What's new

AI Orchestral Music

was answering about the mention and merits of "In C" in the conversation
In that case, Mozart had already done the thing with assembling pieces from fragments of melody and harmony based on chance. The Würfelspiele was intended for a different context but the core idea was there, so it's hard to say Riley was entirely original except in the application.

This is the problem when determining inclusion or exclusion using absolute terms like "original", especially in the context of art and music where originality is really more a matter of adding a twist to what has gone before.
 
You objected to sampled loops. I mentioned musique concrète, which uses tapes/samples. So, I'm not sure which part of that you don't understand.
No I didn't object to anything. I said that using a series of previously constructed loops to make a track is arranging, not composing. Musique concrète is something entirely different, because nobody has composed anything for the building blocks.
 
I wouldn't hold my breath seeing such technology with virtual instruments. As of now developing and training ai models need a lot of resources and A LOT of money. It's not financially worth it to apply it to a niche sector like composing music (and sample libraries are even a niche of the niche). Also, I think it's great there is still a gap between sample libraries and the real thing. Sample libraries are not supposed to replace the real thing (although they obviously do it in many cases). Working towards things is great. Dreams are great. So are unreachable goals. I think for many composers it's a dream to record their music with a real orchestra. Why not work towards that and appreciate and cherish the moment when it might happen? In my opinion it would be best this technology gets buried and destroyed for good. And while we're at it, we could bury streaming sites and social media right with it.
 
At the current state you cannot prompt Udio or the other current AI music s(h)ites for specific things like that.
Fascinating; I've never tried them (bit of a purist myself) but I would've guessed that you could at least reference instruments, styles, tempo, other soundtracks, genres, etc. Sounds like none of this is an option?... If so then seems to be somewhat less useful than even free stock music.
 
No I didn't object to anything. I said that using a series of previously constructed loops to make a track is arranging, not composing. Musique concrète is something entirely different, because nobody has composed anything for the building blocks.
Basically, you made an assumption about what mikrocosmiko meant by "loops" and just riffed from there. Your assumption seems to be quite different to what many people do when employing loops or samples in their compositions, those with contracts at least. It has led to some weird anomalies, such as the royalty situation with Bittersweet Symphony vs anything involving the Amen or Funky Drummer loops. But the idea that Richard Ashcroft just "arranged" the resulting song from Loog Oldham's string recording is patently absurd. With Aphex Twin's Xtal, the link to the original is much stronger and could be regarded as more a rearrangement, but even then it's not as if anyone went out in search of vinyl of the original library track shouting "shut up and take my money, I must have this original version". I don't doubt Carewe and Greig should have received credit for the sample but Xtal winds up in quite a different place. Somewhere in the middle, you wind up with Meat Beat Manifesto or FSOL's early stuff.

Also in your personal definition of musique concrète, you ignored this:

That is made from precomposed works.
 
Last edited:
Are you old enough to remember "Paint By Numbers?"

That's basically the 'music-loop' world.

Assembling loops is not 'musical composition'. It's merely 'cut-and-paste' similar to a graphic artist moving images around in order to create the outcome they desire.

The majority of Graphic Artists are not 'Artists'....just manipulators of others work.

Same with musicians that use loops all the time. They're not composer...just manipulators of music composed by other people.
 
It's merely 'cut-and-paste' similar to a graphic artist moving images around in order to create the outcome they desire.

The majority of Graphic Artists are not 'Artists'....just manipulators of others work.
Who are all these graphic artists? It shouldn’t be an issue to identify them being as it’s so common (apparently).

Are you sure you’re not confusing them with layout artists/designers?
 
Are you old enough to remember "Paint By Numbers?"

That's basically the 'music-loop' world.

Assembling loops is not 'musical composition'. It's merely 'cut-and-paste' similar to a graphic artist moving images around in order to create the outcome they desire.

The majority of Graphic Artists are not 'Artists'....just manipulators of others work.

Same with musicians that use loops all the time. They're not composer...just manipulators of music composed by other people.
I like your analogy there. Although I do hesitate to cast a sweeping label over all composers that utilize loops, it seems to me that it detracts from the creative value of music to consistently make use of prewritten snippets of music rather than composing something new. I mean, if AI got a splice subscription then it could conceivably make the exact same music as a composer using splice. Not to throw loop-using composers under the bus, but it does feel a bit like paint-by-numbers to me. That's not to say that there isn't talent exhibited by these people, but I don't have any wish to use loops in my work.
 
it be nice if it generated midi and/or stems.

Also, just a more deep thoughts, but even thought the music might sound great, my feeling is that the majority of ai generated music will be used for marketing. Sort of like midjouney does with images. Mostly for the social media side. These guys dealing with social media content need a lot of content and these tools help them the most.
Because doing this for music library or Spotify seems just like a waste of time imo. Theres been ways to make music with construction kits, loop based composing and there still no one claiming they are making bank with those tracks. The ones being affected the most, imo are the royalty free crowd and royalty free libraries. Im sure those royalty free sites are starting to get filled up with ai music.
I might even consider submitting tons of ai music to these sites but ill make ai music thats "passable" at best. Generic stuff with tons of random metadata so those trying to find music for their royalty free projects just get tired of it and decide to just find a real artist ;)
 
Sorry but how is this not composing?

Well, I think there be a bit of unintentional equivocation here. If you ever happen to study music theory, then you'll find composing is broken down to the basics of SATB. That right there my friend is sampling, as the title of the video so kindly pointed out. I think that the mastering and mixing of the loops could fall under the producing category, but there's a difference between production and composition. The original composer wrote a melody, harmonized the melody, added a bass line, and set it to a drum track. The samplist just ripped pieces out of it and stuck them together differently.
 
Lordy, it's like Life of Brian.

When the thread got underway, I thought the enemy was AI not human artists you don't like (including graphic artists because....reasons).

Splitters!
 
This is just early days for AI. This is the TRS-80 of AI. What will things be like in 5 years? 10 years? It’s accelerating rapidly.

I sympathize with those just starting out with a music/composing career, because it’s hard to imagine how that career will pan out after just a few years of this tech developing, let alone maturing.
 
The majority of Graphic Artists are not 'Artists'....just manipulators of others work.

Same with musicians that use loops all the time. They're not composer...just manipulators of music composed by other people.
It’s not the notes, it’s the order in which they put the notes together - that makes the composer. 😃
 
Well, I think there be a bit of unintentional equivocation here. If you ever happen to study music theory, then you'll find composing is broken down to the basics of SATB. That right there my friend is sampling, as the title of the video so kindly pointed out. I think that the mastering and mixing of the loops could fall under the producing category, but there's a difference between production and composition. The original composer wrote a melody, harmonized the melody, added a bass line, and set it to a drum track. The samplist just ripped pieces out of it and stuck them together differently.
You forgot that they put a new melody over the… em… reworked samples. I understand what you’re saying but I think that taking composing down to the basic 4part harmony is not adequate in the 21st century. For me, that “ripped pieces and stuck them together differently” is producing a totally different work. I mean, it seems to be a lot more creative that putting down, say, a i-VI-III-VII progression at the piano and improvising a simple melody over it, what I think we all would agree that is definitely composing

I hope not to sound angry, this conversation I find very interesting ☺️
 
You forgot that they put a new melody over the… em… reworked samples. I understand what you’re saying but I think that taking composing down to the basic 4part harmony is not adequate in the 21st century. For me, that “ripped pieces and stuck them together differently” is producing a totally different work. I mean, it seems to be a lot more creative that putting down, say, a i-VI-III-VII progression at the piano and improvising a simple melody over it, what I think we all would agree that is definitely composing

I hope not to sound angry, this conversation I find very interesting ☺️
Once they finished with the samples and started the actual production process it was great. And I hate SATB composition with a passion, but there is a lot of theory behind music theory. (Crazy huh?!) I'm not detracting from the creative process involved in sampling; but I don't think it's as difficult or as complex as composing, and there is no way they are the same thing.

That being said I feel like I'm contributing to the derailment of this thread right now. Down with AI!!!
 
This may prove to be one of the Achilles heels of AI in the short term. When it becomes too hard to knock the results into shape and there's no easy to way to apply small manual edits without continually regenerating the material through prompts (which call on the user to second guess what the AI will produce), more experienced people will use it for a bit and abandon. Even Rick Rubin types who may find the interface a reasonable one and cheaper than interacting with human musicians may just find it too onerous to deal with.

There will be a "that's close enough" crowd and for underscore type work maybe it's all a lot of clients will pay for.

But the way these things are implemented are very characteristic of computer scientists bolting stuff together rather than finding out how a practical FenbyAI should work. I know of one or two startups who have consulted with composers and musicians but they are very much in the minority and I suspect they will run out of cash quite quickly, if not already.

You don't seem to understand.

"Media music" has various parts to it.

1. Stock library music used in trailers, promos, ads, documentaries, reality shows, etc etc
2. Stock library music used as underscore for films, usually lower budget.
3. Music directly written to picture.

AI can be used in various ways...

1. Generating the ideas and used on a spectrum from very little to 100%.
2. Same as above, but using the actual audio recording, chopping it up, using the stems, remixed and written over the top remixed.
I wouldn't hold my breath seeing such technology with virtual instruments. As of now developing and training ai models need a lot of resources and A LOT of money. It's not financially worth it to apply it to a niche sector like composing music (and sample libraries are even a niche of the niche). Also, I think it's great there is still a gap between sample libraries and the real thing. Sample libraries are not supposed to replace the real thing (although they obviously do it in many cases). Working towards things is great. Dreams are great. So are unreachable goals. I think for many composers it's a dream to record their music with a real orchestra. Why not work towards that and appreciate and cherish the moment when it might happen? In my opinion it would be best this technology gets buried and destroyed for good. And while we're at it, we could bury streaming sites and social media right with it.
You're not paying attention.

There's already many being worked on.

Google Deepmind's Lyria is the most exciting, since it's most Udio-like but a clsoed limited beta on youtube



I am not sure what you think needs to happen for an AI virtual instrument. You'll just be able to sing in your part, or use your current samples, and it will generate it however you like.

But there's a Ai drum/sound plugin that is out right now where you can generate infinite drum samples, and use reference audio etc. i forget the name in this moment.
 
You don't seem to understand.

"Media music" has various parts to it.
Thanks for the detailed tutorial on egg-sucking.

And missing the point.

However:

1. Generating the ideas and used on a spectrum from very little to 100%.

If you meant prototyping a concept, I’d agree. From the way it currently works we already know that using it to generate ideas is not a great idea. Check the recent ruling from the US Copyright office for an example (from graphic novels in that case).
 
Top Bottom