Hey, a new sub-board!

[theboxinargentina]

Has anyone come across attempts to use this to capture the timbre of instruments? This technology may only be tuned towards voice capture, but if not ....

BBC recordings could be low hanging fruit for a test - the same performers, on the same equipment, recorded by the same people all around the same time. Some of the poorer BBC recordings could possibly be brought back to life with enough decent stems for training from the well recorded performances.

I have a training environment set up that I could use to have a go if anyone could steer me towards some good stems. I'd need about five minutes of, say well recorded rhythm guitar, and a stem of a poor quality one and I could give it a test if anyone knows of a separated BBC release.

This is a great idea and could be the beginnings of what I imagined as "upscaling" sound, for want of a better term.
I appreciate your efforts and I'd join in but I haven't mastered the modeling end of this tech.
I'll try to get some well recorded instruments when I get the chance.
I think you're really on to something with this!

[theboxinargentina]

I don't know what you mean by "the real thing". You don't seem to have a problem with fan made remixes - those are surely not "the real thing", so I don't see the harm in taking it a step further. Personally, I am loving the current spate of "AI Paul" songs. Sure, most are not very good but there are some genuine delights, and I am deriving a lot of pleasure in being able to hear them.

I'm having flashbacks to decades ago when certain others were telling me, with a religious fervor, that the Beatles SHOULD NOT EVER be remixed.
I'm going to try not to get into it this time with the "yelling at clouds" people.

[cliftdean74]

If this AI fad gets to be the real thing here, you can count me out. I'll stop re-upping stuff and will will delete all my "permanent" links.

We need the real thing, not these so-called wonders.

1,000,000% Agreed.

[ToddS]

If it doesn't sound good , don't post it . If it doesn't sound exactly like Paul , or John or George or Ringo , Please do not post it . OR = Call them Mashups so it alerts the person who wants to download it . Thank You for letting me express my opinion .

[bsizzy_27]

If this AI fad gets to be the real thing here, you can count me out. I'll stop re-upping stuff and will will delete all my "permanent" links.

We need the real thing, not these so-called wonders.

1,000,000% Agreed.

There's some decent AI stuff but overall, I'm with you guys.

[BeatleG]

Major edit: I made a mistake a posted something I wrote earlier. Sorry make a dumb mistake like that. This is what I meant to post this time and I deleted what I posted a little while ago.

The last time I commented on AI music I said the genie was out of the bottle but hoped for respect of artists and not watering down their works by flooding the market with creations of wish lists and the fun of ‘composing’ and mimicking and pretending that John should sing Paul or Ringo should sing Adele or whatever. I decided to shun all the AI Beatles and other artists. I could stay pure and for the first time live up to my ideals. Then I listened to some samples of the stampede to come. You know, for research. But now I see that I can’t even avoid it or even resist it. I see coming a licensed Beatles channel with everlasting new Beatles material. We never have to move on and keep Beatles collecting a part of who we are, a wonderful hobby. Now on our way to work and home again and beyond we can listen the rest of our lives to our favorite group. We don’t need remixes and multitude of takes anymore. We’ll just have new material by the Beatles whenever we want it. Can we count on the AI to make them progress like the great progression we (those of us that were alive then) from 1963 to 1970? Wow, maybe each year that goes by they will progress like that. Imagine 10 years away. That would be mind blowing. The Beatles still progressing when they’re 110 years old. No need to explore new or a wide variety of artists. Just the Beatles. Or maybe the guys will let EC join them and they can be a fivesome. Why stop there? Elvis can join the group, too. In a way, all that maybe would be OK except for one thing. We’ve stopped exploring, taking in new voices and people with different views of the world.

To tell you the truth, I kind of stopped exploring a long time ago so I’m part of that problem. I have my Beatles and Moody Blues, Neil Yong, Van Morrison and Laurie Anderson and a few others. I couldn’t seem to take in new stuff. I’ve tried, giving some attention to SNL performers, for example. Through the years the acts did seem really good and professional and even if it wasn’t my cup of tea I generally made it through the songs w/o skipping them after a minute and realizing it was just another group or artist changing how they attack a song to sound different and skipping along to Weekend Update. Sure, they usually were competent artists, but it wasn’t like the 60’s and 70’s when groups just didn’t try to sound different, they just were different and creative. If I like a SNL musical guest I find some of their larger body of work to listen to but am generally disappointed, it sounds too derivative or worse. Some of that is on me. But, unless I’m totally wrong, I’m not seeing stellar artists.

Anyway, I’ll make my last point. Since we’re off to the races with the music AI, let’s try to not forget to live some of the time in the present and keep growing as individuals, which in part means opening ourselves to taking in new information and letting our souls enjoy new music from real people, living just part of our lives in the past.

[urasam2]

Artists may suffer monetary loss as people buy new Beatles instead of original releases.

Who's going to be buying this stuff? It's being shared for free.

[Ants]

As long as an AI-generated performance is not passed as genuine (like an outtake, a new discovered song, etc.), I'm fine with that.
I see these AI performances as some sort of "imitator": as long as I know it's not the real thing, it's alright for me.

A bit like the famous "Have You Heard The Word?".

I'd really like to hear some vocals "swaps", like Lennon singing Maxwell's Silver Hammer or Yesterday.
I'd find these very funny.
I wouldn't listen to these creations more than two or three times, though.

Also, I would never use this technology to improve a deteriorated tape and/or reconstruct a bad recording.
One thing is extracting something that already exists to boost/improve it, another is creating sounds that originally didn't exist.

[Lord Reith]

If the new experiments are confined to this sub-board then I don't see what the problem is. Part of the history of BZ and this place was that people could upload their home brewed restorations and remixes and get opinions from other people. It is a very helpful thing to have feedback. When I first started I uploaded some truly awful stuff. I can remember how tentative I was and how appreciative that people just didn't tell me to bugger off. Their feedback helped me evolve and get better.

Because this tech is brand new, the only way to learn how to use it is for those who are interested to share freely their ideas and experiments whether they be good, bad, crazy or brilliant. At the moment this vocal synthesising may sound primitive but it has every prospect of becoming a brilliant creative and restorative tool. We have a duty to find useful ways to use this, to counter the douchebags who as we speak are devising evil uses for it. When I first heard of this tech, I was apalled because I immediately thought of all the wicked uses it would be put to. Now I see that it has the potential to be something positive, and it is people like us who will find those uses.

Sure there'll be a honeymoon period while everyone jumps on this new toy and puts it to every use they can think of, and youtube will quickly fill up with AI mashups (and deleted accounts!) It'll get pretty silly for sure, but so long as we know what we're listening to then we can always avoid it if it doesn't appeal. I don't think anyone will deliberately try to deceive people with a new "genuine" lost mix because if it's good enough to fool then they are also going to want credit. I suppose the bootleggers might jump on some of these mixes and tout them as being genuine but, really, who gives a rat's ass what they do. They've been mislabelling things for years.

[Lord Reith]

I think of the attempts that swap singers, or rejuvenate Paul's voice more along the lines of outfakes than anything else - interesting and perhaps enlightening, but not revelatory.

The use of this technology as a production tool for audio repair is perhaps another matter.

Has anyone come across attempts to use this to capture the timbre of instruments? This technology may only be tuned towards voice capture, but if not ....

BBC recordings could be low hanging fruit for a test - the same performers, on the same equipment, recorded by the same people all around the same time. Some of the poorer BBC recordings could possibly be brought back to life with enough decent stems for training from the well recorded performances.

I have a training environment set up that I could use to have a go if anyone could steer me towards some good stems. I'd need about five minutes of, say well recorded rhythm guitar, and a stem of a poor quality one and I could give it a test if anyone knows of a separated BBC release.

That's what I've been thinking for a long time, but it was a fantasy till now. I think sovits has opened a door for this kind of thing.