The Smiley Smile Message Board

Hey all,

I came across some new technology and decided to have a little fun with it!

Here's one to test the waters:

https://www.youtube.com/watch?v=13ICunDuIcg&ab_channel=DaeLims

I've been debating posting, but it was too overwhelming to think I'm the only who has heard any of this stuff.  ;D

The technology is here. Should we have some fun before all the sh*t takes start popping up?   :lol

PS - Lots more where this came from, if people are interested!

Keep 'em coming. While still in its infancy, I could see this kind of technology used to make unfinished parts of Smile a reality.  Imagine hearing a dream cover by Dennis or Carl. Certainly Nothing will stop fans of any band from experimenting. In a few years that dream Beach Boys, Beatles, or Band reunion album could only be a few keystrokes away from happening.

Already started playing with it. I'd say SMiLE AI (follow up to SMiLE AD?) could definitely happen.

In the meantime, I've got some stuff that I'm positive will blow people's minds. Brian circa early 70's singing a powerful vocal on "My Diane" is HUGE, but it's only scratching the surface...!

.

What technology did you use to make this?  It honestly sounds more like Carl than Brian but still not bad.

It's a script that analyzes samples of (in this case) Brian's voice and forms a sophisticated voice model that can be applied to previously recorded vocals.

So basically, I'm singing the stuff - and after running it through the script, it spits out a Brian vocal. His consonants, vowels, breath sounds etc. I have a lot of years of doing Beach Boys covers (read: attempting to imitate Brian), so I can approximate the emotion he'd put into the vocal - and the AI takes care of the rest. There's a lot of trial and error to get the AI to, for instance get it to sing "softer" or go into falsetto. I've actually made two models already. The one you hear on "All I Wanna Do" is 2.0, which is smoother. "My Diane" is mostly 1.0, which had a bit more of a "shouty" tone and had more noticeable artifacts.

I can assure you that there's no Carl in there - it's 100% based off of Brian's voice. Of course, it's AI so it's not gonna be perfect - but I think it's damn close. I've been living a dream over here. I realize this AI stuff could be a lot to take in (it has been for me), but I hope other people get a kick out of the results of my play time. :3d

Wild.

BWPS Heroes & Villains with AI model Brian vocal?
Would it sound more like 1967 released H&V or 1966 cantina vocal?

Although I don't feel quite right saying it, whole BWPS album with AI "vintage" vocals is intriguing.

"Vintage" Brian on Summer's Gone?

Dennis on A Day in the Life of a Tree? Or Brian...

My first thought on hearing the recent Kanye examples was what an AI Dennis singing Holy Man could sound like. There are so many ethical issues with AI and art/music (and AI in general), but the technological developments are extraordinary and the possibilities are becoming both incredible and quite frightening.

Does a really solid AI voice model require lots of isolated vocal tracks?

VERY interested to hear more. It’s a strange experience listening, knowing it’s AI and still hearing some very clear Brian in tone etc. Very impressive work!

.

It’s important with this stuff to try to remain level-headed, and also polite and not dismissive. I think it’s important to note the obvious, which is that folks who are most enthusiastic about AI applications right now also not coincidentally seem to be the most effusive in claiming how close to seamless it all is.

It’s all undeniably intriguing, and I can’t claim it doesn’t strike an eerie and intriguing (and creepy, and….) chord. As I was just telling someone, I’m not sure how great I feel about the sort of “this may be unleashing something unholy on the world, but someone else was gonna do it anyway, sooooo….”, but the stuff is what it is and it’s happening.

I could tell immediately that someone was singing to guide this AI, as the phrasing and timbre and all of that was both not matching the original respective vocals, nor was it “nailing” Brian in any particular era either. I’m not saying it’s a million miles off. But we’re firmly in uncanny valley territory, and I suspect as long as a human being is used to guide the AI vocal, that’s going to introduce some noticeable anomalies to those who are super familiar with Brian’s (or whomever’s voice).

There’s also the anachronism of having a 1970-style Brian singing a 1976 song. In one way this is easier to not be as unsettled by, because it’s an obvious anachronism, like hearing a 90s Carl voice laid on top of the 1973 “Sail on Sailor” track.

Even just within the realm of goofing off in BB land, there are very different types of applications of course. There could be cases of trying to replicate something that *should have* existed (e.g. 1966 Brian and the other guys singing the missing vocals on the ‘66/’67 Smile tracks), or something that once actually *did* exist (e.g. a Carl or Dennis contemporary lead on “Sail on Sailor”), or something that at least *could have* existed (Brian singing, well, a ton of stuff). On the other extreme, I guess one could try to do a “What if 2023 Mike Love sang ‘Surf’s Up’?”, or “What is Dennis Wilson sang the lead on ‘Looking Back with Love’?”

I’ll stop giving people ideas….

.

This stuff definitely sounds "close enough" to Brian Wilson that when I hear it, yeah I do start to think, huh, what if we had the 1966 Beach Boys' voices singing the 2004 lyrics to "Blue Hawaii" or "On a Holiday"?  Could a complete, seamless Smile LP be created with AI?   And the answer to that question would appear to be "yes."  But should it be done?  I must admit that I'm a little uncomfortable with the idea, and we get into philosophical questions about whether something "artificial" can be an acceptable substitute for something "real"... and the extent to which the lines between "real"l and "artificial" are blurring. We're  in brave new world

I definitely get that. I think it's important that anyone who does these labels them as "AI" so it's documented. Who knows, it could become illegal one day. I'm surprised deepfakes are still legal.

Yeah, I mean, no one can say who would have been tricked had the stuff been leaked and presented as legit. My best guess is that those who have the knowledge and familiarity with the music/voices and also some overall analytical skills would have easily been able to tell some sort of fabrication was taking place, and that something was clearly off. I don’t know who would have been able to specifically say “that is clearly a present-day person singing, guided by AI software”, but there are so many circumstantial things about these recordings to poke holes in, I think only casual listeners who aren’t discerning would be incredulous about it.

On the tech side of things to the degree it’s audible, I hear auto-tune type digital jumps in the notes. I hear points where the notes are just sung/bent in a way the vocalist never would have (I don’t know if the like span of time one has been listening to the actual voices matters or not, if someone who is a huge fan but became a fan like 5 or 10 years ago can’t always hear stuff someone who has been listening for like 40/50 years, etc. can, all I can say is that it doesn’t take much for it to quickly *not* sound right). There are other auditory things not specifically to do with the AI tech that indicate something is amiss. The murky sound of the vocal and mix overall. The recording sounding like a pre-existing mix of something has had something grafted on to it (e.g. if a rough acetate mix of Brian doing a Mike vocal all of a sudden popped up, there would usually be other differences/anomalies in the mix to differentiate it beyond the vocal).

I’m not speaking to the intention of any given AI thing being created. I don’t know when or if someone would intend to fool people or not. But when we’re talking in the realm of “It sounds just like Brian, doesn’t it?”, then I think a lot of knowledgeable, detail-oriented fans are going to notice *everything* about a given recording/mix/presentation, and there’s no way it really can sound “just like” Brian if it’s 1970 Brian singing over a backing track that we know was recorded in 1976. I suppose then the task becomes, I dunno, re-recording a 1970-sounding backing track? That’s obviously going to not pass muster as well.

But if we attempt to pull everything out of the equation other than that actual voice, I think then we get into territory where it can quickly become difficult to have a constructive conversation, because I’ve seen this happen with other stuff where some people think it’s “just like!” whatever it is in question, while others genuinely immediately can tell it’s not that thing. It’s like colorized photos (which are often also done with AI), or colorized film/video. Some people look at some of those examples and think it’s uncanny. I’ve seen examples that have been so processed that they look like an oil painting of wax figures, but some people think it looks just like Technicolor film.

I suspect with these AI voices, there is a surprisingly narrow gap/line between “wow!” and “ugh, not even close”, and that’s just looking at folks who are presumably making good faith arguments/observations.

Worth also keeping in mind that the possible quality of this sort of thing will likely develop rapidly. You might imagine a scenario in years to come where AI voices are licensed out for all sorts of purposes, with the theoretical possibility of having AI Elvis “sing” whatever you choose. All sorts of problematic issues here, but I also can’t help but imagine hearing lost vocals, unfinished work completed etc.

Personally, I’ve always struggled to listen properly to “alternate” track listings, lost albums etc, possibly because they’re just not what the artist intended. That’s not me being a purist, I’ve tried many times to construct those sorts of tack lists, but I always come back to the original albums. I can enjoy stereo remixes and not feel the same, including those with digital extraction used. So I don’t know how I’ll feel about a partially AI Smile, for example. Nonetheless, I very much hope to hear more of this and would dearly love to hear an AI construct of some of those things I’ve only ever heard in my head (Sail on Sailor with Carl’s vocal, Holy Man with Dennis etc). We’ll never know how Carl might have phrased his take on Sail on Sailor, but maybe an AI version would be worth it just to hear an approximation.

I can see a reasonable use case for finishing things that weren’t completed, but I think there are debates to be had about widespread use in commercial releases using AI vocals.

Agree it is "only" intriguing and eerie. But wild nonetheless. And the possibility of rapid development certainly adds to the intrigue.
Is anyone suggesting it sounds "just like" Brian?

More Today Than Yesterday... with Brian on lead:
https://www.youtube.com/watch?v=Xj9c1UuuXxM&ab_channel=DaeLims

Next they need to use AI to write imitation Beach Boys songs.

[quote author=seltaeb1012002 link=topic=28316.msg682044#msg682044 date=1680110716
Ooh I got one! How about a super rare acetate of the Til I Die demo with a long lost vocal track?


https://www.youtube.com/watch?v=GR8hz3uzljw&ab_channel=DaeLims

[/quote]

It can't be a Brian Wilson demo without him going "andthenitgoes" every ten seconds! While I think some of the phrasing is a bit too exaggerated (i.e. the "hey hey hey" lines are a bit oversung IMO), this is extremely impressive, I hope this program can be released somehow!

While that precise verbiage hasn't been used, I think "it's AI so it's not gonna be perfect - but I think it's damn close" is something very near such a claim. "Damn close to just like Brian" as it were.

I think what I was generally trying to allude to in what I've seen from a lot of folks who use AI applications and are most bullish about it is that there often isn't much moderation in how they sort of self-assess the work (or assess the work of others using AI). The sensibility isn't so much artistic as it is trying "prove" the tech.

I don't see a lot of "this was just for a lark; I'm not sure how good it is, but it's just an experiment", and I see a lot more "this is amazing stuff and we're on the precipice of the world changing, and while I maybe can't claim it's 100% indistinguishable from the real thing, I think it's like 99% there", if not folks (and I'm referring in this case more to loud, AI people who are often also like crypto dudebros on social media, who think their AI art built off of Google image hits of other artists is perfect) who say they are just as much an "artist" as someone who actually makes the stuff for real.

This is unbelievable....like in a weird, good, bad, amazing way :lol

Seriously it wasn't perfect but moments in there sounded like straight up Brian. And to heck with perfection, that this is even remotely possible is mind-blowing. I can't even imagine how this technology will transform even in just the next year.

I'm rather concerned about this technology for many reasons, but this is a case where the creative in me is excited for the possibilities.

We have MIDI orchestras, pianos, etc that can help us replicate an orchestra/band. And now we can add "Brian Wilson", "Carl Wilson", etc to the list of instruments people can use to create their own little slices of beautiful music.

I'm sure a lot of bad sounding stuff will come from this, but I can imagine that we'll all be pleasantly surprised (as we already are) at some gems that come from this new era of technology.

Great work, seltaeb!

Probably not technically a deep fake (as it's clearly satire), but the one where George Lucas critiques the Rise of Skywalker (https://www.youtube.com/watch?v=_MuxVqB3I7E) trailer is one of the greatest things to come from this era/tech.

The other day Elon Musk and a bunch of other techies said that a 6-month halt on AI technology was needed - now I'm wondering if Elon's less concerned with ChatGPT and more concerned with the possibility of hearing Mike Love singing lead on 'God Only Knows' ;D

Holy sh*t, out of body experience! https://m.youtube.com/watch?v=AAGJUZWh45I

This is all so freaking fantastic. As Brian seems to be pretty much done recording, these sort of projects will fulfill a real need. Keep it coming, please!

I wonder if it would be possible to, for example, take multitracks of vocals from a Beatles song, including backing vocals, and replace every single track with an AI Beach Boy equivalent (using the recorded tracks as guide vocals, rather than re-recording one's own). Here, There, and Everywhere?

(Well, heck, this would be the best way to replace vocal tracks on BWPS.)

And what do you know, the legendary BBS historian is talking sh*t…

I don't think the criticisms being lobbed at this are entirely accurate or fair. To quote Hey Jude from earlier, "It’s important with this stuff to try to remain level-headed, and also polite and not dismissive."

To say it sounds "nothing" like Brian is obviously excessive criticism. I liken the criticism to someone mocking a CGI explosion from a 1990s Hollywood movie "that looks nothing like an explosion" - when in reality the computer helped digitally generate an image of fire, smoke, and a shockwave. Is it a real explosion? No. Does it look perfect? No. But to claim it looks "nothing" like an explosion is to melodramatically ignore that a digital artist took real-life visual elements/cues and created a digital representation that is very similar to a real life explosion.

Ethics aside, I think this is absolutely stunning. Reminds me of the story of the creators of Pixar making a very rudimentary short version of Toy Story - it looked utterly creepy and disjointed, but the visionaries (including Steve Jobs) saw the potential and invested heavily...they didn't say "those look nothing like toys, this fails miserably" - they saw the potential and ended up making a multibillion dollar company that changed digital animation forever.

I'm not saying that seltaeb is going to become a billionaire using someone's AI software to recreate Brian's voice, but I am saying that to quickly/rudely dismiss/criticize this work is fairly pretentious and ignores the possibilities to come. I can see why someone wouldn't like what they're hearing with these tracks (for a couple reasons), but it's also another thing to be excessively dismissive.

Again, I have to reiterate that, based on what I've seen from a broad number of reactions to AI stuff (pics, video, art/"paintings", sound), people need to accept that, setting aside examples of people who have ulterior motives in praising or panning this stuff, even among folks making good faith attempts to take this stuff in, there is a BROAD range of opinions.

We're seeing this right now. Some folks are hearing this stuff and are stunned with how amazing it sounds. Others think it sounds awful. Truly. I've played these recent recordings for a few people, asked for a few reactions. They've mostly been *not* good, and these are people who know their s*it and are not just old cantankerous fans who don't like change/technology.

As with many polarized things these days, this doesn't bode well for the future of AI simply from the point of view of reacting to it and gauging its quality. And I've definitely seen with visual art a bit of the Dunning-Kruger effect in play, where non-artists make AI art, and to them it looks *just as amazing* as something a real artist would make. Just like some people hear autotune and think that doesn't sound completely unnatural as far as what a human voice should sound like.

Do these AI recordings sound like Brian Wilson? It's a bit like being "pretty pregnant"; you either are or you aren't. A lot of new tech these days is being sold on "just you wait, it's almost there", and in some cases with AI I think that's almost certainly true. I think ironically that AI-ing a singing voice is going to lag surprisingly far behind any visual medium (especially photo fakes) as far as what can actually "fool" people, or what sounds so close or just like something to the point where one can listen to it and, like, enjoy it and think (or pretend?) it's the real thing.

To me, the AI vocals sound much more convincing than AI photos.

Imagine the possibilities. Using Brian's current voice to do a lead vocal for What a Fool Believes, and then slipping it into his jukebox...

Actually, I'd like to hear that.

Yeah the possibilities seem endless. And like with anything, I'm sure there will be very bad implementations of it, but also very good implementations of it. It's kinda like some SMiLE mixes aren't great, unoriginal, but others really capture the magic...some Smile mixes may not be true to the history of the project, but they sound amazing. And like Smile mixes, there are so many variables that can make them good or bad.

Likewise I think there are so many variables with this new technology. And I also think that it opens up an entire new universe to explore and tinker with. These mixes are cracking the door open. Just imagine what we'll be hearing, good or bad, in the years to come.

It's interesting that just about every actual artist/creator really finds AI stuff to be pretty awful. It's easy to dismiss these people as all just s***ing their pants because they're going to be replaced.

But these are also people that know these mediums better than those who only consume it. I think that's worth something too. As are the ears of people who have been listening to these voices for 30/40/50/60 years.

Awhile back someone did a piece of AI art that was something like "Calvin and Hobbes if they were a Laika film", and some people looked at that and thought "holy s**t, that's amazing!", and then a bunch of people pointed out how *not* like that it looked. Actual artists (and just Calvin and Hobbes and Laika fans) were able to point out the 37 things that were totally off.

I played those Brian AI tracks to someone who told me "Wow, that *does* actually kinda sound like Curt Becher!"

I for one find the AI separation technology to be far more fascinating, with far more room to credibly potentially "fix" things or "right" things in a convincing way that can truly be enjoyed as a listening experience. Random example: Fix those two Beatles reunion tracks (which I actually love even as-is) so Lennon doesn't sound like he's singing from inside a toilet (Free As a Bird) or on helium (Real Love).

I work in a field that is very much targeted by AI right now (graphics//video). But I heard a great commentary on it recently - that AI, like any other computerized tool in recent years, is a tool...it doesn't have to be the end of your job as long as you utilize the tool. It's much more complicated than that, but that's the simple version of it. I look at the graphic/art AI technology the same way I'm looking at this audio AI tech: it may seem rudimentary now, but wait for the tech to evolve.

Such technology could very well mean the end of my job in the future, but oddly enough doesn't make it any less fascinating for me.

I played the AI tracks for my wife and she was also blown away that it was AI...yeah it doesn't sound perfect and there are clear flubs/imperfections, but to think that this is the first iteration of the technology that has Brian sing on songs he never sang on. It's quite a marvel.

These are amazing. Especially Guess I'm Dumb. I would love to hear A Day in the Life of a Tree. Subscribed!

The implications of something like this are horrifying.

WOW that 'Guess I'm Dumb' is great. 'Be My Baby' is off the charts great!

I can see why this would worry people - but just for the "what if" this is incredible. This stuff isn't going to replace classic Beach Boys, but for those of us who wish we could Brian take the lead on certain tracks, what a treat.

I just want to say, the recent "Be My Baby" one really sounds convincingly like a classic 1964 era Brian vocal.  It even sounds double tracked the way a BB vocal from the period would have been.  It probably helps that there is a wealth of BB acapella material, both officially and unofficially released, to train a model on.

Obviously the wider implications of this tech in a wider sense are distressing for artists who are still actively recording, but in this capacity you are basically using AI tech to reproduce something that is lost (Brian's classic voice) and realizing something that never existed.  This could be used to recreate songs that Brian never recorded vocals to ("Sherry She Needs Me" should be at the top of your list) and the covers that you've posted show this could be used on songs outside of the Beach Boys canon.  It would also be interesting to hear BB songs sung by Brian with different members singing lead, Dennis singing lead on "I Get Around" etc.

When I first saw the topic title I was excited that AL and Brian were making an album 😆

But this AI stuff is fascinating and great fun!

It's kind of fascinating, and I'm truly not singling out anybody in particular, because I honestly don't keep track of how each person has come down on any given topic, but it's pretty fascinating to see so many fans up in arms because like the remix of a track has different sounding drums, while some fans are fine with a dude and computer algorithm trying to sound like Brian Wilson.

I brought up the idea of this tech bleeding out into more joke ideas (e.g. let's hear Dennis sing "Looking Back With Love"!), but I honestly think that might be the *best* use of the tech, where there's no pretense to making it, and it's just truly for a laugh.

To the degree this tech would ever be used to "fix" or "right" something, I'd say re-singing vocals the actual BBs actually sang seems pretty low on the list. Actual missing stuff (which means we're mostly talking about "Smile" and a few odds and ends like "Sail on Sailor") strikes me as the best of a list of not-so-great applications for this. But, like, when 1976 Brian sang a lead, *that* voice is part of the whole thing. That's part of what that recording MEANS.

And, as a total aside, I find subbing in younger Brian on "Matchpoint of Our Love" particularly bizarre, as he actually sounds good on that one on the actual original recording! It's kind of infamous for that.

Fascinated to see this! Amazing work so far. Training the models can be quiiiiite time consuming!

I *may* have been working on a new Smile mix using all this fun new Ai stuff... keep your eyes and ears peeled...
There's a lot more to this AI stuff than generating new vocals, even the ability to separate specific voices and instruments can mean amazing things for good, clean new mixes of tracks we have no masters for. (Barnyard? Child is Father of the Man??? Wonderful?)

As I've mentioned, using AI to isolate pre-existing parts is very interesting/useful and sometimes revelatory, and a *very* different application of AI tech than generating new sounds/images/video, etc.

As we can also see from the recent "Revolver" remix, if this AI isolation software is used on actual multi-track sources, where some things are already isolated on high-quality sources, that isolation software is even more useful and effective.

.

You're doing great work  :) Just keep focusing on that feeling of "oh crap, it works!"

I've been messing with models for the last few weeks whenever I have a free moment... trust me, I respect how great this has come out so far!

Ain't easy to set up!

I was listening to the "Feel Flows" set on a long drive today, and Carl's stunning live rendition of "This Whole World" from 1988 really knocked me out.  CW sings it with such passion and ferocious energy, that it struck me as something that an AI bot can never completely replicate. 

With that said, once the genie of any new technology is out of the bottle, then it's out of the bottle. For good or ill, if something is useful, then, by golly, it will be used.  Let's face it. If Brian had been able to use AI bots for Smile vocals instead of the Boys in 1966, many unpleasant arguments over lyrics would have been avoided.

I'm enjoying these mock-ups.  The possibilities are truly endless.

God Only Knows "1966 home demo" is pretty good.

Does the tag use any of the authentic vocal from the released version?

What seltaeb1012002 really should have done, is held off posting anything until April 1, and then on that day introduced the above mentioned "home demo" as a genuine long lost article. It would have been a cruel but interesting experiment and perhaps with a free pass if posting on that day  :lol

Who remembers the incident during the What Love Can Do era, which also included a remake of God Only Knows, when promotional video of the new God Only Knows was posted, but apparently tampered with in some way to be part original version, part BW remake, and part assistant studio engineer's brother singing a snippet, or something like that? There was plenty of commotion trying to figure out what on earth was going on before the reveal.

AWESOME work on the latest one! So inspiring!
Can't believe this is where tech is... imagine how the geniuses of the era would feel listening to these "ghosts"  :o

I agree, you're only gonna get the AI to replicate whatever you feed it - with maybe some variation, depending on how much it has to go off of. At least in the iteration of the technology we have at this point.

And cool, glad you're enjoying them! I'm working on AIs for Carl, Dennis and Mike now. I'm in South Korea for 2 weeks and Google decided to lock me out of my Dae Lims account this morning without option to get back in, unless I log in from a previously known location... Enable 2-step authentication people!


Aww man, that would've been perfect! Yeah, I was too caught up in the matrix to think clearly enough about how to market the thing. But whatever, the cat's out of the bag, lol.

And nope - that's all AI Brian.


Thank you! Yeah, I mean it's kind of a dream come true for fans. So much potential. And who knows, maybe one day these tools will be utilized for official releases. I could imagine a full AI concert experience featuring "live" vocals, holograms, etc. Imagine seeing Pet Sounds or SMiLE live, or Sgt. Pepper, the White Album, etc... Or even being able to sort of program your own concert at home. "Hey ChatGPT, play me Knebworth 1980, only replace all the songs with Love You in its entirety!"

It had been my dream for years that Brian's band would do Brian Wilson Reimagines Love You - and instead of the synths (which I do love) we could hear those songs with Pet Sounds-esque instrumentals.

Perhaps, years down the road, AI could recreate the instruments/vocals on Love You to make them more of a 1965/'66 Beach Boys style. Not dying to hear this, but it would be cool if possible.
________

On another note, I find the response to these tracks to be incredible. I don't have the brainpower to decipher the varying opinions (nor the patience to read every post relating to this topic). Hey Jude seems to be right that there are a swath of people claiming they love it, and a swath of people saying they hate it, and I haven't really seen many in-betweeners.

While I have really liked what I've heard thus far, I am also hesitant about this stuff. But I also find myself hesitant to adopt a lot of mainstream modern technology, feeling nostalgia for simpler times and how people seemed to be much more inwardly/spiritually content 20+ years ago without Alexas or robot vacuum cleaners.

It's odd, because I have almost equal-opposite reactions to these AI tracks: I really am positively floored by the technology, and it's cool to be tricked into a fantasy of hearing Brian at the piano doing a demo of GoK. But I seem to also want to shut my computer down and travel back to simpler times. Something just feels wrong about it this AI stuff - like a feeling in the back of my mind - and I don't think I'm alone in feeling that.

Again, this is really cool stuff, and I'm happy that AI tech isn't just being used to help students cheat when writing essays. I love that the creatives here on the forums are delving in and trying to create something cool for us fans to get lost in. Can't shake that uneasy feeling though - though maybe that's just normal when a vastly new technology becomes available. Can't imagine what people thought of telephones after thousands of years of writing letters to one another.

First off, seltaeb, if you would like me to start my own thread I surely will. I don't wanna hijack your awesome work, but I feel it's a little redundant to create another AI thread.

Second... I have a question for you all...

Do you Like Worms?
https://youtu.be/Gdeh72-mFoA

I'm working on the full track, as well as a few more pieces of Smile. I have a separate model from seltaeb I've been training myself. I'm not perfect at performing for it yet either! That said, let me know what you think! I figure, since it isn't "real" I should add a bit of my own arrangement while keeping it true to the vision as closely as can be with the limited info and lack of recordings.

So excited to share the whole Smile mix with you all! I have a lot more going on here than just the AI vocals!

I’m currently thinking of this technology as another step on from other advancements in the recording industry. Maybe that’s wrong, maybe this is fundamentally different somehow, but so many technological changes have left people feeling uneasy, or concerned that the music they’re hearing isn’t ‘real’, that I see that as a (currently) more extreme end of a spectrum.

Personally, I love the examples so far and ver much look forward to new ones. There’s clearly a lot of work involved and the results are very impressive and great listening, in my opinion. I love Brian’s voice and a lot of these examples sound very much like it.

Ethically, I’d have a problem if someone was exploiting someone using this, and the industry definitely has form in exploiting talented people, so I can see the potential for all sorts of problems there. I’d also have a problem with ‘new’ material using an AI voice, but purporting to be the original artist etc.

But the potential for using this to fix, complete or as part of a suite of tech in production could add real value, I think. E.g. as a fix for Carl’s “shunshine” flub, or to present what Dennis might have intended for some of his music when his voice had given out etc (e.g. AI Dennis plugged in to the Taylor Hawkins vocal might make the listening experience of that track amongst other Dennis tracks much smoother and more enjoyable).

What better way to test the capabilities of the tech than with the greatest unfinished album of all time? Vey much looking forward to whatever Seltaeb and others can create with AI.

Patiently waiting for AI fix the vocal flub in Soulful Old Man Shunshine...

Great work! Being totally ignorant of how hard the process of making this stuff is, could you possibly make an alternate version with Brian's found melody from TSS? I would love to hear someone's interpretation of that.

Thank you!! :)
Absolutely! This was one of the things I tried, I just don't know how the rest of the lyrics would fit into that scheme  :lol Definitely worth exploring though!

Awesome! Yeah I remember a thread on here from back in 2011 (can you all believe that was 12 some years ago?) where some people tried to piece how that melody would complete itself....but that was just in that one line...like you say, making all the rest of the lyrics fit into that will be a challenge! Looking forward to this and your other Smile AI creations!

I may be jumping the gun a bit on the mix (I will clean that up for the full release!) but I couldn't wait to share my full-vocal-ed mix of Do You Like Worms!

https://youtu.be/_7lOI-tU1lQ

The notes on the video explain a lot, so I'll let it speak for itself. Please enjoy!!

That is absolutely mind-blowing stuff...stay with it and take all the time you need (to quote a certain sourpuss over at the nearest faraway place) to bring it all back home. This represents a fascinating new phase in the still-ongoing saga of SMiLE, and I think if it's handled responsibly and respectfully (as it appears that you and seltaeb are both doing), we'll all have some very intriguiing things to listen to...and possibly even have world peace!  :3d

This was my 2011 attempt to fit the DYLW lyrics to the melody found on the Smile Sessions. It incorporates a bit of Little Pad to round out the line. And while I’m certainly not a singer, I bet AI Brian could make it sound pretty decent.

https://m.youtube.com/watch?v=2RpdU8Bep28

That is a really great interpretation! I just, personally, can't get the other lines in the song to fit into that structure  :lol

Don't wanna overshare, the surprise is part of the fun after all, but I put up a short of some progress.
https://youtube.com/shorts/58GyNt3kCGg?feature=share

Having a lot of fun with this! Hope you all enjoy it as much as I do making it.

Seriously blown away by DYLW with Brian on lead. AI obviously has a huge potential and I hope it can be used to further furthere "restoration" to the Beach Boys sometimes incomplete archives of incredible material. The way I view it, this is no different than reassembling and augmenting the columns for the restoration of the Parthenon in Athens. We know what the lyrics would have been but we just don't have them anymore. That we use our technology to attempt and replicate what might've been is no different.

First off, thank you and thank you to everyone else who has commented on YT :) I really appreciate the feedback and I hope the full project lives up to the expectations set!

Cabinessenceking, I completely agree. When I started BOOKOFF Smile, I just aimed to remaster all the tracks into the cleanest FLAC Stereo mixes modern AI could facilitate. I found myself singing along to the songs and imagining... but being able to hear just a tiny bit of how it MAY have sounded... it gets the brain juice pumping and really starts to fill in the blanks. I hope the feeling is just as strong with the other tracks!

Love to Say Dada... what a track. In Blue Hawaii has some of my favorite vocals on BWPS... I'm giving it my all and really trying to capture the arrangement as close to a Brian sound as I can. Obviously, the more vocals missing the harder the task... DYLW was close to finished compared to these! 

Can I ask whether the idea of using an isolated BWPS vocal is possible, desirable, or effective as the input for processing? I have to admit that reading the steps required for this tech left me no closer to understanding how it works or how to use it, so I have absolutely no idea if using a modern era Brian lead would work, but I’d be interested to know whether that’s beneficial at all.

Similarly, replacing an isolated Carl vocal with AI Brian etc, is that an option and does it work in aiming for a Brian-like sound?

So 'On A Holiday' is one of my favorite tracks from BWPS but my least favorite from TSS (or any other vintage mix) due to the quality. Has there been any improvement to the sound of that track with the AI stuff?

I've never understood why the stereo version of Pet Sounds sounds so pristine (like you're sitting there in the studio) and yet for many Smile tracks it sounds like they were recorded on a portable tape recorder then dubbed to an acetate, played 12 times, etc.

That must be what I'm remembering! Great idea mixing in some Little Pad melody.

Hello! Logging in from the distant past to say: VERY much enjoying the Smile stuff, BUT:

It's BEADED cheering Indians, not BEARDED!

As myself an "Indian" (Anishinaabe / Metis), I can tell you: growing beards is not our strong suit in any way! Many tribes haven't facial hair at all. Beads, on the other hand, are a staple of most tribes!

(In all other ways, that's a thing of beauty, and I'm sure Blue Hawaii will be same)

Ah shoot! You're damn right! I'll be sure to correct this, thank you very much!!

In addition to the Old Man "Shunshine" fix, I propose 2 more:

- Correct the offkey introduction to Spirit of Rock N Roll (from the Hallmark album)

- Add some other BB voices to Brian's I've Lost That Lovin' Feeling

Excellent work you two.

The way this is going, it's insane to imagine where this technology could end up, imagine 60s Brian Wilson singing I Feel Love as produced by Giorgio Morodor  ;D

Are we *sure* Carl didn't mean to sing "Soulful old man SHINE shine shine your light divine on me"?

Hey all, this will be my last update for a bit... I've been working on this basically non-stop!  :lol

(https://i.imgur.com/Gl0Am9Y.jpg)

BOOKOFF Smile will release tomorrow at 5pm EST. I've got a YouTube Premier set up here:

YouTube ruined all the fun! Despite more than 80% of the video being heavily modified content, I doubt my claim against it will win... so I've posted it to Vimeo below! I have *FLAC quality sourced versions* of the songs that I hope to find a way to legally share at some point. As I mentioned, I have put plenty of blood sweat and tears into pretty much every track, so fair use should cover most of it (especially unmonetized)

Here's the link to check it out: https://vimeo.com/817502458
(Download link is in the main thread)

I'll probably make a thread for it in the Smile section of the forum, since I have quite a few more things planned to post alongside it (isolated vocals, the AI model, tutorials... etc)

I sincerely hope you all enjoy it! As much as it is a labor of love to Smile, it's also a labor of love to the fanbase and this site in particular. I remember being here 10 years ago talking about these songs, and here I am today still doing the same!  ;)

It's sure to move a million units... today!

I haven't posted here in years but I did just want to add a note here to say that I think you're both doing a great job and, hey, if anyone wants to PM me to let me know what you are using to create these sounds, please don't hesitate!  ;)

Guess I’m dumb, but I don’t know what you mean by “the main thread,” and I can’t find a download link. Muchas gracias!

Not dumb!! Here's a link :)
http://smileysmile.net/board/index.php/topic,28324.msg682179/boardseen.html#new

Thank you!! I have never in 20 years or so wandered into the Beach Boys media section (or even noticed it existed). I will download this and listen to it this afternoon!

Hey everyone!

So, after giving things another once-over in my head... I decided that I would still like to release the model.
I discuss my thoughts in a video here: https://youtu.be/7svhZEOz0ZE

While I don't explain everything in the tutorial, I do explain just enough to get a dedicated fan across the finish line.
I'm so excited to see what you guys are able to create!

You can learn more here: https://youtu.be/qrojawZbKzU

Not worth of its own thread but AI related:

Chat GPT thinks this is the chords and lyrics to Surf's Up. Bizarre.

The chords to "Surf's Up" by The Beach Boys are:

Verse 1:
F#maj7 B7 E
A diamond necklace played the pawn
A E B7
Hand in hand some drummed along, oh
F#maj7 B7 E
To a handsome man and baton (Bygone, bygone)
A E B7
A blind class aristocracy
F#maj7 B7 E
Back through the op'ra glass you see
A E B7
The pit and the pendulum drawn

Chorus:
F#m E/G#
Columnated ruins domino
B7 E
Canvass the town and brush the backdrop
A C#m F#m
Are you sleeping? Hungry?...Later, friend
B7 E
Cocktail napkin -- found a slot machine

Verse 2:
F#maj7 B7 E
Savior and crucify (gonna keep that airplane)
A E B7
Reasons to live denied (ah, c'mon)
F#maj7 B7 E
Spot light on the fearsome terrain (the aviation nation)
A E B7
T.V. actor on a one-man reign
F#maj7 B7 E
Wave after wave will flow with the tide
A E B7
And bury the world as it does
F#maj7 B7 E
Tide after tide will flow and recede
A E B7
Leaving life to go on as it was

Chorus:
F#m E/G#
Columnated ruins domino
B7 E
Canvass the town and brush the backdrop
A C#m F#m
Are you sleeping? Hungry?...Later, friend
B7 E
Cocktail napkin -- found a slot machine

Bridge:
C#m F#m
Hang ten Hang ten, hang ten Hang ten
C#m F#m
Hang ten Hang ten, hang ten Hang ten
C#m F#m
Hang ten Hang ten, hang ten Hang ten
C#m F#m
Hang ten Hang ten, hang ten Hang ten

Verse 3:
F#maj7 B7 E
The music hall a costly bow (the music all is lost for now)
A E B7
The music all is lost for now
F#maj7 B7 E
To a muted trumperter swamped in stars (swamped in stars)
A E B7
Ages ago in fact -- last week
F#maj7 B7 E
Engineer -- cap in hand -- begged, "leave the can"
A E B7
'there was no room for doubt'
F#maj7 B7 E
With fingers crossed -- the geraniums died
A E B7
Violet stars decayed -- goodbye

Chorus:
F#m E/G#
Columnated ruins domino
B7 E
Canvass the town and brush the backdrop
A C#m F#m
Are you sleeping? Hungry?...Later, friend
B7 E
Cocktail napkin -- found a slot machine

Outro:
E A E A
Hang ten Hang ten, hang ten Hang ten
E A E A
Hang ten Hang ten, hang ten Hang ten
E A E A

I'm very glad -- I've been meaning to do some additional stuff (the Worms melody from the box set, etc) to supplement your Bookoff Smile -- plus "Can't Wait Too Long"!

(PS: hello, I am a former member of the "Smile Community" and used to run the site that preceded this one, The Smile Shop, nice to meetcha!)

My first modest contribution:


Brian Wilson - I'll Bet He's Nice (1966 piano demo): https://youtu.be/mTWsuKlxH8k (https://youtu.be/mTWsuKlxH8k)

Just a heads up - seems most of my videos were taken down by YouTube.

It appears mine will be too, I'm likely going to take down the rest tonight to avoid hitting the copy-strike limit. DYLW was the first one...

Yep, down to three left Dae's channel. Grabbing 'em while I can!

This is amazing.

Suggest you take these videos over to Vimeo, where you can make them private with passwords if you have to...

.

That stinks. Maybe consider posting them at them at vimeo?

I just wanted to drop in and say, especially to the two main creators posting their works here, job very well done. It has people talking about Smile once again, not just talking but really digging deep into the music, and letting it run through their imagination, which is one of the strengths of this music. That's been sorely lacking in my opinion over at least the past 5 years or more, and has been pushed aside by various pockets of this fan base seeming to determine what should be discussed and how it should be discussed and perceived, which is why my interest waned more than a bit in recent years.

This is getting the music back into the discussion via a new technology, and I'm only spotlighting the Smile efforts but it's probably the element of this band's history that has the most lineage with fan creation and technology in the past. Which is why it's also nice to see one of the founders of The Smile Shop visit and comment as well, given that the Smile Shop also took what was a fairly new use of technology at the time to bring together a pretty specific niche of fans and provide a forum to discuss this incredible music. When technology and a passion for specific works of art combine to create new experiences and bring people together to talk, it's a great thing. Again, kudos and thanks to everyone making this happen and creating new experiences. One of the great unknowns in all of the music history I've explored has been what Smile would have sounded like in late 1966 and 1967 when Brian was creating those tracks, and unfortunately the only answers we have exist in the surviving audio and attempts to recreate it or "finish" it decades after the fact. BWPS stands as the creator and collaborators closing the book on a finished work, but the great "what if?" is still what would it have been with the finishing touches applied in 1967 and the voices of that time singing the parts. Thanks to a new technology and creators putting that technology to use, we can come closer to hearing a realization of the illusions many of us have had in our minds for decades. That's a good thing.

Re: YouTube copyright strikes and the like. Unfortunately YouTube, as great of a resource as it is and as much as I love watching it, tends to be corrupt in terms of the way it selectively enforces their rules and regs, how it uses their algorithms to boost versus blacklist certain creators and content, and how their ban policies often make no sense. How some channels are allowed to feature and monetize (i.e. make money) off copyrighted works while others like these various AI Smile and BW tracks are given strikes is a mystery to me, and ultimately it makes no sense on a basic level. And YouTube, like other social media, takes no definitive positions on their rules and enforcement of their rules, allowing some to freely do exactly what others get strikes and bans for doing. I'll go no further, I'll get too political.

But I mention this because of what is currently happening in the music business, via an ersatz AI-created Drake/Weeknd track called "Heart On My Sleeve" that is solely the creation of an anonymous uploader which began to spread online, went viral, and actually started to climb the streaming charts before the labels and lawyers stepped in and had it removed. Neither Drake nor The Weeknd had anything to do with this track, yet it sounded so much like them that their listeners and fans flocked to the song. It's an interesting debate, but I agree with those who are saying this is another seismic wave of an event that is just starting to jolt the music and entertainment business. If anyone is interested in one side of this developing debate, here's the article from Billboard as of yesterday:

https://www.billboard.com/pro/drake-the-weeknd-fake-song-ai-generated-music-illegal/ (https://www.billboard.com/pro/drake-the-weeknd-fake-song-ai-generated-music-illegal/)

I guess we'll have to wait and see how far it goes before AI creations truly become "illegal", whatever that means or whatever form that may take.

In the meantime, it's great to see Smile being discussed and enjoyed again. It's been too long.

Heh. It’s only a matter of time before someone turns “At My Most Beautiful” and “Pancreas” into full-fledged Pet Sounds and SMiLe outtakes.

Maybe I've missed it, but I find it interesting that FatheroftheMan's version of SMiLE hasn't been talked about at all elsewhere. Makes me wonder if the people enjoying these AI creations are simply avoiding the constant backlash of negativity from the usuals, so they've stopped bringing it up in a positive light. Love it or hate it, the Bookoff mix deserves to be talked about. Firstly, AI vocals aside it is a sonic wonder*, and secondly, the AI vocals make it the most complete vintage SMiLE ever mixed (if you're able to pretend/imagine for 30 minutes that the AI vocals are real).

*"AI vocals aside" - want to clarify that even if one doesn't like the AI vocals, this mix is worth the listen for all the other sonic gems it has to offer.

Speaking of Smile, I ran across the Surf's Up instrumental from BWPS just now on YouTube (https://www.youtube.com/watch?v=oNETTrZ_aAE (https://www.youtube.com/watch?v=oNETTrZ_aAE)), were all of these made available at some point? I was only aware of 'Heroes and Villains being' an available instrumental from the 2004 album. Would love to get my hands on all the tracks.

Speaking of which, if anyone from Brian's camp is reading this thread, I would love a 20th anniversary BWPS boxset. Would be amazing to have the album remastered/remixed (apparently the vinyl album sounded amazing but the digital/CDs did not? would love for the digital version to have that warm sound people talk about when referring to the vinyl). Also, vocals-only tracks, all the instrumentals. There could be commentary by Darian and others between some of the bonus tracks. Put the audio of those amazing interviews between VDPs and Brian Wilson from the BWPS DVD onto the tail end of the set.

I just love that I can go to Metacritic, click on "Top albums of all time", and BWPS is still sitting there at the #3 position. This album deserves some love!

.

That is actually my YouTube channel. I also have the Roll Plymouth Rock backing track posted on it too (https://www.youtube.com/watch?v=3jByW0jO4cQ (https://www.youtube.com/watch?v=3jByW0jO4cQ)). In addition to the Heroes and Villains instrumental you mentioned, there were also released instrumental versions of On a Holiday and Wind Chimes from BWPS, but I'm not sure where those are from (https://www.youtube.com/watch?v=p8YBmDVfEn0 (https://www.youtube.com/watch?v=p8YBmDVfEn0)). For the two tracks on my channel, the Surf's Up instrumental was released on the Beautiful Dreamer bonus CD released by David Leaf in 2005 (https://www.discogs.com/release/2909586-Brian-Wilson-Beautiful-Dreamer (https://www.discogs.com/release/2909586-Brian-Wilson-Beautiful-Dreamer)) and the Roll Plymouth Rock instrumental appeared on a bonus CD that came with the November 2004 edition of ESQ (https://www.discogs.com/release/9549874-Brian-Wilson-On-A-Holiday-Roll-Plymouth-Rock-Instrumental).

Hey, so was it you who had the first "Guess I'm Dumb" Brian AI vocal that's since been taken down? Any way you can share a link to the audio file or post it on Vimeo? I'm dying to hear it!

.

Wonderful—thank you so much! I'm loving this version of "Guess I'm Dumb."

There's an unflubbed version on the Feel Flows set. Personally I prefer the Endless Harmony version (the vocals are delivered in a different timbre), but I'm sure with proper editing the two can be combined seamlessly.

Might be worth mentioning that YouTube is now filled with a glut of AI vocals from all sorts of artists, and they are almost all truly awful. As in, far, far worse than the first few things posted in this thread.

Few may be prescient enough to predict exactly how this AI thing will play out on the technology side of things, but I what I can tell for sure is that, as with everything else on the internet, it already appears that everybody thinks they can do this, and they can't, and most of it is a total mess. I'm not sure how much of it is bad ears, versus poor use of the software, versus poor vocal execution, etc.

So I feel pretty comfortable predicting that whatever the best is that AI will have to offer, *most* of what you'll see and hear will not be that. It will be exceedingly awful.

Like, I'm hearing stuff with two or three simultaneous voices, stuff with not a few but many, many weird glitches all throughout. And a lot of stuff that doesn't sound like Brian Wilson or Paul McCartney (or whomever), but sounds like someone with a thick Eastern European accent trying to sound like Paul sounding like John sounding like Paul.

Multiple attempts answer the question I think we've all asked many times, which is "What would it sound like if Paul McCartney sang a Badfinger song, but while eating a sandwich and having a stroke?"

Also, it appears most folks have quickly zoomed past any attempt to make logical/clever attempts at its use (e.g. subbing in a younger voice, or having Paul sing a John song, or someone replicate a known of but lost vocal), and have gone straight to dumb, goofball novelty ideas, like having Kurt Cobain sing the Archies.

A lot of AI crap, but some real gems float to the top. I heard an AI rendition of Lennon's "Grow Old With Me" that was fabulous.

Give me human vocals, with all their mistakes, over this Al stuff. Give me human hands working the fretboard, drummers dropping their sticks, I mean, people thought Autotune was bad!

"I see.....

I see Jamaica, my moon above...."

It's also very clear that certain voices sound worse than others with even extensive information input into the AI.

Some voices have such a distinctive thing going on, it's easy for a fake vocal to pass muster for the average person. I'm sure everybody thinks Rick Astley singing "Billy Jean" sounds "just like him!".

The folks doing these things seem to be having more trouble with certain vocalists. Lennon stuff sounds particularly awful. Michael Jackson stuff sounds way off. Like, literally some impressionists can do this better in some cases.

I think some people *think* they have the McCartney stuff nailed, but you have epic uncanny valley status going on with his "Beautiful Boy" AI vocal, where it's a robot with sort of McCartney's tone, but then he pronounces/sings everything the way Lennon does. Proof that the nuances of McCartney's voice (in its prime) are actually *very difficult* to nail.

Also, no offense to anybody or any region, but it's also painfully obvious when someone with a heavy non-English accent is singing the guide vocals on some of these AI tracks. It's like, "Wow, that sounds 27% like Paul McCartney and then all of a sudden he chokes on his Russian accent!"

The work Giles Martin did on the Beatles "Love" soundtrack, mixing multiple songs up, or the "Smile Sessions" box laying those understated "Surf's Up" code vocals onto "Child is Father of the Man", stuff like that is far more creative and far more impactful.

The best thing AI has offered as far as the BBs so far is creepypasta nightmare fuel like this:

(https://scontent-sjc3-1.xx.fbcdn.net/v/t39.30808-6/344884917_991200292044801_6640508243618564625_n.jpg?_nc_cat=110&ccb=1-7&_nc_sid=730e14&_nc_ohc=8HRirWrUYmMAX-UamO_&_nc_ht=scontent-sjc3-1.xx&oh=00_AfAd-XivN548eX-ep5loqcRk2rB6CRt2lnL33s3TJo4-Ug&oe=6459B37F)

(https://scontent-sjc3-1.xx.fbcdn.net/v/t39.30808-6/345046026_257674096743291_3548329475712795476_n.jpg?_nc_cat=101&ccb=1-7&_nc_sid=dbeb18&_nc_ohc=Id9CN_KWGTEAX8Pp8Ds&_nc_oc=AQlnSqoq6zYX5QBOjUcDk4zT-sl8-Wu7FF9flcUvCtrHQf5ooF_4oiHm-A4RFHOueN0&_nc_ht=scontent-sjc3-1.xx&oh=00_AfB-zKm97nONiPirLU1YznsA3AW9Z4d4zcfmKSRld6qaHQ&oe=6459398B)

These are incredible.  The first image Carl and Mike look like super Bee Gees circa 1977. Can't identify the guy on the left though.

The second image appears to show a young Quenten Tarantino supporting the group. Who knew?

Another one for someone to try :

She’s Leaving Home sung by Beach Boys.
Mind you the “Beatles” version of God Only Knows is not good even though Brian’s Twitter shared it.

Has anyone mentioned or suggested ‘Proud Mary’ with a BW early 2000s voice. Maybe even a whole R&R album with some classics he may or may not have hinted at?
I’d say it’s the only way we will get this long promised project now.

Walk on By...holy crap. That is all.

This is stupid and a great shame. I hope you can make them available elsewhere. Ol Man an the Bacharachs are superb.

I told my wife, when bookoff smile came out, that the best thing that could happen is that it would be "outdated by summer" and it seems my highest hopes are about to be met!!

https://youtube.com/shorts/YwGAn9QSVuk?feature=share

Hearing Mike sing those lines feels so right. I am excited to hear more.

Holy moly. That is incredible.

And FOTM, that's an awesome attitude to have. That being said, I think you've solidified your spot in Beach Boys lore for being one of the first to ever attempt to fully complete that album (and being very successful at it).

After hearing Bookoff Smile and 'Thank Him' (and the 'In Blue Hawaii' track), I am becoming more comfortable with the idea of The Beach Boys and AI coming together. There will probably be a lot of garbage, but there will be a lot of treasures that most of us will appreciate. Case in point:

Hearing 'Thank Him' just blew my mind - ever since hearing that old lousy-sounding recording, I have wished that Brian and the band would've recorded a full version of that. And a few weeks ago I was randomly looking up songs on Youtube and came across that. It honestly took me back to that feeling I got when I first heard 'Don't Worry Baby' in proper stereo. Crazy time to be alive haha.

I'm wonderin', has Rio Grande been AI'd yet??

I've probably mentioned this before, but the idea of AI helping create a Pet Sounds-era-sounding Love You album is intriguing to me, and in addition to that I'd love to hear the '88 album given the same treatment (and saying that, I love the synths and production on both of those albums, I'm just intrigued by the idea of making them sound like they were recorded in the 60s with Pet Sounds era instruments).

If only AI could re-write Landy's contributions to '88 ;D

Can we  AI OSD singing Mike’s songs? ;D

Trailer dropped today!

https://www.youtube.com/watch?v=TXxn0_rT8iM&lc=UgyRzZe-VV4VyG0ND1p4AaABAg

Oh yeah SB! It's gonna be the box set of the year! Wait until you hear my AI version of "Summer of Love"!!!    :woot :woot :woot

I know this wouldn’t happen either way, but I’d rather have someone like Darian, perhaps with or without contributions from Brian and Al, handle this sort of a thing. Brian’s band knows how to recreate this stuff to a T, and maybe with Scott Totten and John Cowsill no longer with Mike, something magical could happen. After all, Al did tease a possible Brian album next year. Darian and the band could handle the arrangements and help Brian and Al (if they chose to contribute) with lyrics. Obviously, I know this is extremely unlikely, especially potential input from Brian and Al, but as Lonely Summer said above, I don’t want to listen to AI stuff and justify its existence/support future creations. I’m sure those of us here are doing this with good intentions and a lot of hard work, but this stuff is too dangerous to me. Someone wanting to attack any of the band members could create deepfakes audio instantly, and that could spread fast. I think AI should be left to demixing, as HeyJude has said, and if Mark Linett, Alan Boyd, and Howie Edelsoj could get a hold of that (type of) technology (for a reasonable/fair price), that could be interesting for attempting new mixes of 62-63 and 65 material (LDC, Today, SDSN, etc) if they ever receive archive releases down the line.

Looking forward to this.  Dae Lims is mad talented and putting the whole AI thing to good use.  His "Return To Pepperland" reworking is phenomenal.  I just hope he has a backup plan for when the Tube inevitably shuts this thing down.  Which they totally will.

Darian already did this back in the '90s on his version of "I Wanna Pick You Up," and it was revelatory. And as good as you'd expect which is to say very good.

The SHOWMANSHIP

https://youtu.be/PImkald8cQU

We're going multimedia on this... so hyped!

Also his work on "Do You Have Any Regrets?" Brilliant!

This is fantastic, I can’t wait to hear the full SMiLE work you’ve done once complete. Really impressed!

The more AI deepfakes that are created, the more people will distrust new audio/video. So, in my opinion, it's a good thing that there are a plethora of people using AI to deepfake The Beatles or The Beach Boys - because if something nefarious (and fake) does surface, there is a vast swath of songs/audio evidence that will show how easy it is for people to deepfake, and thus people will be less likely to believe that the nefarious content is legitimate. Whereas if deepfakes were a rarity, deepfake nefarious audio/video would be more likely to be believed by the masses.

In the AI era, I hope that the widespread use of deepfakes will help people to not rush to judgement, not jumping on the bandwagon, that this era will help us realize the importance of first-hand evidence and common sense. For years the media has edited video and audio of events to set narratives, governments "deepfaked" evidence to push them into war(s); I hope that with a new wave of deepfakes and mis/disinformation that the general populous will grow more discerning of what they see in the media/social media world.

And as for the specific use of AI in The Beach Boys world, I think that it is overall harmless. For those that want to stick with the classics, these YouTube AI songs don't affect the beauty of the original tunes. For those who are intrigued by the creative possibilities of this new technology, we now have the ability to hear things that never were. And I completely get why some really don't like this stuff (both the moral arguments and that they don't think it sounds anything like the original Beach Boys), but thankfully there's nothing forcing them to listen to this music, either.

Yeah I have conflicted feelings about the usage of AI in entertainment.  On one hand it is artificial and not "pure" so to speak so there is an ethical question to be had.  Do we really want new Beach Boys material artificially sung by "the Beach Boys" long after they're all gone?  On the other hand it could be useful to clean up audio or isolate specific tracks (we could potentially finally get a fully-realized stereo mix of "Good Vibrations").  That's what they're doing to John Lennon’s vocals on that upcoming Beatles track they've been hyping. 

As for all the deepfakes popping up on YouTube, they at least serve as a fun curiosity more than anything else.  And when you have talented guys like Dae Lims who but AI to good use and do something constructive with it then I'm happy to listen to even if it's not authentic.

Dae Lims' SMiLE is live.  Grab it before it gets shut down (no pun intended). 

https://youtu.be/n7W_2JS12CY

Early into it. Worms is a psychedelic dream and a highlight.

Edit: Ok, nothing is beating Surf's Up. WOW.

Awesome! now this is an example of how to use AI in a creative way, as a tool. Monumental work really👏🏼

Not enough time to share all my thoughts right now... but what a daring, incredible, and creative work. Absolutely monumental.

I won't belabor the point at the moment regarding how accurate or inaccurate these AI vocals sound to me, but I'm confused as to why a fantasy land "what if" Smile where the big selling point is attempting to fake how the Beach Boys would have sounded singing this stuff in 66/67 would leave out stuff like "Look/Song For Children." I think hearing that bridge between "Wonderful" and "Look" on the 2004 "Smile" was one of the biggest moments on that project, one of the most "a ha!" moments.

I also lose patience when AI vocals are clearly in place for songs that already have perfectly fine vocals. Not sure what the point is of laying a fake "Brian 1966" vocal on top of stuff like "Wonderful." I mean, I guess it allows for better stereo mixing options? But when we're going past the "What if?" stuff and I find myself just listening to someone run a vocal through AI to replicate a Brian Wilson voice that already exists, I quickly lose patience. Why am I listening to an algorithm sing "Wonderful" and "Surf's Up" when *ALL* of the constituent elements on those songs already exist? This isn't "Mike singing 2004 Smile lyrics" type stuff. It's me listening to a person and/or an AI algorithm replace the actual tracks that already *do* exist.

There is obviously far worse stuff on YouTube as far as AI vocals, and I appreciate things being labeled as "AI" (I guess we'll see how much this stuff proliferates via repostings that aren't properly labeled), but, well, I'll hold off on larger piece/critique until I can elaborate more. I will say that I guess I kind of understand why some fans might have, or *want to* have a "holy s**t!" moment hearing this, but to me the whole thing is far more murky and complicated on multiple levels and, if I forego the complications and take the reductive approach in condensing this all down into a simple point, I'm just left wondering not so much "what's the point?", but more "should I burn much time listening to fake voices when the real voices exist?"

To echo Hey Jude a little and repeat what I said on Endless Harmony, a brave move to replace existing vocals with AI generated ones.

Regarding Surf's Up, it is one of my absolute favourites, an yet - is it greed? - I can't help but wonder if a somehow "better" Surf's up could have existed. And the suggestion of harmony vocals during the "aboard a tidal wave" section is tantalizing.

Interestingly, this particular work actually has enough decisions that are outside of the “do these AI vocals sound accurate?” debate that it’s actually in some ways easier to debate those.

I think that version of “Surf’s Up” sounds like kind of a mess. I guess, because it’s using a weird AI-assisted vocal, it manages to sync up the “Brian vocal” to the ’66 backing track a bit better than either fan-made sync jobs or the official releases that sync Brian’s ’66 demo vocal to the ’66 backing track. But just about everything else sounds wonky to me. The weird key change at the end sounds totally off. The key change on that discovered “Wild Honey” demo of “Surf’s Up” works because of the nature of that solo piano Brian performance. But pitch-shifting the entire track on this “finished” version sounds really strange.

I obviously have my own take/ethos on this song and the material in general. Much like I think Brian/BB fans ignoring “Guess I’m Dumb” because Glen Campbell is singing it and they can’t have a Brian lead vocal is kind of missing out on a great song with a nice vocal (Glen sounds good on it, and sings it like Brian), I think there’s a point where the novelty wears off on trying to shove Brian on the lead on the first movement of “Surf’s Up” and it makes more sense to listen to what I think is the best version of the song, the 1971 version with Carl singing it. Carl sounds amazing on it, it retains the “Smile” vibe while fleshing the song out with a *very* important bass line near the end, and amazing group backing vocals, not to mention Al’s excellent part at the end. I’m not opposed to the Brian “what if” vocal as heard on the “Smile Sessions” (and that fans had done years before to varying degrees of success). But at the end of the day, in a world where I’m happy to admit when the “official” version of something is *not* the best, I can safely say the 1971 version of “Surf’s Up” released on the album of the same name is the definitive, and best version of the song. Better than the ’66 paste jobs, certainly better than the weird AI versions, and better than Brian solo versions from the 2000s.

I mean, I guess it takes some balls (it takes something anyway) to go beyond the AI generated leads for non-existent stuff and just decide to erase actual Brian/BB leads that are definitive and re-do them with weird AI, and then also start grafting on weird overdubs. I think this stuff, already questionable with the AI vocals, quickly becomes a weird mess when it’s getting futzed with on multiple levels in multiple ways. Yes, yes, I know, “it’s just for fun”, etc. I’m not saying anybody wants to delete the BB albums and put this out. But I gotta rate it and analyze it for what it is, and as much as some fans seem to really, really want this to be a mind-blowing thing (and if it does that for you, I’m of course not going to stop anyone), but to me it’s a weird melting pot of the questionable AI vocal stuff, brazen but ultimately sub-par artistic liberties, head-scratching “fan mix” decisions to move around tracks, and a weird overall goal/ethos (e.g. This seems to be a weird, truncated “Smile” that’s missing key things like “Look” and “Holidays”, etc.).

And, in the realm of the actual AI voices, there are a few here where I honestly couldn’t tell who it was trying to sound like. I think at one point it’s going for 1966/67 Dennis, but I guess the AI has less 60s Dennis vocals to learn from, so that one sounds particularly off.

Also, one more thing and I’ll take a breather: While I guess on some level it’s interesting to hear the alternate phrasing/melody for the “Do You Like Worms” verses employed, I think the 2004 iteration sounds much smoother and less clunky.

I think the whole thing is amazing. A lot of very creative decisions have been made throughout, and I love some of the new backing vocals. On just two listens nothing about it annoys me, except perhaps the strings in Surf's Up which sound a little weedy, and the French Horn tag at the end of same which is just silly, and breaks the mood somewhat. Overall a stunning effort.

There's definitely some spine-tingling moments from this version.  I'm not sure that I like the overall sequencing of the songs though and was there any reason to omit "Good Vibrations" which was the song that tied the whole thing together?  Also some of the wacky tempo changes seem jarring to me.

But on the plus side it was nice hearing a fully-realized version of "I Wanna Be Around" and kudos to Dae Lims for being daring enough to update "Surf's Up".  This is probably one of the most remarkable fan edits of SMiLE that I've ever heard.

Agreed. I'm not into the sequencing either, but an amazing mix. Surf's Up is inspired.

Like it or not, AI is the future of music production.

I caught a little bit of Smile AD's livestream today where he talked about the making of these tracks. For those who don't really dig this mix, it may be worth scanning through those 'making-of' livestreams. Not because it'll change your mind about how you feel about the mix (though it might), but it really shows the amount of time, energy, and care that went into this project. Just from a technical perspective, this mix is an astonishing achievement. And hearing Smile AD talk about his reasoning for mixing things certain ways, adding in AI vocals in certain places, it really makes sense once you listen to his reasoning.

I was blown away at how many tracks he had for each song - double tracked vocals, bass lines that were beautifully (and seemingly perfectly) extracted so he could mix them better. The clippings/edits he did just to one of his vocal tracks was very cool to see. Other highlights were how he added in various sounds to help achieve a more impactful sound. He would scroll up/down through the tracks for whatever song he was talking about, and I feel like sometimes it was an endless scroll because he had so many tracks (in a good way, it just shows how much work he put into each of these songs).

The artist/musician in me is completely floored by the creativity of this mix. There are so many beautiful moments. Not to mention how AI Brian, Mike, Al, etc sound just like the real thing.

Some songs I'm not too crazy about, some have really grown on me (especially Do You Like Worms - I love how he tackled the melody of the 2nd half of the verses - but I admit it took some getting used to!). But overall I love how genuine it sounds - genuine in both quality and in the happiness the tracks evoke...and I can't stress that enough. It really is worthy of a smile!

Above anything else, I have to commend Smile AD for not being afraid to try new things.

I know this wasn’t directed at me, rab, but I realized I should (briefly) further explain my thoughts on the AI topic, since I can’t remember if I said one specific part before. I am not denying the effort that some users such as Smile AD are putting into these projects (though I will still insist that many other users are just doing something without effort for the sake of it). I definitely understand the intrigue and what if’s regarding the band’s music. There is likely some alternate or something that never existed or used to exist that each and every one of us wants to hear. However, I don’t understand why we’d listen to AI recreate The Beach Boys when we have the real thing, and so much of it? The amount of material the band has released (solo material included) is a ton of material, let alone the amounts of alternates and unreleased and/or previously unknown material surfacing constantly on the archive sets (since 1981/1993, but especially since 2013). Ok, I think I’ve bored you all enough now.  ;D I’ll keep quiet.

I do understand where you're coming from. From my point of view, while the real thing is SO much better, but it's also kinda neat to step into an alternate world for a few minutes and pretend The Beach Boys completed something that never was.

The only three projects associated with this AI trend that have truly captured my long-term interest are the recreation of 'Thank Him', and FOTM's & Smile AD's Smile projects. For me, it's the act of completion of something that should have been completed long ago. So I think that's where my interest stems from and likely where it will stay. The Beach Boys recording SpongeBob music, or Brian covering the Ronettes - while technically interesting, doesn't hold the same value to me, because it's not completing something that was, historically, almost in the grasp of the band.

But that's just me. Honestly, I'm still on the fence about the morality of it all, and I worry about what could come from all of this (the entirety of the AI boom). For now, I'm glad that people are using it creatively, in an attempt to bring more interest to the band.

https://www.youtube.com/watch?v=MDcJKC5RWpI (https://www.youtube.com/watch?v=MDcJKC5RWpI)

Wanted to post a link to my AI-enhanced version of Worms. I tired to keep it a little more minimalistic -- my verse vocal is a little bit like a stripped down version of Dae Lims' version, that was pretty much where I was heading too. Also added in the "East or West Indes" vocal at the end. I think maybe it's a little more historically accurate than his but less fanciful? If that makes sense?

I am working on the others!

The Daelims AI Smile is my favourite ever Smile listening experience, without a doubt. Smiley Smile, GV box Smile tracks, bootlegs, the H&V demo from Endless Harmony, BWPS, Smile Sessions etc - I love it all, but I've never enjoyed a fully formed mix as much as I've enjoyed this Daelims one. Given where things are with the development of AI and how quickly this was created, I'm stunned at how good the vocals are, and the Q&A youtube stuff was really illuminating in terms of choices and why AI vocals were used in certain places. The expansion of the backing tracks to enhance etc is just as impressive to me.

I'm not a purist, but I can enjoy the integrity of the Pet Sounds mono mix (Wouldn't It Be Nice in mono is as good as pop music ever gets) and the Stereo, and the sessions etc. BWPS is stunning as a completed album, but the vocals will never compete with the Beach Boys own voices, no matter how well executed. I have always found any blend of BWPS vocals with original tracks jarring and distracting in a way that I simply don't find with this AI mix. There is no historically accurate final Smile mix by the Beach Boys and there never will be, but my own view is that this is the closest anyone has come to a full-album mix that is cohesive and not hampered by loads of missing vocals or varying fidelity/sonic issues etc. This mix is an extraordinary piece of work and I love every second of it.

There is no perfection possible with this sort of thing - we'll all differ in how we experience it and what we prefer, and I'm still working out what I think about the ethics of AI in general etc, but for anyone who might want to hear the Beach Boys sing In Blue Hawaii, then what's not to like about this? Or a version of You Were My Sunshine without Dennis being squashed in the mix - listen here for that too.

I don't disagree particularly with views of those that aren't so keen (different perspectives etc) or who simply prefer to hear the original voices rather than AI recreations etc - but this whole topic (smile, this board etc) was always very much based around a celebration of a lost album, missing pieces and the many "what if" scenarios and this mix responds to so much of that with an incredible sounding response. I have literally woken from dreams in which I'd heard rediscovered original Beach Boys vocals on some of these tracks that were either never recorded or lost, but in this mix I can hear them and they're just as amazing as the things I heard in my dreams. To someone like me, this mix is an extraordinary gift and I am beyond grateful that someone took so much time and effort to make something so wonderful.

Love the East or West Indies bit! Good work! Also just picked up on your Smiley Smiley remix - great to listen to.

I like the idea of using AI to fix or complete BB/BW songs.  With respect to fixing, the intro to The Spirit of Rock and Roll (by BW) on the Hallmark album is a top candidate for repair.  Other candidates include several songs from BW's Getting In Over My Head and the Danny Hutton duet (Hold Back Time) on the VDP/BW Orange Crate Art album. 

Jon Hunt's "Smiley Smile" remix is DEFINITELY worth a listen...very, very cool. Smile with the Smiley Smile production and sound. For those who haven't heard it:

https://www.youtube.com/watch?v=Fh5v80aaweM (https://www.youtube.com/watch?v=Fh5v80aaweM)

AI and creativity raises a fascinating complex of issues (in part because it is also so harrowing in its implications). That said, I think we need to give folks leeway to try things out as much as possible--within some established and fairly exacting set of parameters. In saying that I am focusing mostly on music/musical reconstruction as we've begun to experience it in general (and, of course, specifically with SMiLE--a project that in its still-infinite aftermath is made for AI tools and approaches.

For SMiLE as tackled by Dae Lims, we're probably going to need a full transcription of the vocal reconstructions in order to fully assess the ultimate level of efficacy in his SMiLE. That may be possible for some unusually enterprising individual who can collate his lengthy explanations of his process. Of course, such a compendium won't eliminate lingering skepticisms about the decisions made (particularly the choices between AI vocals and actual vocals). But we probably need to see all of that in some kind of consolidated flow chart or other type of summary before we can draw any truly substantive conclusions.

One thing that I hope we'll get from one of the SS stalwarts is a song-by-song interpretation/analysis, which doesn't seem to have emerged yet. I wonder if one or two of our folks who are established as research experts might take a crack at that at some point. Those folk may wish to tackle the subject of "liberties" taken and how likely some of those choices might really have been, just for the sake of grounding that in a way where it can be dealt with and not continually threaten to sidetrack the discussion. Dae Lims made some major choices (particularly in jettisoning GV, Look and Holidays) in making what we might call the January '67 version of SMiLE (although with some of the '71 "Surf's Up" managing to travel back in time).

It's clearly a very personal mix, and a labor of love, and we should grant him license to build such a version as a testament to what we can never know, but still have an intense desire to imagine--and now we can hear so many more things that can be imagined than was previously the case. I hope this is just the beginning of an effort to produce "editions of the in situ SMiLE" that will have similar astonishing results. BWPS is clearly not the SMiLE that would have emerged if it had not gotten sidetracked, pole-axed, abandoned and reconfigured, so all's fair in love and ProTools. Let this be just the beginning!

And on such a note, let me second GF's endorsement of the "Smiley" SMiLE by Jon Hunt. More food for thought, more intriguing sonic variations for music that still has mysteries left to be revealed to us...

I’m not trying to be reductive, but I think that if one take the maker of these AI projects at their word, and I see no reason for them to lie about their motivations/reasoning for their decisions, then I don’t think analyzing it is too complicated. Meaning, the people on this board know their stuff, and can pretty immediately spot what’s AI and what isn’t, and also know the stuff well enough to develop an opinion about what they like and what they don’t like.

I guess, for folks that are *really* into this stuff, flow charts and detailed notes might make it all more enjoyable. For me, as someone who, setting aside the more broad implications of the use of AI on music (a most decidedly complicated topic), finds projects like this to be more a curio than anything else, I don’t need voluminous notes and charts. I can detect what’s going on more than well enough to *understand* it; I don’t need the precise reasoning behind each of the decisions. Again, this is all said with an understanding that I trust the maker of the project is being honest and that they certainly feel they had reasons for the decisions they made.

I'm not too concerned with giving people leeway or license to make this stuff. They're going to do it and are doing it. Nobody is stopping them. I truly actually do keep an open mind about this stuff. I think the AI music stuff online went off the rails long ago (if it was ever on the rails in the first place). This particular maker of this AI Smile, and their previous work, is really an anomaly as far as time spent, knowledge of the source material, and proper labeling of the end result. I appreciate all of that. But even this maker's stuff has kind of shot past what *I* feel is a possibly useful ethos in making the stuff, and has landed in fan mix/fantasy tracklisting territory, all while using those AI tools. As I said in a previous post, it's not so much a question of "What's the point?", and more a question of "Is my time well spent listening to this stuff rather than X?".

I think one of the issues here is the push and pull between AI creations as a technical application of software, versus being actual “art.”

We all have varying degrees of familiarity not only with AI-sourced material, but varying degrees of experience coming up upon (or against) people who create it and/or are very bullish about it (not surprisingly, those two things are usually one and the same).

However many months back (not that it has let up more recently) when people started posting visual/digital art and saying “There, see? I’m an artist now too!”, it immediately got pretty icky. There was and is a large chasm between people who feel those people are *not* artists versus those who feel they are, and also a wide gap between the people who think the art they’re executing is just awful versus those who think it’s amazing.

When it comes to AI-sourced music (meaning mainly in this case generating fake voices meant to sound like a real person, using AI trained on recordings of that person), there is a very similar gap on the *reception* side of things. Some people listen to this stuff and think it sounds amazing, others think it sounds awful. Also, within the group of people who think it sounds *accurate*, there is some level of disagreement on the choice to make the stuff, or the implementation of it.

My subjective observation has been that, especially because a lot of the AI music stuff has been admittedly done to sound like a person, trained on recordings of that person, the people *making* this stuff have generally seemed to be much less egotistical and display much less hubris about what they’re doing as opposed to, say, visual AI artists. I’ve gotten the sense in the past that, when seeking out a specific prompt (e.g. make this recording sound like *that* person from *that* year), the people making this stuff have given off a vibe less of “I’m an artiste!!!” and more like “This is a technical process that requires some level of skill.” While that doesn’t address the misgivings some have about the choice to make it in the first place, it is a lack of bullishness that I appreciate.

However, regardless of the intent or the expressed level of humility or hubris from any given person making this stuff, when it goes from a specific prompt of just filling in a vocal to sound like a specific person at a specific time, and moves to making subjective, artistic decisions about melodies, chords, arrangements, and eventually ends with wiping actual existing vocals to then replicate those *very same vocals* with AI, that’s where it gets a little more dicey. I’m not even talking about the morals/ethics and all of that. Just in terms of how I feel about the end product, that level of “fan mix” added to the “AI” thing makes the whole thing just a lot more wonky.

And then that leads to the question of how to talk about this stuff. I have no desire to sh*t on someone after they spend hours and hours working on something, whatever it is. But also, the presentation suggests wanting people to consume it and provide feedback. So there’s an uncomfortable point where it ends up being a reaction of “Uh, yeah, I’m not sure if it’s worth that much effort to produce *that*.” But clearly there is an audience for this stuff, so I don’t feel too bad saying maybe *I* think my time is better spent listening to something real and made with whatever original artistic intent it was, rather than listening to too much of these large scale AI music projects, because I don’t think my mixed feelings about it are going to hurt anybody’s feelings.

I'd say that 2006 Hallmark take on "The Spirit of Rock and Roll" is too limp and lethargic to salvage. It's got more problems going on than some wonky vocals.

I also think the "Orange Crate Art" material is fine as is. It was already futzed with back in 1995, and frankly the "1995 Brian voice" is a specific flavor that I think is worth preserving. Note how his '99 re-done lead on "You're Still a Mystery", for all its smoothness, lacks of the energy and quirk of the '95 vocal.

Yes, I believe "Hold Back Time" is the one that sounds wonky enough that one rumor persisted for a while that the album "accidentally" used a guide/temp set of vocals. I don't know if that's the case, but if it is then it implies a later, more polished set of vocals exist somewhere, which is the main thing I'd want to hear. If it actually exists.

I think the only way to "fix" this later-era solo stuff would be to still make it sound like latter-day Brian, just more polished. And I don't know what the efficacy would be (or possibly the point would be) of making "2004 Brian" sound more like, what?, 1998 Brian?

I think at this stage that the AI-assisted isolation/demixing technology is a million times more interesting and useful (and ethical if that's one's concern). I love Brian's productions, but it would be interesting to hear some of the backing tracks pulled apart more. Joshilyn's videos have certainly opened many eyes to all the intricate stuff buried in those recordings/mixes.

Another interesting application? Vocals. AI is getting better at separating multiple singing voices locked into one microphone/track. I'd *love* to hear the harmonies picked apart further. I was recently listening to some Beatles vocals pulled apart (e.g. the separate Paul, John, and George parts on "Eleanor Rigby" and "Because"), and it's super interesting. It really drives home how those middle parts, those George Harrison/Al Jardine parts, are some of the most difficult.

Some thoughts: https://youtu.be/nrEoGv13Cm8?si=yovpp-n2WHO2GqL1

Can't wait till you guys hear the rest of this, but I wanted to show a bit of how the sausage is made.
So many folks have the wrong idea about these tools!

https://www.youtube.com/watch?v=Tnm_DW0azIc

Any guesses what this goes to? Hmmm...

Oh! And these models, much better ones than my Brian model, will be available soon as well.  ;)

(https://i.imgur.com/B7fRqMG.png)

BOOKOFF SMILE PLUS - Completely new ai models, tracks, and mix! Coming... when it's done!

https://www.youtube.com/watch?v=ep_6aF32OjI

Song for Children - featuring Al, Mike, Brian and Carl!

https://www.youtube.com/watch?v=C_LBW8qDZFM

><

Fair enough :/ This is literally a bedroom project ahaha, I think it sounds pretty good but... valid criticism!

I think the constructive criticism here should be based on a better understanding of the "training" process and how it's applied to each song. In comparing Mike's work to Dae Lims (you folks out there realize, of course, that "Dae Lims" is "Smile AD" backwards, yes?!) we should ask whose models/process has produced the most credible "AI surrogate" vocals, and why this is the case--better vocalists being used as "source," better/more exhaustive-thus-closer-to-the-sound-timbre-vocal style of each actual BB, etc.

AI has got to get closer and closer to a more authentic (if that's the right word, and if it's actually possible...) replication of the BBs voices in order for re-envisioning of vocal arrangements that never made it to tape in '66/'67 to pass muster. Many of them kinda sorta do that now, but this is the battleground for whether this promising area of work can go beyond the stage of being a fascinating curiosity. Dae Lims has (IMO) gotten there on more than a few occasions; Mike is lagging behind in that respect, but his work still holds promise.

It's clear how much work is involved in all of this, so none of the above is meant in any way to be perjorative or judgmental, and I hope it's not taken that way. Hoping that folks will keep at this, as the music is special and unique enough (and still tantalizingly incomplete despite the existence of BWPS '04) to warrant ongoing investigation and engagement...

It's the advance in the technology that is pretty stunning, in that none of this was available nor was it happening even a short few years ago. What did exist was pretty bad, and now at least the median standard is that it could be called "convincing" in the average AI modeling products. I compare it to the various technologies that were used back when they cleaned up the Beatles BBC tapes and released them: At the time, it was new and exciting, however now it can sound kind of wonky and primitive. And look where the extraction and cleaning technology has progressed solely in the Beatles' universe, where Jackson's Get Back and the newest remixes used that same extraction software to basically reinvent the original mixes.

So that's 30 years difference, from BBC and Anthology to Get Back and Revolver. Compare that to where AI was in 2021 to where it is now, and it's staggering in how fast the tech progressed. I predict within a year, there will be AI that will allow fans to create full videos for a random song, using AI technology to create the imagery of the artist, the music, and whatever visual cues the maker wants to feed into the AI generator to create the video. And then consider what that will mean for film and the whole process of making a movie from writing the script to shooting the scenes to editing it all together. That's the area which I think many have issues with. It's already affecting the world of academia where ostensibly a student could write and hand in a paper generated with AI and the teachers/professors would have no way to know if it were the student or the AI creating it.

No offense meant toward thelonleyc, but to ask if people are using professional singers (that have the exact same muscle and breath control as the real Beach Boys) on what is clearly an unofficial project is kinda ridiculous. For this being a bedroom project, it is incredible what we've heard. I don't think SMiLE AD or FOTM claim that their works are without hiccups/flaws, but I gotta say that even with the flaws these tracks sound incredible - especially considering what Guitarfool wrote above: "none of this was available nor was it happening even a short few years ago"

No doubt constructive criticism should be given so these guys can work out the kinks in their AI models, tracks, etc. But asking these guys to hire multiple singers with 4+ years of studio experience is asking a bit much considering none of these guys are making money on any of this (and the tracks can disappear from YouTube quickly).

FOTM, I really like what you did with Song for Children! Those background vocals at :45-:50 seconds are amazing!

><

Well I see your point about it being a musical community. And I agree that it would be awesome if more people from across The Beach Boys musical community were more involved with these projects (kind of like The Beach Boys cover collaborations we did years back on this forum)...and I'm sure, in time, they will be. That these are the very first steps into this new musical landscape says a lot about the potential of this new technology.

Also, I'm not sure what Smile AD's musical background is, but there is definite talent there (I implore people to watch his behind-the-scenes videos on YouTube where he uncovers the layers of vocals/instrumentals he worked on for his Smile mix). And FatherOfTheMan has been making records for the last 10 years? I think it's just new technology, combined with time limitations, etc..

That being said, if it's not convincing for you then it's not convincing...and there's nothing wrong with that. Honestly, some part of me is happy that we can tell the difference between AI Brian and real Brian. Another part of me is excited to see just how convincing these tracks will become in the next few years.

All good conversations above, I would always love to collaborate with anyone interested in this tech!
I've noticed so far, most folks just want to pay me (or just beg) to have their songs made with Paul McCartney's voice... or a specific request from the Beach Boys...

One of the things I most want to show is that, honestly, ANYONE can use this tech. It really isn't that complicated! I am a programmer by day, but I'm just a bedroom musician by night. I do believe that is the biggest leg up that Dae Lims has... his knowledge of sound engineering, mixing, and just his skill as a musician is a HUGE asset to his productions with the AI. That said, I'm just a hobbyist and I believe any of you (with the right amount of passion and effort) could make some bomb stuff with these tools.

I've been trying to make more "behind the scenes" stuff for this project. A lot of folks, here and elsewhere, would assume that I used BWPS vocals to "de-age" Brian... but that isn't the case. While I personally wish to stick to the words and arrangements pretty closely, I am singing all the parts which as mentioned is a blessing and a curse in some cases.

I think it's important that people know this. It takes musical skill to do these things, and it's the main thing I'm trying to improve to catch up with Dae Lims... who has a pretty decent lead on that skillset!

Little clips like this may help to explain to the general public how this stuff works, and perhaps legitimize the tech a bit more?
https://youtu.be/8F5ljjBppSk

Is there a reason you wouldn't use BWPS vocals as the input where possible?

Made a quick video to answer this question :)

https://youtu.be/SED0pLJj844

Interesting. So it sounds like there would be potential to get good results from a later day Brian Wilson vocal if one could get their hands on a clean and dry track.

This provokes another question...what are the inputs for the machine learning thingy that generates the AI model? Presumably the ideal for that piece would be raw and dry vintage era Brian Wilson vocals.

Wow, the Dae Lims Do You Like Worms has to be the best version of that song I've heard thus far. That really blew me away.

I'm at a bit of a crossroads here, and I can't decide what to do.

I have a finished version of On A Holiday, with Mike and Al sharing the lead. On the middle section, the pirate rap, I had van dyke parks beat rapping the part I was so excited about it and couldn't wait to share.

I was encouraged to send it to the man himself, and I was excited to receive a reply. I told him that I wanted to make something special to celebrate my love of BWPS and the original tracks for the 20th anniversary. I made sure to let him know that if he didn't like it, I wouldn't release it as is.

Very respectfully, he responded that he didn't feel the track honored the legacy and that I should do something else with my talent (paraphrasing). He was very kind and I want to stress that I don't feel anything negative towards his response.

However, I can't help but feel that if he isn't a fan of the project... I should just drop it. I don't feel like a compromised version of my vision would be worth following through with. I'm not sure what specifically about my project he didn't like, like just the use of his voice on the track, but I just feel off about the whole thing now.

I know others have mentioned he gave some encouraging words to Dae Lims Smile, so that's cool... oof!  :lol ::)

Thoughts?

><

Never should've asked him. *This* is the thanks integrity gets you!

Is this Brian's vocals from BWPS getting the AI treatment? Or is it Dae Lims' voice?

And what apps are being used to do this? Or is that some sort of state secret?

Dae Lims and I both sing our parts, for reasons outlined in my video clip above.

AFAIK he is also using RVC, you can find some easy tutorials on YouTube, it's not that tough to get set up!

Well, despite the hang ups, I decided not to let my existing hard work go to waste! I had a different idea for the "rap" section... I hope you all enjoy it!!

On a Holiday!
https://youtu.be/Al3F7OD-fk8

I had a go of it a few weeks back:   https://www.youtube.com/watch?v=7wOS9yPVgoo
For the rap part I used Paul McCartney, but I much prefer yours. Singing that pirate part just feels right.

I remember seeing yours on YouTube a while ago! Great to see others taking a swing at it. I definitely think it's cool to hear Paul doing the jaunty pirate rap ahahaha

So glad you enjoyed my version, thanks for listening!

Hey! Just wanted to alert y'all -- I have a new version of Child Is The Father up on YouTube, with some AI verse vocals. I have always posited that the coda melody for "Surf's Up" had its origin as the verses for Smile's "Child" especially as we have two verses essentially written on the lyric sheet (the 2nd was apparently partly forgotten and never used). This was my chance to do proof of concept and back up my theories with some backing vocals that really create a bed for that melody. I'd still bet a buck these started here.

UPDATE!! NOW WITH SLIGHTLY BETTER MIX!

https://www.youtube.com/watch?v=ZKABpRkjWMA (https://www.youtube.com/watch?v=ZKABpRkjWMA)

And here's a link to my version of "The Old Master Painter," featuring OMP lyrics!!

https://www.youtube.com/watch?v=sGI6EVJyfZY (https://www.youtube.com/watch?v=sGI6EVJyfZY)

Enjoyable take on Child. The melody doesn't really slide into the overall backing the way I think it would had Brian intended it that way.. at least not as a literal insertion. Maybe with a few changes to the melody?

I really liked the addition of the children playing in the background.

Reading this again, I'm wondering if VDP remains just as abashed about his own vocals as he always seems to have been. He's clearly not much of a singer (as he has noted frequently himself), but he has had some good moments (such as on his arrangement of "H&V") but my sense is that he would somehow be appalled to be an actual part of the vocal SMiLE, seeing that as a kind of desecration of what Brian was trying to accomplish in stretching the band's vocals as far as possible.

Given that interpretation, you could take his response with a grain of salt and read between the lines as to why he was not so keen on your choice of using him.

It's a tricky business getting in touch with one's idols, you know--so much randomness is possible in their response, and they have their own issues with things in ways that will probably always remain opaque to us...

Not sure if this has been posted here yet, but this is one of the coolest AI tracks I've heard (Old Man River):

https://youtu.be/IJjqmt7wrbE?si=Y1PRvl2MGnYr19KT

May I kindly suggest uploading them with the audio reversed, and/or with a download link in the description? Even without a link, people will be able to rip the audio file from the Youtube video and re-reverse it in even the most basic editing program. That's what I've done to keep my own SMiLE mixes up on YouTube. Id love to hear the rest of your work, both of you!

Dae Lims is crazy talented.  His "Walk On By/Are You There" medley is stunning.  And his Beatles mock-up "Return To Pepperland" is even better.