Jump to content

Aiwendil

Members
  • Posts

    171
  • Joined

  • Last visited

  • Days Won

    15

Aiwendil last won the day on January 22

Aiwendil had the most liked content!

7 Followers

About Aiwendil

Profile Information

  • Gender
    Male
  • Occupation
    Physicist
  • Interests
    Music, writing fiction, analytic philosophy, history
  • Favorite Composers
    Mozart and Beethoven
  • My Compositional Styles
    Classical, Baroque, Romantic, Early Music, Jazz, Progressive Rock
  • Notation Software/Sequencers
    Finale, NotePerformer, Reason 3.0

Recent Profile Visitors

3,692 profile views

Aiwendil's Achievements

Community Regular

Community Regular (8/15)

  • Concertophile Rare
  • 2025 Christmas Participant Rare
  • Very Popular Rare
  • Twelve Years in!!
  • Young Composers Nerd Rare

Recent Badges

135

Reputation

  1. I don't think we should be Draconian about policing this. I think the policy should be no posting music composed by AI, but I really think it would be a mistake to start policing and interrogating every post and assuming it's AI until proven otherwise. If there are indications that a piece might be AI (e.g. no score provided, telltale signs of Sonus, posted by someone with no prior history on the forum, etc.), then it makes sense to inquire further into it. But demanding proof of authorship from every composition would be a drastic overreaction to an issue that has, as far as I'm aware, only actually cropped up once here so far. I'm against AI compositions on the forum, but I'd rather waste my time once or twice giving useless feedback on an AI composition than chase real people who are posting real compositions off the site.
  2. Haha, yeah, I love Haydn's musical sense of humour.
  3. @Henry Ng Tsz Kiu Thank you so much for your thoughtful reviews of each movement! I'm glad you mentioned the timpani solo - I thought it would be amusing if, when the orchestra sets up the cadenza with its 64 chord, the timpanist suddenly decides that they're the soloist and starts playing a timpani "cadenza" before the trumpet gently nudges them aside and proceeds with the real cadenza. I could even imagine a bit of stagecraft where the trumpeter gives the timpanist a "what the hell?" look and maybe the conductor pretends to try to get their attention and make them stop. Anyway, I'm glad you liked the concerto. It's certainly one of the better pieces I've finished, in my opinion.
  4. Those are very kind words coming from such a skilled composer as you! I guess I was just in the proverbial zone when I wrote this. Usually I'm much slower, though that often has to do with me getting distracted or losing interest in a project only to return to it later. With this concerto, I managed to stay very focused and worked on it nearly every day.
  5. This is quite nice. I like Persichetti's book but I've never done the exercises in it; you're making me kind of want to give them a try, though!
  6. Thanks very much! The Hummel concerto (as well as the Haydn) was very much in my mind when I was writing this. It took me a little over a month to compose this; I think I started it in early November and finished mid-December. I have no contacts at all in the music world, so no, I'm afraid there's no way this would ever be performed.
  7. Thanks to @Churchcantor, @Henry Ng Tsz Kiu, and @bkho for listening! I'm quite gratified by your comments, because I really am quite happy with the piece.
  8. Thanks very much for listening, and for your comments! I try not to let adhering too strictly to a Classical style get in the way of experimenting with harmonic surprises like that episode in F# minor. As for the end of the cadenza, I don't feel that a concluding trill is a strict requirement for a cadenza, and I'm happy with the way I ended this one. Though really, the cadenzas I wrote are just intended as suggestions; in the very imaginary world where this would ever be played by a real person, I think the performer should be free to make the cadenza their own. Looking forward to reading your reviews of the other two movements!
  9. This sounds like an interesting and inventive subversion of a waltz. I'm curious why you describe it as two movements, though. I guess you're considering a new movement to start at m. 106, but it's in the same tempo and style. Also, you probably know this, but the stuff in the right hand at m. 158 and forward would normally be written as tremolos. I assume you wrote out the notes so that the notation software played it back correctly, but in such cases I'll usually prepare separate versions of the score for playback and for display.
  10. Very nice piece, with the cross relations between E and E flat. And very nice performance by Henry, as usual. Well done to both of you!
  11. Here's the thing, though: "AI", by itself, is pretty much meaningless. For decades, people have been using "AI" to refer to the programming controlling NPCs in video games, for instance. Sometimes, "AI" is used in a sense that covers any machine learning or neural network application. More recently, "AI" has caught on in the context of Large Language Models like ChatGPT that take a prompt as input and then use some minimization function over their corpus of training data to predict what the most likely response to that prompt would be. The broader class of software including LLMs and similar models that output images, sounds, etc., are called "generative AI". And because "AI" is such a hot topic now, anyone whose software product uses anything resembling a neural network is going to be sure to advertise it as "AI". In my opinion, it would be better to use more precise language and, when we mean LLMs, say "LLMs". But I suppose that ship has sailed. In any event, clearly what is under discussion here is whether music composed by a generative, LLM-like software should be permissible here - and as I've said, I don't think it should be, since this is a composition forum and there's no sense in posting a piece of music you didn't compose. But by the same token, if you composed the music, then of course you should be permitted to post it, regardless of whether you used tools that are advertised as "AI" to produce the audio mock-up of the music you composed.
  12. I really enjoyed this! My only complaint is that in a few places, the mixing sounded off - the flute almost inaudible when playing pianissimo. But that's a minor nitpick. It's an excellent composition.
  13. Exactly. Completely agreed. It wouldn't be appropriate to post, say, a Haydn string quartet in the "share your work" forums, because it's not your work. In exactly the same sense, it doesn't make any sense to me for someone to post an AI-composed piece here, because it's not the poster's work. What would be the point? Best case, you fool people into thinking it's your own work. What then? People give you feedback, assuming you wrote the piece, and that feedback is totally useless to you since you didn't, in fact, write the piece. I suppose maybe somebody will say, "Good job," to you, and OK, congratulations, you tricked someone into praising you. (Though, I must say, to date, I still have not heard an AI-composed piece of music that I thought was anything better than mediocre). This is exactly what I can't understand. What would anyone expect in response to a piece of AI-generated music? Sure, I could pretend that a human actually wrote it and write a critique of it based on that fiction, but I can't imagine why I would ever do that, nor can I imagine what use that would be to the person who posted it.
  14. I'm curious how quaternions were used to generate this. I assume by a, b, c, and d, you mean the coefficients in a + bi + cj + dk, but I don't understand how thinking of this as a quaternion, as opposed to just a set of four values, is relevant to the composition process.
  15. My feeling is that, regardless of your opinion on AI music, this is a composition forum. Prompting an LLM to generate music is not composition. Therefore, it has no place on this forum. It's as simple as that for me.
×
×
  • Create New...