[HN Gopher] Generate Musical Accompaniment with R ___________________________________________________________________ Generate Musical Accompaniment with R Author : hackoo Score : 74 points Date : 2021-11-28 12:54 UTC (10 hours ago) (HTM) web link (flujoo.github.io) (TXT) w3m dump (flujoo.github.io) | CornCobs wrote: | Cool! Machine learning analysis of music seems to largely be at | the audio or midi level, and not really at the symbolic level. | What kind of learning techniques would be applicable to | structures like musical meter? | zeckalpha wrote: | MIDI is semi-symbolic. | TheOtherHobbes wrote: | MIDI happens at the note level, and the controller level for | parameter management. MIDI has no concept of chords or chord | languages. | CornCobs wrote: | It's still on the physical time scale and presents music as a | linear sequence. I believe symbolic music representation is | more of a graph than a list | pvaldes wrote: | system('vlc Lorde-Melodrama.avi') | sidedishes wrote: | I've been doing one-hour time limit song writing challenges and | in those I need to write accompaniment fast. I gotta say, this | generation is a really close reduction of what I actually do -- | come up with some accompaniment figure over a chord or two, then | transpose it and tweak chord tones for the next measure. Maybe | add a part or do something different if it's a section change. | IMO this post is a great example of how to mix variation and | repetition to balance interest and familiarity. | | Even if the generation itself has limitations it seems like a | great springboard for manual touches like dissonance tweaking, or | new sections, or fills / transitions -- arguably the cool | creative part after all anyway! There's a lot of repetition in | music, and when writing, it's hard to decouple it from points of | interest. A workflow that enables such is very exciting. | | (Taking this opportunity to share my multiplayer piano roll | editor: https://yuxshao.github.io/ptcollab/ ) | adamnemecek wrote: | I've been working on and IDE for music composition. Launching | soon https://ngrid.io. | billfruit wrote: | Will it have a tracker interface? | adamnemecek wrote: | No, an advanced piano roll. | bambax wrote: | Isn't this fundamentally describing an arpeggiator? | | Using Reaper I made a simple arpeggiator extension: it takes | input from one track for the chords and from another track for | the notes, and it maps the notes from the notes track to the | available notes from the chords track. It's like a "snap to | scale" but it snaps to the current chord. | | It's much better than a normal arpeggiator plugin IMHO because | you don't have to program patterns in the UI of a plugin; you can | do anything in the notes track and it will always work. | | The limitation, as with any arpeggiator, is that it can't play | notes outside of the current chord (but it's possible to have | "chords" with many notes on the chords track to allow more | freedom). | | One way that I use it is I write a chord progression, then I play | over it using the extension, and then I adjust the recording by | moving notes to more interesting positions. This song was made | this way: https://open.spotify.com/track/5TxVfIf9JUAhCEL3O5cWXT | (It's not perfect and the sound is kind of bad, but it's a | start). | setgree wrote: | This looks like a cool project and I look forward to trying it | later. | | General comment: one reason that the R package ecosystem often | feels creaky and brittle, I think, is that a lot of packages | (like this one) are pretty far afield from R's original intended | uses of statistics and visualization. Here, we see the use of | integers to denote pitches strung together as a vector to | generate a sequence. That's not precisely a misuse, but it's | definitely misaligned with the original vision, and trying to | build things on top of that misaligned infrastructure is a | potential vector of packages' breaking and general | irreproducibility. | adeelk93 wrote: | R's package ecosystem is creaky and brittle even for statistics | and visualization. I don't think "misaligned with the original | vision" is really relevant. Like, it has nothing to do with | someone making a package to work with music in R. | | Not sure how any of this is really relevant here. This is a | general R gripe. | malshe wrote: | > R's package ecosystem is creaky and brittle even for | statistics and visualization | | Explain what you mean by this. It's a big claim without any | supporting evidence. ___________________________________________________________________ (page generated 2021-11-28 23:00 UTC)