Strudel REPL – a music live coding environment living in the browser | Hacker News Reader

Strudel REPL – a music live coding environment living in the browser

strudel.cc

244 pointsbirdculture53 comments

Open articleView on HN

There are some pretty amazing live coding sessions of Strudel on YouTube. Some examples:

https://www.youtube.com/watch?v=HkgV_-nJOuE

Switch Angel is awesome. She also has some Instagram tutorials.

Dj Dave is the other creator I've found doing strudel content.

https://youtu.be/E1K6Sv-oIb0

"Hard Refresh" and "Airglow" made it onto my "On Repeat" playlist almost immediately.

Yea, Switch Angel and DJ Dave have a collab together somewhere. Probably on their Instagram, they are pretty active there.

I meant to include a link to DJ Dave, but it looks like I copy/pasted the same URL twice!

So random seeing her mentioned here today. I just discovered her yesterday, saw a Youtube short where she's operating a Eurorack synth at Switched On in Austin, TX. It's a cool little synth shop that's worth checking out if you're ever in town. Looks like they've moved locations recently though.

> So random seeing her mentioned here today. I just discovered her yesterday

I literally came across her on my youtube feed an hour back, and thought it'd be something HN would love.

Still distracted, I opened HN's default news page. Aaaand I saw this thread and your comment. So, no free karma for me lol, but still hope to see good discussions about the software. It feels very hacker-ey in an old-skool kinda fun way!

You accidentally pasted the same link twice. What was the second link meant to be? Would like to see it also :)

DJ Dave - someone else posted a link above.

(silly copy/paste error!)

Thanks :)

This was one of my favorites -- with the voice filter the narration feels like it's part of the song which I found especially fun.

https://www.youtube.com/watch?v=GWXCCBsOMSg

That channel has been all over my recommendations, it's awesome so much skill!

Me too! Came on YouTube feed today. Blown the fuck away. I Google Strudel REPL to learn more and found this thread as well. :)

So stoked to play with this.

Someone should remix the audio in this video, including vocals.

Strudel is a JavaScript port of TidalCycles (Haskell). While TC uses SuperCollider for the synthesis, Strudel uses superdough which seems to be a custom implementation. I'm currently learning SuperCollider sclang and waiting for a version upgrade to have a reason to submit it here - usually some of the discussion is quite insightful. Anyway sclang is the PHP of music - just uglier and less consistent. But it's also powerful and and quite fun.

I was goofing around with TidalCycles and really wanted to use it for the Haskell syntax but Strudel’s interface is so slick I suffer the JS syntax.

Thanks for mentioning superdough I hadn’t seen it anywhere while I was playing with all of the above. Piqued my curiosity :)

Might be worth checking out Tidal's Mondo Notation, which while not quite Haskell syntax is far closer to it, being a proper functional style notion, that unifies with mini notation, so no need for wrapping many things in strings.

Looks like this:

    mondo`
    $ note (c2 # euclid <3 6 3> <8 16>) # *2 
    # s "sine" # add (note [0 <12 24>]*2)
    # dec(sine # range .2 2) 
    # room .5
    # lpf (sine/3 # range 120 400)
    # lpenv (rand # range .5 4)
    # lpq (perlin # range 5 12 # \* 2)
    # dist 1 # fm 4 # fmh 5.01 # fmdecay <.1 .2>
    # postgain .6 # delay .1 # clip 5

    $ s [bd bd bd bd] # bank tr909 # clip .5
    # ply <1 [1 [2 4]]>

    $ s oh*4 # press # bank tr909 # speed.8
    # dec (<.02 .05>*2 # add (saw/8 # range 0 1)) # color "red"
    `

If actual tidal notation is important, that has been worked on, and would look like:

    await initTidal()
    tidal`
    d1 
    $ sub (note "12 0")
    $ sometimes (|+ note "12")
    $ jux rev $ voicing $ n "<0 5 4 2 3(3,8)/2>*8"
    # chord "<Dm Dm7 Dm9 Dm11>"
    # dec 0.5 # delay 0.5 # room 0.5 # vib "4:.25"
    # crush 8 # s "sawtooth" # lpf 800 # lpd 0.1
    # dist 1

    d2 
    $ s "RolandTR909_bd*4, hh(10,16), oh(-10,16)"
    # clip (range 0.1 0.9 $ fast 5 $ saw)
    # release 0.04 # room 0.5
    `

Only the actually implemented functions, and implemented custom operators are available even when that works, so not all tidal code can necessarily be imported.

But it is currently broken on the REPL site because of https://codeberg.org/uzu/strudel/pulls/1510 and https://codeberg.org/uzu/strudel/issues/1335

In the supercollider forum there is talk about a wasm port of supercollider: https://scsynth.org/t/webassembly-support/3037

I wonder if that could be used at some point.

When I last played with SuperCollider I used Overtone, that wraps everything in a Closure API. With that you use s-expressions instead of sclang to define your sounds. I am not sure what the state of Overtone is these days, but there seems to still be some activity: https://overtone.github.io/

sclang came across to me as something like a hybrid of Smalltalk and Ruby. It's indeed very "inconsistent" and weirdly familiar-yet-alien, but I would consider it a lot more elegant than PHP.

This is cool because a lot of the current tools are a bit old and I feel a bit like they suffer from NIH (not invented here) syndrome, where what is actually needed is for things to just be in javascript.

This wasn’t possible as much when the last gen of tools came out (sonic pi etc) but I think the time is right.

The next iteration that would be cool is a true two-way interface between the visualizations and the code. Right now the slider is a really awesome element, for example. I think Bret Victor would be proud.

I'm not very musically inclined but this is what I was able to make:

$: arrange( [4, "<sh09_bd>(4,8)"], [4, "<sh09_bd>(4,8)"], [1, "<sh09_bd mfb512_sd>(6,6)"] ).s().fast(2).layer(x=>x.add("0,2")).gain(".4!2 .5").phaser(2).phasercenter("<4000 800 4000 4000>")

$: s("gm_tinkle_bell").distort("<1 2 1 2:.5>").crush("<8 8 8 6 6 8 8>").chop(4)

$: arrange( [2, "<c4 e4 g4>(3,8)"], [1, "<f4 a4 c5>(3,8)"], [1, "<c4 e4 g4>(3,8)"] ).note().chop(4).fast(4).distort("<3:.5>").phaser(4).phasercenter("<800>").fm(4).fmdecay("<.05 .05 .1 .2>").fmsustain(.4)._scope()

I don't know what half this stuff does but it was still so much fun and this is probably one of my favorite projects ever. What made it most fun for me is that the reference docs are in the page so it's really easy to pick something at random and just see what it does.

Oh and there is flok[1] which combines the strudel repl with visuals from hydra. Also there are sclang and other algorave environments available. Everything is synced (with crdts i guess) so it’s live collaborative. Which is nice to remotely jam with friends

[1] https://flok.cc

And quite new dilber.io: https://github.com/Omodaka9375/Dilber

Also Gibber: http://gibber.cc/

You can also use hydra in the standard strudel repl, and vice-versa.

I like this intro video: https://www.youtube.com/watch?v=QRJ0xrjLj6A

  Uncaught (in promise) ReferenceError: AudioContext is not defined
    ln https://strudel.cc/_astro/spectrum.Bf7jMx6O.js:1
    W https://strudel.cc/_astro/spectrum.Bf7jMx6O.js:1
    un https://strudel.cc/_astro/spectrum.Bf7jMx6O.js:1
    c https://strudel.cc/_astro/spectrum.Bf7jMx6O.js:1

What would I need to change in my security settings for this?

This being text based makes it really easy to have AI generate the music. Now waiting for Strudel agent that will transcribe music into strudel notation.

I like doing my own composition, but AI being able to just do sound design from a prompt ("hey can you make a synth that sounds like the lead from XYZ song but buzzier, add parameters to control ABC") would be revolutionary.

Yes that's possible. Here's Sonnet 4.5 for your request: https://claude.ai/share/5672940d-99b3-4b01-9a9f-fb22ca59fa41

What the implications are for music, I don't know.

Before long we can get agents to listen to the music, too. Then based on our prior data can tell us how much we liked it.

That'll save me so much time listening to music!

> This being text based makes it really easy to have AI generate the music.

It's already really easy for AI to generate music without using text as an intermediate format.

Can it make music like e.g. Extempore (see https://www.youtube.com/playlist?list=PL_eJ0XdLbWzzq_03wTIMV...)?

All examples I've heard from Strudel so far are pretty boring (constant beat/chord machine music).

Are there examples in other styles?

Strudel can use custom samples in addition to the built in synths and samples. The language is really expressive. I’ve not gone too far into playing with it but from what I’ve seen it’s pretty flexible.

That said I’ve only seen people making house/techno/drum-n-bass kinda stuff with it.

The syntax is pretty relevant for the kind/compexity of the aspired music. The music from the examples is quite simple compared to what Soerensen does with his Lisp-like syntax. Strudel seems to go more towards SuperCollider syntax, which from my humble point of view is better suited for offline productions.

I added an example of some music other than what you describe in comments above (below? in some direction)?

Awesome!

Tide Cycles doesn't work on Fedora, so I might use this instead. Anyway to get it running as a node js script so I use it locally?

It will run locally in a web browser, it’s open source and if I recall correctly the authors provide a downloadable way to serve it from your local machine.

Here's my (Lil Data) remix of DJ_Dave's Array

https://strudel.cc/?mTeJt_ICoPrw

https://www.youtube.com/watch?v=QrMk5HTVwJQ

thanks to share! I like having examples to see how the stuff actually works :)

quick question, where do strudel load those sounds? snd/bass/build_0.mp3 (line 11) ? are those built in samples?

If you're new to this, you can check this video: https://www.youtube.com/watch?v=vuSZQnkOB_Y

It will explain the process DJDave uses to create music under strudel

is it still possible to run the latest version in appimage on linux? only found this version, it's from 2023: https://club.tidalcycles.org/t/strudel-0-8-0-released/4769

i remember seing a .appimage hosted on github in the releases, but they moved to a new host now

oh and if you don't know what inspired the name of the software ; a strudel is really yummy: https://en.wikipedia.org/wiki/Strudel

For a much more open ended (but advanced) option, one can run Csound live in the browser too now over WASM.

For some music ed stuff I work on, I actually have s7 Scheme in WASM controlling csound in WASM, both were surprisingly easy to get going!

There are plenty of instruments in there. I did a quick and dirty encoding of the first bar of “City of Star” with the piano when I first discovered it.

note("G2@2 A#2 D@2 G@2 ~ G F@2 D@1.5") .sound("piano")

Like others I saw this in YT recently. It's funny how the interface makes you think it's a console but you'l see a slider in there.

I always wanted to try tidal cycles but the setup always seemed daunting (at least when I last looked). This is pretty cool

That's amazing and actively developed, I wonder if we'll see some actual tracks being remade in this

Strudel did something many other approaches to live coding have failed to do imho. No hosting needed. Just open the web based REPL and go. Great entry to live coding, but you quickly run into limitations.

Here is a piece inspired by Dawn of Midi and my attempt at taking a piano synth and trying to make it sound like a lof of different things.

Copy, paste, modify.

const bpm = 138; setcps(bpm/60/4);

// Pattern 1: lower melodic pulse

const bass_pulse = note("<d2 f2 a2 g2>") .s("piano") .slow(4) .gain(rand.range(0.45, 0.65)) .attack(0.005) .decay(0.8) .sustain(1.2) .release(1.2) .lpf(800) .room(0.3) .delay(0.15) .delaytime(0.375) .delayfeedback(0.25) .pan(0.5);

// Pattern 1.5 (?): layered base

const bass_pulse_2 = note("<d2 f2 a2 g2>") .s("piano") .slow(4) .gain(rand.range(0.45, 0.65)) .add(note(12)) .attack(0.005) .decay(0.8) .sustain(1.2) .release(1.2) .lpf(800) .room(0.3) .delay(0.15) .delaytime(0.375) .delayfeedback(0.25) .pan(0.5);

// Pattern 2: Mid-range polyrhythm

const mid_pattern = note("<a3 c4 d4 f4 a3>") .s("piano") .struct("x(5,8)") .gain(rand.range(0.25, 0.45)) .attack(0.008) .decay(0.4) .sustain(0.05) .release(0.6) .lpf(perlin.range(1200, 2200).slow(8)) .room(0.5) .pan(rand.range(0.3, 0.7));

// Pattern 3: repetitive pulse

const high_pulse = note("d5 [~ d5] d5 ~") .s("piano") .fast(2) .gain(rand.range(0.18, 0.35)) .attack(0.01) .decay(0.3) .sustain(0) .release(0.4) .lpf(2800) .room(0.6) .delay(0.25) .delaytime(0.1875) .delayfeedback(0.3) .pan(0.7);

// Pattern 4: Sparse accent notes (3 over 4 polyrhythm)

const accents = note("a4 ~ f4") .s("piano") .slow(2) .gain(rand.range(0.35, 0.55)) .attack(0.5) .decay(0.6) .sustain(0.9) .release(0.9) .lpf(1800) .room(0.45) .pan(0.2) .sometimes(x => x.delay(0.3).delayfeedback(0.4));

// Pattern 5: Extended mid-range polyrhythm (13 over 16 - cello thingie)

const mid_long = note("<a3 c4 d4 f4 a3 c4 e4 d4 f4 g3 a3 c4 d4>") .s("piano") .struct("x(13,16)") .gain(rand.range(0.32, 0.48)) .attack(0.06) .decay(0.9) .sustain(0.25) .release(1.1) .lpf(perlin.range(800, 1400).slow(12)) .lpq(4) .room(0.55) .delay(0.18) .delaytime(0.25) .delayfeedback(0.3) .pan(rand.range(0.35, 0.65));

const high_long = note("<d5 a4 f5 d5 c5 a4 g4>") .s("piano") .struct("x(7,8)") .gain(rand.range(0.28, 0.42)) .attack(0.05) .decay(0.8) .sustain(0.3) .release(1.0) .lpf(sine.range(1000, 1600).slow(8)) .lpq(3.5) .room(0.6) .delay(0.22) .delaytime(0.1875) .delayfeedback(0.35) .pan(0.7);

stack( bass_pulse, bass_pulse_2, mid_long, high_long, mid_pattern, high_pulse, accents );

It's cool and all, and I like it. I have TidalCycles installed and have played around with it.

My only criticism is it makes music feel like CSS. In some ways it helps with theory, yes, but the DX is more like Tailwind.

Just got introduced to this today, honestly as an ECE major finishing up his BS, I legit want to do this as an edm hobby & skill honing

Reply on news.ycombinator.com