version of SOAD without releasing a full version... it's unfair. We need more of this. Keep up the good work my man.
I would like to hear full versions of your interpretations, these were amazing and sounds exacly like something that the bands would do, manly the Queen part
oh yeah his versions are fantastic
Thundercats HO!
I now want the entire Toxicity album in the style of Queen as done by you. That Toxicity cover was on point.
It was on point, but not sure if Bounce would work. actually fuck it yeah he should do Bounce
Those little dead air parts on the AI really do knock me out of it. Hits that uncanny valley part where it's too fake.
Thank god haha
All AI is uncanny valley to me.
Uncanny means too real
@@HardKore5250 Non exactly
@@HardKore5250 more like its trying to be real but something is off
This is no contest, you win!! I love this concept so much. Beautifully done
That Queen style toxicity is some of your best vocal work, I want a full version of that!
You not only win but I loved the fact you identified the songs used as the source was not the originals. Nicely done.
Hands down, your covers are a lot better! We need full versions of the songs or at least a part 2!
Honestly, the only time AI impressed me musically is when you ran your version of Black Parade in the style of Queen through that Freddie Mercury filter - and that's only because you already did all the heavy lifting by capturing Freddie's style, his cadence, and all the little variations that made the performance so good to begin with. The voice filter filled in that last final gap and made it truly eerie, but it could never have done that without the actual work you'd put into the performance first!
I'm glad Boi What puts the work in his songs
AI will be able to do it all and can. We just have to acknowledge our strengths and ai strengths.
@@HardKore5250I think you’re entirely missing the point
@@HardKore5250 it sounds like you're doing a great job of not acknowledging the strengths of AI & the strengths of humans here.
You absolutely won each time. Although I feel like a person would win in most human vs. AI scenarios, you especially are great and your voice is just awesome.
I would love to hear a full version of the Toxicity cover.
One thing I noticed on these AI covers is that they can`t change the metrics of the singing line to fit the style properly. It is just the voice that is changed to match another artist and then the rest of the song comes around that. You really put time thinking how a different artist would aproach a song. That is why you win every battle in my opinion ( that and the fact that you do everything with your own voice).
No contest! All of yours are superior because they have genuine depth and emotion. Not close in my view!
You win on all of them hands down. There’s a warmth to your tone, plus your creativity with the music. AI is doing a “cover”, whereas you are doing the songs “in the style of”, which I prefer. That being said, we need a whole album of various metal songs in the style of Frank Sinatra. Outstanding! Never knew I would want to hear something like that, but now that I have, I want more!! ❤❤❤
Seems like you went for old Lincoln Park and the AI went for new Lincoln Park. I say you still win all around since you're doing that with your own voice and muscles and it's infinitely more impressive as your ability to mimic other singers voices is literally amazing and never ceases to impress
Lincolm Park 😂
me at this comment for last 30 minutes trying to understand if Lincoln Park is a multiple occured typo or a some joke with a inner meaning that I cant understand due to my language level lol
@@adoculos1046 My guess is autocorrect.
@@adoculos1046 🤣 No it's a multiple typo. That's what I get for using my phones and it's voice to text function to leave a comment.
Agree, He did the actual work and it sounded really great.
The Linkin Park cover is really the first *true* AI cover I've seen, where both the instruments and vocals are fully done by an AI (with a lot of human tinkering, of course), and I do have to say, I think the AI had you beat in *that* one, for both accuracy and sound. I don't know if it's a fair comparison to make, since I, like you said, can't imagine that was done by just plugging it into a machine learning algorithm, pumping out something and sticking it on the internet; I'd imagine it was done piece by piece, perhaps even each segment was made by a person, then put through an AI to make it sound more Linkin Park'y, *then* stuck back together. That's how I'd have to imagine it was done for it to sound so good and have such unique touches.
The other ones though... don't stand a chance, your Queen cover of Toxicity needs to be a full video!
Since everyone wants to be a 'fan' and say you, I'll play Devil's Advocate. You are talented in ways no one in the world is. But of course an AI song, done in the actual voice of the original artist is going to sound more like the artist than another person trying to replicate that voice. Plus the AI song had more time put into it, while you were just trying basically in real time to do the same thing. That being said, you're the reason I started loving cover songs in the first place on TH-cam. Never ever stop doing what you're doing.
That Paul Anka album is fantastic, by the way. I remember playing it in my car the year it came out and my mom saying something like "I didn't realize some of these songs were so old"
I loved your versions. Specially Sinatra's rendition.
But the power of AI is scary. The Freddie Mercury impersonation stunned me.
Really? I thought AI Freddie was the least alike
Not even worth a discussion. You win.
Exactly
Wonder what we'll be able to do in a few years
Congratulations!!!! YOU WIN!!!!!!
You sound excellent in the style of Sinatra! 😊
you voice is so incredible ! thanks alot for these songs... best regards from austria
i think the main way people do AI singing is they record themself doing an impression, and then the ai tweaks it to be more like the actual singer. could be fun to see how good AI versions of your videos turn out since your impressions are already so close
It's exactly how they do it
I knew that music of AI Queen sounded familiar. Yours DEFINITELY won! Please do a full version of this one!!
I think the only one the ai kind of won at was the Linkin Park, because it doesn't feel like them to me without the rap part also being included. Every other one yours was so much better. Your way of singing in so many styles just blows me away.
You'll always be our best Ai, you definitely win 🏆
Awesome video! This really showcased your talents and musicians in general.
Your versions are so emotive and musical, while the AI version are just cheap copies!
You win easily😊
Dang nirvana mixed with frank sinatras style is actually pretty good.
that's because it's an overall great song. No matter the arrangements good song writing always comes through
@@saysthemanbehindthemask true but that's not always the case.
@@saysthemanbehindthemaskI think the songs have to fit like a glove or a person has to redo the arrangement to fit other genres. It is not just "good song is auto good in any genre"
You win every time because you’re a real human and it shows.
I prefer the AI version of Toxicity, but moreso I think that the AI cover creator chose to go along a later Queen approach of slower and softer ballads like "Don't Try So Hard", "Don't Try So Hard", and "I'm Going Slightly Mad", hence the Robin sample.
Not sure your Night At the Opera Queen style really fits the somber tone of Toxicity. But I know that's also the sound most people associate with Queen, so solid work trying to make it fit.
For the first two songs, I think you won out over the AI. I enjoyed your versions more, but I didn't hate the AI version either. It was funny to find the Frank Sinatra one was just a slightly tuned Paul Anka version. I don't have that critical of an ear for music, so none of the AI versions bothered me. I do think that the AI cover of the third song was better, but only becuase they just changed the female singers voice slightly to make it sound like Freddie. It was hard to reconcile your version when you were doing a more growly scream... don't know that I have heard much like from him before.
Great video, and I would like to echo the other people and say I want a full version of the Linkin Park and Frank Sinatra covers, lol.
I do think that the Ai got you on the Linkin Park song, but only that one. I took into consideration that the volumes were completely different which impacted the feel of the Ai version. I would also imagine that a lot of human intervention took place, but Ai is definitely going to eventually be better. For Gods sake just look at Sora or whatever that video thing is called. It’s insane!! I don’t say these things with glee as I once was a musician in a past life and I poured my heart and soul into my songs, but the Ai is just coming man, and it’s coming fast! I really enjoy your stuff though. Just found you.
Hmm... I have comments that I can give to both for the Linkin Park cover.
Yours: Clean sound. Joe's signature DJ input is missing, but it suits the song better. The slower paced, rockier style better suits over the early Nu-metal/grunge style used by the AI, a great choice. Chester's voice tends to be a bit higher when he's singing head voice, but for the most part, spot on. Mike's lines are usually at least a little varied from Chester's, but it's nice that you remembered he's there, lol (I sometimes don't).
AI: Appears to be blending bits of Hybrid Theory, Meteora (not particularly appropriate for this song). Voices sound a bit like they've been run through a radio static filter, and Mike's rap lines, while appropriate for the style chosen, does not fit the song. I'll give him One thing (lol), he at least remembered to give Mike something different in his lines. It doesn't really work though, later on there's like a whole verse that he tries to give to Mike, but it sounds really crap because Mike doesn't tend to do that sound-bite-style rap for whole verses. This might sound odd for describing sound, but it sounds almost like the song is wading through treacle. The style calls for something a bit higher tempo, and the lines never hit that pace, kinda like gears grinding in an engine. There's also a lot of dead air, it doesn't sound like it's been sung so much as played like an instrument, which really doesn't sound right. If you could get a band to cover this version, it would be a vast improvement, but this hodge-podge emulated Linkin Park definitely does not suit this song (that said... if a person made some real effort to try to emulate the process Linkin Park went through when they were in their early Nu-metal/grunge era, instead of getting a computer to skip all that, might have a very different, possibly more fitting, result). Speaking of... he used AI to generate the video too? I don't think the video stopped at 2003 Linkin Park, pretty sure they went beyond and just chucked in anything (notably images of Chester from well into the rock era can be seen at several points), so he can't even get that right.
I think this is definitely proof that you're better than the AIs shown here, since the other two songs are just stolen covers from other people with the new artist's voice veneered over the top, while the Linkin Park song just sounds (and looks) very mushy, not at all like Linkin Park (although Mike himself has proven he's not above messing with AI himself, and yeah, his experiments have also been pretty awful).
While signing up for BetterHelp, I was asked to okay having advertisements. In other words "We'll help you with your mental health, after a word from our sponsors".
Lol that's not good
Toxicity in the style of Queen would make for such a good full version, I'd totally love to hear that! What an odd combo that works surprisingly well!
Your Queen Toxicity was amazing.
Maybe I'm biased considering SOAD is my favorite band, on top of also really liking Queen.
But I actually really liked that one.
Maybe... maybe a full song?
I think the AI does a better job at replicating the voice of the singers but they can't capture the overall feel of the artists. I think you really excelled yourself on these, especially on Toxicity - that was so good.
This video was incredible, I recently discovered your channel and I'm loving it, we definitely want more Ai videos like this. In fact, your cover of Toxicity in the style of Queen left me very surprised and happy, it was incredible!
I'd say AI wins only in mimicry of voice, it's pretty decent in how the artist themselves would sound singing those arrangements - but you actually understanding how the artists would actually have arranged it and how they would've subsequently sang it is no contest.
Plus if anything this highlighted how much we overestimated AI or the people who've "used" it by basically copy pasting a different professional arrangement then running someone else's interpretation through a filter. Essentially all AI did was change the vocal cords and nothing else ingenuitive
Honestly the AI cover of Somebody That I Used is literally the lyrics of the song put over LP's Numb. Anthony's actually mixing Linkin Park and the Somebody, not just Somebody into Numb
HANDS DOWN, you win every single one, Anthony.
I must agree....it´s a bit scary to hear those songs, specially the one with Freddies voice....makes me wonder what songs we would have heard if he hadn't died so early... :-(
Most AI covers, the point is literally just 'replace the voice'. They're not actually trying to get 'what if X artist *made* Y song *in* Z year', they're just trying to get 'what if X artist *sang* Y song *as* it was written in our version of reality'. The Linkin Park/Gotye one was an oddity in the 'AI cover' scene, and the Nirvana/Anka/Sinatra one was the closest most of them ever get.
Ty
I want the full version of that Toxicity cover of yours! I've been listening to the AI one for quite some time now, and you nailed it!
PS. I bet all your covers will go through the AI cover of the actual singers' voices in question, so the AI vicious cycle goes on… 😅
You captured the nuances of these styles way better of course (: AI needs a lot of human work to make anything impressive but at that point, ai is just a small feature and it’d be disingenuous to solely contribute the outcome to it. As a whole, your understanding of each genre and artist and ability to recreate it for different songs entirely is way cooler lol
The moment you started to sing, put a smile on my face! Toxicity was a strong closer. I would love to see you work with postmodern jukebox. Your vocal range is breathtaking.
The second two ai ones were barely anything; pretty much just the cover they took from.
I really enjoyed your covers; if anything the ai videos are a prompt/challenge for possible projects?
Yeah man it’s been bugging me since I saw that linkin park ai cover, I think you captured them a little more truly.
Anthony you definitely took the cake over ai. There is alot of things that computer generated covers need to work on. You gave it more depth and emotion. Also you know how each of these singers differ. Been watching you for a long time so I know your techniques have grown over the years. Much love brother
For the Linkin Park one, guessing they did each instrument/vocal track separately, with a lot of refinements. Shinoda bits might even a straight text to deep-voice speech thats been quantized .
And probably did actual production and mixing on it afterwards.
Generally though, they sound kind of obviously fake and theres little clipping bits (or occasionally actuaally skips, like Paul Anka missing words in SLTS). Compared to the actual performance. On paper one could use the AI to test/demo arrangements but its a long way from actually sounding like something other then a meme.
Are we not gonna talk about the banger of a sponsor ad that was? LITERALLY
I LOVED the Toxicity cover you did...my mouth was agape!
We absolutly need a full version of 'Toxicity' in the style of Queen! That was one hell of a mix and song!
Human voice will always win over AI. You have heart and soul and nuance, AI has none of those.
Not always, but I agree with "most of the time".
You've done better in all of them and the fact that you didn't have to steal anyone else's work to do it is just another positive for you.
I think you should try to do more stuff like this in the future, AI song covers are basically everywhere. I also like the general idea of challenging yourself to do a cover of something completely by random!
I was literally asking myself this exact question. Hands down you nailed it this time. I wonder where we will be 1 year from now.
When the AI Chester hits that first scream. Your smile says everything 😊
Not gonna lie, that's true. I hate AI (well, not exactly, but cba with elaborating atm), but when the opposing side has as strong of a point like Chester's unique vocals and style of singing, you can't help but feel the appeal, even if reason will shout otherwise afterwards.
Unfortunate, but it is what it is.
Many AI "covers" use the acapella vocal with an AI-applied "wrapper" to make the vocal sound like the desired singer... however, they still have the styling of the original performer.
The ultimate results come from someone who can give the singing the inflections of the desired artists, and then applying the "wrapper" to that performance. In your case, since you can get so close to the original artist, I would expect you would be able to make flawless AI "covers."
Having said that, I truly admire and appreciate you showing off your natural ability to get so close to so many artists. There's nothing like a completely natural performance!
You absolutely win ! Somebody I used to know is just the same song with a different voice and some more guitar. YOU make it sound like a really reinterpretation. The other two were ok, but not because of the ai, only because humans rearranged the songs before. That’s creativity, and only humans can do that
Yeah the ai covers were just unsettling. Yours were fitting tributes, especially to Chester and Freddy
I like the "AI" better overall, but as you established: its not actually AI. Its someone elses stuff just slightly modified by the AI.
Most decent "AI" covers remind me of when DAWs became accessible to everyone. They used to say DAWs made songs now. Same thing now. Some people believe a single prompt plops out an entire record...
Man, you are a real talent. Always top content. AI will allow us to do crazy stuff but you are the winner here
I think everyone is wanting Toxicity in the style of Queen, Please bring a full versionnnnnnnn :)
My heart is heavy hearing AI Chester. He had the most amazing voice. He’s still missed so much.
For Teen Spirit in the style of Sinatra, as soon as you started the AI my brain went "Better arrangement but the vocals sound less like Sinatra." You crushed it my digital friend. Keep up the amazing work.
I'd love to hear Rise Against - Swing Life Away in the style of Pearl Jam
Queen one had to be the hardest one you to pull off vocal vise xD (I dont mean anything negative with this comment just saying you did a really good job!)
The battle was wonderful! Really exciting spectacle! Sorry, don’t know clearly how AI works with music. I understand it like AI takes micro pieces of real sound and sticks them all together. So, technically, it goes that AI songs are real singing or real singer in some point. You know, modern singers (not you of course), sing their vocal line for million times, then cut and recut, then put effects. At the end they are getting much more artificial fake product than AI work.
Anthony, how I know, you are not copysinger. You are stylish artist. And have your own unique vocal. “To sing in style” for me it’s to take bright details of singer, put it through own energy and get new masterpiece. You are much more interesting than original!
So, I think, in general it doesn’t make sense to compare, cause everything for their own aims.
About songs: if you wanted to do copy, I’m amazed by Frank Sinatra. So warm velvet timbre - true gentleman. My compliments!
You will always win against AI. You have vocal cords not some digital representation of them. Love you and your immense talent!
Probably your best work so far. Absolutely brilliant!
No danger of being replaced any time soon. For the love of everything that is holy, full versions of all of these please!
You my man, you win. As always awesome stuff and an amazingly talented artist.
Sir you definitely won! Your voice with deep emotion is beautiful!
Damn, you do a really good Mercury impersonation!
*edit* The A.I. hasn't quite left the 'Uncanny Valley' yet
We NEED a full version of that Toxicity Queen style cover holy crap was that the craziest thing ive heard!
Most beautiful sinatra ive ever heard
Ahhhhh!!! Your Toxicity in the style of Queen was SO GOOOD
I think you won all three because yes you may not sound identical to the singers but you mimic all the little things that make them unique like the way they range and the way they pause how they breath it just doesn't sound cheap.
For the first one, I prefer the ai arrangement more, but the performance and sound of yours
Dude you are awsome and your voice is splendid as well as your singing skills. On the first song I could not have any attitude, because I am 45 years old and that kind of music is way out of my interest, but on the second one, with Frank Sinatra performance, you left me speachless, great voice and than again without worthy arangement, way to go, really good. I have to admit that your version of Queen was not so good, after all you can't make a vocal impression of any singer, it's impossible. You are very talanted at all and keep doing what you do best - make music, sing and enjoy life. Be blessed!
You gotta do more of these. They are so fun. Big fan of both of the queen ft. Soad versions. Yours and ai.
I hope you will do a full version.
thats one of the coolest ideas to do with AI. You are so creative with your videos dude
can we please actually get a full version of your cover of toxicity in the style of queen. that little snippet was ace.
Anthony wins every time and its not even close. I fuckin hate AI.
Somebody that I used to know... we need a full version of it!! Looove your style!
Yours beat all three, except as you stated, the arrangement on Anka’s Smells Like Teen Spirit. I’d LOVE to hear a full version of Toxicity in the style of Queen. That was brilliant, especially the backing vocal style. You nailed that, along with some of Freddie’s mannerisms. Keep up the great work.
AI loses the depth of vocal range and the deep rich sound of breath. Also your rendition of Freddy's voice makes me miss him so very much. How proud he would have been to know he is still so loved.
More vids of this, please!! Great job!
I really like these challenges and hope you do more!
One of your better series of recently. Hope this can continue
Also I want a full version of Queen toxicity
Let me echo other comments by saying I need that full Queen cover of Toxicity!
You are so incredibly talented 😍 also youyouyouyouyou
I love this concept!