sporkl

I've been working on something similar, implementing a relational language on top of a tensor library[0].

Mathematically, einsum and database joins are the same thing, just over different semirings (real numbers for einsum, booleans for databases). A lot of papers about datalog explore this sort of thing in more depth. In particular, Dyna[1] might be interesting.

[0]: https://arxiv.org/abs/2509.22614 [1]: https://dyna.org/

Would be curious to know how algebraic music theory is being applied here. Overall this seems pretty basic; no real voice leading, or even chords with more than 2 notes in them. Algebraic music theory runs pretty deep and I'm not seeing much of it applied here. I honestly can't imagine paying $7 for a two-voice, repetitive 40s long midi file.

Seems potentially LLM'd, account is new and I couldn't find any other sign of existence of the creator. They've also submitted this 4 times in the past month.

I think that's what Soloist[1] is trying to do (unaffiliated, but I've met the founder)

[1]: https://www.soloistapp.com/

Sounds kind of like Arcan https://arcan-fe.com/about/

The framework used in the book, malt[0], is currently not GPU-accelerated, but it's being worked on.

Maybe interesting, I used it for a toy implementation of the GPT architecture[1] in about 500 lines.

(I studied with one of the authors, Dr. Daniel Friedman; wasn't super involved here but proofread a late draft and TA'd for a course based off the book.)

[0]: https://github.com/themetaschemer/malt

[1]: https://github.com/sporkl/malt-transformer

Lilypond is the only music engraving system I'm aware of that can handle polytempo scores. The TEX-ness really comes in handy.

I'm personally a fan of Stylus Labs Write: https://www.styluslabs.com/

Not exactly infinite canvas, but pages can grow outward. Cross-platform and open source! And has some cool features which make working with handwritten text nice.

Not this project, I've seen this something like this done with Google's MediaPipe models to get face tracking data into Max/MSP.

https://cycling74.com/forums/n4m-facemesh-handpose-google-me...

Racket! This is the language that the common author (Dr. Daniel Friedman) uses, and many of the Little books use custom DSLs implemented in Racket.

Racket now runs on Chez under the hood (inheriting the performance), and has a pretty decent ecosystem as far as schemes go.

(I TA for Dr. Friedman’s programming languages course)

Reminds me of the method used for sonifying periodic table elements here[0], but in reverse.

[0]: https://youtu.be/Z9dpHWrgzMw

Anecdotally, I feel like OCaml is growing in popularity, probably due to ecosystem improvements. Stuff like dune and other OCaml Platform tools becoming mature, multicore support, recently first-class Windows support, etc.

yeah, in general I do want to open source it, but for the time being I’m a student and every little bit counts

Overall a good intro on the subject. I feel like at the start it might benefit from being a bit more explicit about pitches being frequencies and ratios being intervals; probably just worth a reminder for people who aren’t as familiar

Might be interesting to talk about how the usual ratios come from the harmonic series. For sounds that don’t produce a harmonic series, other potentially non-integer ratios can actually sound more consonant. The youtube channel New Tonality[0] has a bunch of great videos about this

Also wanted to mention that I’ve been working on a piece of commercial software[1] for working with freestyle/adaptive just intonation, if anyone’s interested

[0]: https://www.youtube.com/@new_tonality [1]: https://www.dmitrivolkov.com/projects/pivotuner/

Cool stuff, nicely fleshed-out. Sort of like a Risset rhythm, which I’ve explored a bit myself [1]. One of my favorite pieces that uses this technique is Black Rain, by Daniele Ghisi [2].

[1]: https://www.dmitrivolkov.com/misc/risset-polyrhythm/ [2]: https://soundcloud.com/danieleghisi/black-rain

This kind of reminds me of what happens when you implement an interpreter in a relational programming language, which lets you do cool stuff like generating quines by specifying that the program and it’s output should be the same.

Quick search lead to this paper which is exactly what I’m talking about: http://webyrd.net/quines/quines.pdf

If anyone’s curious to see complete pieces built out of this sort of technique, check out Conlon Nancarrow’s Studies for Player Piano (he does some other pretty crazy polytempo stuff too, like changing the speeds of different parts at different rates).

Personal favorites that use the same polytempo technique are No. 36 [1] and No. 37 [2].

I’ve been studying his music lately and am working on a piece with similar rhythmic experimentation, so happy to try to answer questions!

[1]: https://youtu.be/IOlId8O6D5w [2]: https://youtu.be/g0gNoELvpPo (I prefer the Werco recordings, but this is a cool vizualization)

Thanks! We were originally thinking of approaching things more from the educational side, but it’s kind of a hard point to make without experimental evidence, which we don’t have.

As far as playing a quantum circuit through the end, that’s what our python and max implementations do! Should work with any gate; but because it’s simulation-based, there’s a limit to the number of qubits. We’re also working on a follow-up about sonifying arbitrary hamiltonians.

I couldn’t make in in-person, but from what I heard the audience was pretty cool

a lot of it does end up being that kind of thing, but not all. some of it’s kind of like how quantum math has been used to model (if I remember correctly) seismology or weather, and some of it’s about sonifying the entire quantum state, including superpositions (assuming no error/noise, the random part of quantum doesn’t happen until measurement, where superpositions are destroyed)

I co-authored a paper that was presented at the 2nd symposium mentioned in this article, and I’ve read the textbook mentioned in the article, happy to try to answer any questions

The paper we wrote isn’t on airxiv yet but can be read here: https://github.com/sporkl/superposition-rhythms/blob/main/is...

It’s still pretty new, but people are taking advantage of GPUs for audio these days: https://www.gpu.audio

From what I've seen, the music industry (all genres including pop, classical, etc.) tends to be more about personality than actual content these days, so it's not really an issue.

Virtual singers are already pretty good anyways, I feel we've already passed the point of diminishing returns.

And even if you look at virtual singers, the ones which are popular are the ones from ten years ago, not the newer ones with more realistic voices.

I don't remember the details, but the reason the 2019 paper was challenged was because Google did not take into account quantum simulation optimization techniques when estimating how long the computation would run on classical computers. I wouldn't be surprised if there's a similar situation here, where unconsidered optimizations prevent this from being quantum advantage.

Music OCR is in general in a pretty terrible state these days (I’ve tried most of the options on the market). I’ve heard good things about soundslice[1] though, but I haven’t had the occasion to try it yet.

[1]: https://www.soundslice.com/sheet-music-scanner/

I knew some middle school students who would place their school-issued chromebooks upside-down on linoleum and slide around on them. So no surprise here.

I've tried using this in an argument and gotten a response along the lines of "I don't mind if people watch me poop."

The logo CrabLang is using is Ferris, which is not an official Rust logo, and also in the public domain I believe.

https://www.rustacean.net/

My main side project is Pivotuner[0], a MIDI effect plugin for adaptive pure intonation

It enables some pretty cool things like fairly straightforward microtonal modulation! Right now I'm working on a paper to submit to a computer music conference about it. Happy to send the draft if anyone's interested (email on website in bio).

I released it last December. I've had a couple sales, not many, but still enough to cover the cost of Apple developer membership, so at least I'm sort of in the positive!

Should also mention that I've gotten feedback on it from some pretty major artists, including Grammy nominees and winners like Jacob Collier! which is very very exciting

Happy to answer any questions!

[0]: https://www.dmitrivolkov.com/projects/pivotuner/

Same here. Attempting to install or update on my computer gives me a message saying that there's an SSL error.

If you don't have a house, and sleeping when you're in the car, then why would you bother commuting at all?