Sure Jeongbeom, it's actually helpful for me to articulate these ideas, so ask away.
I wrote an algorithm that creates a kind of 'field' or 'space' of music that one essentially moves around at varying speeds, with different patterns of rhythms or chords in different areas. I've been exploring it for a few months, and I am currently recor…
Sure Jeongbeom, it's actually helpful for me to articulate these ideas, so ask away.
I wrote an algorithm that creates a kind of 'field' or 'space' of music that one essentially moves around at varying speeds, with different patterns of rhythms or chords in different areas. I've been exploring it for a few months, and I am currently recording some of these 'journeys' through the space. This isn't an academic exercise, I really feel an emotional response to the space and its sounds, and although it's really hard to be objective about one's own music, a couple of people I sent it to gave very positive responses.
Ideally, listeners would have the program and then a few set journeys built in, which vary each time they are played. But the logistics of this is a bit complicated for a one-person project, so that might have to wait. For now, I will share the recordings, but I see this as an intermediary step to a more ideal situation of people having the program, and it generating the music on whatever device they are on. That way, each time the 'performance' would be unique, although there would be a similarity too, so that it's not completely random, which wouldn't sound good.
So basically two things have been created - an algorithm, and a journey through that algorithm; combined they make the artistic creation.
I'm soon going to share the audio and a post explaining in more detail the process, that I have written and re-written for about a month! I hope what I've said makes sense. I'll try to get it out there as soon as I can...
These ideas are definitely being explored in crypto circles. I find some of that discussion really interesting. The issue I have is with crypto, which appears to be a Ponzi scheme with terrible environmental consequences.
« That way, each time the 'performance' would be unique, although there would be a similarity too, so that it's not completely random, which wouldn't sound good. »
I thought of something similar for my latest composition. Several variations have been composed for each movement of the original piece. By combining these variations, unique versions of the work are achieved.
Everything is recorded and all variations are entirely written, fixed. So it is a lot of work (and maybe less exciting conceptually).
The algorithms I use are very simple. I don't program them. I edit the music accordingly, and I can decide to let go of the rule whenever I want, for a single note or an entire section if I believe that it will serve the music better.
« Ideally, listeners would have the program and then a few set journeys built in, which vary each time they are played. »
At one point I thought of distributing the album as a music player with the possibility to bookmark a combination of variations when the listener likes it. It was not feasible for me, so I went on publishing physical "1/1's", one for each combination of variations, on CDr or cassette.
Fascinating Petar. I had a listen to your music over the weekend and it sounds excellent and original. I will listen again and try to buy your new album in the new year.
Is the latest composition you refer to recorded yet? Do you have something to share? It sounds like a really interesting process and I would like to hear more. Scores, audio, anything.
It's interesting that you mention the feasibility. That is exactly what has held me back. If programming becomes easier over the years, as I expect it will, it becomes more doable.
The problem is that audiences are (understandably) habitatued by a century of listening to recordings. Creating a program potentially represents lots of work that might very well be ignored, and yet, it's one of the most interesting things that artists could do right now.
Based on your music it sounds like you program. Are you using Supercollider. I really like the warmth of your synthesis. Great work! If you have the time to post on Substack it would be great to know more about what you are doing.
The latest composition I refer to is "Sept cordes". I can send you a link with variations if you want but I couldn't find a way to send a private message on Substack.
The score is currently a mess. I will finish it in the next few months and would be happy to send you a copy.
"Creating a program potentially represents lots of work that might very well be ignored, and yet, it's one of the most interesting things that artists could do right now." : Absolutely.
I don't remember ever using Supercollider. I've worked with OpenMusic for decades, but it eventually evolved into more of a "symbolic music writing sequencer" that I use less and less since it obstructs the general flow, especially when exceptions to the rules are needed as mentioned above.
"I really like the warmth of your synthesis." Hmm... I'm not sure what you are referring to. Could you please be more specific ?
Hi HL, I'm afraid only Bach can create Bach. I wasn't sure if your first comment was serious or sassy - such is the peril of the internet. Taken seriously... my aim is not to create music from algorithms but to create music using algorithms; a subtly but significant distinction. Bach used algorithms - harmonic, scalic, and rhythmic - to create his music. The 48 Preludes and Fugues is a highly conceptual work, both at a high level and low level. It combined cutting-edge ideas with humanistic expressivity. No algorithm can, or ever will, replicate Bach's genius.
What I am doing is nothing new. Musicians have used algorithms combined with the latest technology for time immemorial. The only difference is I am using a computer, but it can't replicate Bingen, Bach, or Coltrane, and their unique processes. Their legacies - as far as I'm concerned - should be left well alone. Replication is dull, but innovation exciting.
Back in the 1980s, trying to generate music like Bach's was a big thing in the AI community. A friend of mine did her undergraduate thesis on writing Bach-like fugues. I think it peaked with Hofstadter's book, but the idea still pops up now and then.
I've not read the book. Thanks for the recommendation.
I knew a kid growing up (he was about 14) and he could write baroque music that was indistinguishable from Bach. It was incredible. But when it came to doing his own creative and original work it was unremarkable. He was almost like a musical forger.
I expect something approximating this might be possible with AI, given my friend's approach seemed entirely analytical, however Bach was doing things of an artistic and conceptual nature.
I'll warn you that Godel, Escher and Bach is written in the style of Lewis Carroll, so it can be charming or tedious as you will. Having taken a course in recursive function theory, I'll also point out that the book is a much less formal but otherwise excellent introduction.
A lot of the fascination with writing music like Bach's was about trying to understand what Bach did and how he might have done it. It was like watching a pro-basketball game and then trying to toss the ball through the garage hoop while running. Failure made another's success even more impressive.
Thank you Dom, I’m not really sure if I was being serious or sassy. I was just taken by the idea. I do take the distinction to create music using algorithms.
I’m not a musician but I do love music. My husband made music both technically (I still have his 30-year-old Amiga computer, Sequential Circuits Prophet 500(?), and other things that would be jacked in) and old-school, violin, guitar, and piano. I was the audience.
While I have always loved Bach and many others it wasn’t until Howie that I learned so many things. He could easily fill in all I would need to know. He also talked a lot about how Bach made his music. I think you two would have gotten on well.
And while it may not be new it does sound interesting. Innovation is good.
I do worry a little bit about what may be coming available to us all of a sudden. Legacies seem to be sold off to the highest bidder to do what they will.
Hi HL, I'm glad the idea resonated with you. By the sounds of your message it sounds like your lost your husband - I´m sorry to hear that. He sounds like a fascinating guy. For a lot of musicians Bach is kind of our Shakespeare figure. Immortal, mysterious, and sublime. I think the important thing for musicians is to respect our ancestors and what they did to carry this music to us. In that regard I think we should protect their legacies too. It's not just data to be played around with, there's something precious there, even sacred. Thanks for engaging.
Sure Jeongbeom, it's actually helpful for me to articulate these ideas, so ask away.
I wrote an algorithm that creates a kind of 'field' or 'space' of music that one essentially moves around at varying speeds, with different patterns of rhythms or chords in different areas. I've been exploring it for a few months, and I am currently recording some of these 'journeys' through the space. This isn't an academic exercise, I really feel an emotional response to the space and its sounds, and although it's really hard to be objective about one's own music, a couple of people I sent it to gave very positive responses.
Ideally, listeners would have the program and then a few set journeys built in, which vary each time they are played. But the logistics of this is a bit complicated for a one-person project, so that might have to wait. For now, I will share the recordings, but I see this as an intermediary step to a more ideal situation of people having the program, and it generating the music on whatever device they are on. That way, each time the 'performance' would be unique, although there would be a similarity too, so that it's not completely random, which wouldn't sound good.
So basically two things have been created - an algorithm, and a journey through that algorithm; combined they make the artistic creation.
I'm soon going to share the audio and a post explaining in more detail the process, that I have written and re-written for about a month! I hope what I've said makes sense. I'll try to get it out there as soon as I can...
Async.art has a concept like this being done.
These ideas are definitely being explored in crypto circles. I find some of that discussion really interesting. The issue I have is with crypto, which appears to be a Ponzi scheme with terrible environmental consequences.
« That way, each time the 'performance' would be unique, although there would be a similarity too, so that it's not completely random, which wouldn't sound good. »
I thought of something similar for my latest composition. Several variations have been composed for each movement of the original piece. By combining these variations, unique versions of the work are achieved.
Everything is recorded and all variations are entirely written, fixed. So it is a lot of work (and maybe less exciting conceptually).
The algorithms I use are very simple. I don't program them. I edit the music accordingly, and I can decide to let go of the rule whenever I want, for a single note or an entire section if I believe that it will serve the music better.
« Ideally, listeners would have the program and then a few set journeys built in, which vary each time they are played. »
At one point I thought of distributing the album as a music player with the possibility to bookmark a combination of variations when the listener likes it. It was not feasible for me, so I went on publishing physical "1/1's", one for each combination of variations, on CDr or cassette.
Fascinating Petar. I had a listen to your music over the weekend and it sounds excellent and original. I will listen again and try to buy your new album in the new year.
Is the latest composition you refer to recorded yet? Do you have something to share? It sounds like a really interesting process and I would like to hear more. Scores, audio, anything.
It's interesting that you mention the feasibility. That is exactly what has held me back. If programming becomes easier over the years, as I expect it will, it becomes more doable.
The problem is that audiences are (understandably) habitatued by a century of listening to recordings. Creating a program potentially represents lots of work that might very well be ignored, and yet, it's one of the most interesting things that artists could do right now.
Based on your music it sounds like you program. Are you using Supercollider. I really like the warmth of your synthesis. Great work! If you have the time to post on Substack it would be great to know more about what you are doing.
Thank you for listening Dom.
The latest composition I refer to is "Sept cordes". I can send you a link with variations if you want but I couldn't find a way to send a private message on Substack.
The score is currently a mess. I will finish it in the next few months and would be happy to send you a copy.
"Creating a program potentially represents lots of work that might very well be ignored, and yet, it's one of the most interesting things that artists could do right now." : Absolutely.
I don't remember ever using Supercollider. I've worked with OpenMusic for decades, but it eventually evolved into more of a "symbolic music writing sequencer" that I use less and less since it obstructs the general flow, especially when exceptions to the rules are needed as mentioned above.
"I really like the warmth of your synthesis." Hmm... I'm not sure what you are referring to. Could you please be more specific ?
So that’s a yes to my query?
I look forward to what you come up with.
Hi HL, I'm afraid only Bach can create Bach. I wasn't sure if your first comment was serious or sassy - such is the peril of the internet. Taken seriously... my aim is not to create music from algorithms but to create music using algorithms; a subtly but significant distinction. Bach used algorithms - harmonic, scalic, and rhythmic - to create his music. The 48 Preludes and Fugues is a highly conceptual work, both at a high level and low level. It combined cutting-edge ideas with humanistic expressivity. No algorithm can, or ever will, replicate Bach's genius.
What I am doing is nothing new. Musicians have used algorithms combined with the latest technology for time immemorial. The only difference is I am using a computer, but it can't replicate Bingen, Bach, or Coltrane, and their unique processes. Their legacies - as far as I'm concerned - should be left well alone. Replication is dull, but innovation exciting.
Back in the 1980s, trying to generate music like Bach's was a big thing in the AI community. A friend of mine did her undergraduate thesis on writing Bach-like fugues. I think it peaked with Hofstadter's book, but the idea still pops up now and then.
I've not read the book. Thanks for the recommendation.
I knew a kid growing up (he was about 14) and he could write baroque music that was indistinguishable from Bach. It was incredible. But when it came to doing his own creative and original work it was unremarkable. He was almost like a musical forger.
I expect something approximating this might be possible with AI, given my friend's approach seemed entirely analytical, however Bach was doing things of an artistic and conceptual nature.
I'll warn you that Godel, Escher and Bach is written in the style of Lewis Carroll, so it can be charming or tedious as you will. Having taken a course in recursive function theory, I'll also point out that the book is a much less formal but otherwise excellent introduction.
A lot of the fascination with writing music like Bach's was about trying to understand what Bach did and how he might have done it. It was like watching a pro-basketball game and then trying to toss the ball through the garage hoop while running. Failure made another's success even more impressive.
Thank you Dom, I’m not really sure if I was being serious or sassy. I was just taken by the idea. I do take the distinction to create music using algorithms.
I’m not a musician but I do love music. My husband made music both technically (I still have his 30-year-old Amiga computer, Sequential Circuits Prophet 500(?), and other things that would be jacked in) and old-school, violin, guitar, and piano. I was the audience.
While I have always loved Bach and many others it wasn’t until Howie that I learned so many things. He could easily fill in all I would need to know. He also talked a lot about how Bach made his music. I think you two would have gotten on well.
And while it may not be new it does sound interesting. Innovation is good.
I do worry a little bit about what may be coming available to us all of a sudden. Legacies seem to be sold off to the highest bidder to do what they will.
Hi HL, I'm glad the idea resonated with you. By the sounds of your message it sounds like your lost your husband - I´m sorry to hear that. He sounds like a fascinating guy. For a lot of musicians Bach is kind of our Shakespeare figure. Immortal, mysterious, and sublime. I think the important thing for musicians is to respect our ancestors and what they did to carry this music to us. In that regard I think we should protect their legacies too. It's not just data to be played around with, there's something precious there, even sacred. Thanks for engaging.
You could, eventually, give me Bach, a couple of violins, and concertos we have never heard before but still recognizable?