How to Synthesize a House Loop

Every generation of musicians for the past 8 decades has had the same thoughts. What live coding tools for synthesis offers you is an understanding of the nature of generational technology.

Consider this: there are teenagers today, out there somewhere, learning to code music. Remember when synthesisers were young and cool and there was an explosion of different engines and implementations?

This is happening for the kids, again.

Try to use this new technology to replicate the modern, and then the old sound, and then discover new sounds. Like we synth nerds have been doing for decades.

Look into the JUCE framework for building your own tools. I was using MaxMsp for a while, but would always think to myself "This would be so much easier to accomplish in pure code". So, I started building some bespoke VST's.

There's a learning curve for sure, but it's not too bad once you learn the basics of how audio and MIDI are handled + general JUCE application structure.

Two tips:

Don't bother with the Projucer, use the CMAke example to get going. Especially if you don't use XCode or Visual Studio.

If your on a Mac, you might need to self-sign the VST. I don't remember the exact process, but it's something I had to do once I got an M4 Mac.

For a great example of some (non-live) coded music, I would recommend The Haywire Frontier by Nathan Ho [0]. The whole album was sequenced and synthesized entirely in SuperCollider with no samples, external hardware, or third-party plugins. It's really interesting and a crazy achievement, definitely worth a listen.

For live coding, Switch Angel is definitely someone I would actually go to see live, check out this video of hers [1].

[0] https://nathanho.bandcamp.com/album/haywire-frontier [1] https://youtu.be/iu5rnQkfO6M

> I've watched a lot of live coding tools out of interest for the last few years, and as much as I'd like to adopt them in my music making it's not clear to me what they can add to my production repertoire compared to the existing tools (DAWs, hardware instruments, playing by hand, etc).

Aside from the novelty factor (due to very different UI/UX) and the idea that you can use generative code to make music (which became an even more interesting factor in the age of LLMs), I agree.

And even the generative code part I mentioned is a novelty factor as well, and isn't really practical for someone who actually makes music as their end-goal (and not someone who is just experimenting around with tech or how far one can get with music-as-code UIUX).

Give it some time.

I feel like the newer (ish) tools such as Strudel, and also this here Loopmaster, have a much better toolset for producing stuff that actually sounds great (vs just purely the novelty of "look im coding beats"). Like, Strudel comes with an extensive sample bank of actually quality samples (vs relying on synthesis out of some sense of purity), and also comes with lots of very decent sounding effects and filters and the likes.

Combine that with the ability to do generative stuff in a way that Ableton, FL Studio or Renoise are never going to get you, I won't be surprised if people invent some cool new genres with these tools eventually.

Basically, your comment reads a bit like saying demoscene makes no sense because you can make any video better with Blender and DaVinci Resolve. And this obviously isn't true given the sheer overload of spectacularly great demos out there whose unique esthetic was easy to obtain because they're code, not video. (find "cdak" by Quite for an on-the-nose example).

I'm going to be surprised if this new wave of music coding tools will not result in some madly weird new electronic music genres.

Obviously there's plenty of stuff these tools are terrible for (like your example of nuanced instrument parts), but don't dismiss the kinds of things they're going to turn out to be amazing at.

100% agree.

I think this format of composition is going to encourage a highly repetitive structure to your music. Good programming languages constrain and prevent the construction of bad programs. Applying that to music is effectively going to give you quantization of every dimension of composition.

I'm sure its possible to break out of that but you are fighting an uphill battle.

Procedural generation can be useful for finding new musical ideas. It's also essential in specific genres like ambient and experimental music, where the whole point is to break out of the traditional structures of rhythm and melody. Imagine using cellular automata or physics simulations to trigger notes, key changes, etc. Turing completeness means there are no limits on what you can generate. Some DAWs and VSTs give you a Turing complete environment, e.g. Bitwig's grid or Max/MSP. But for someone with a programming background those kinds of visual editors are less intuitive and less productive than writing code.

Of course, often creativity comes from limitations. I would agree that it's usually not desirable to go full procedural generation, especially when you want to wrangle something into the structure of a song. I think the best approach is a hybrid one, where procedural generation is used to generate certain ideas and sounds, and then those are brought into a more traditional DAW-like environment.

I see it as a neat way for nerds to nerd out about nerd stuff in an experiential way. Like, this is not going to headline a big time rave or festival or anything, but in a community of people who like math or programming or science, sure, why not introduce this kind of performance as another little celebration of their hobby?

Years ago I went to a sci-fi convention for the first time, because I had moved to a new town and didn't know anyone, and I like sci-fi. I realized when I was there that despite me growing up reading Hugo and Nebula award winners, despite watching pretty much every sci-fi show on TV, despite being a full-time computer nerd, the folks who go to sci-fi conventions are a whole nother subculture again. They have their own heroes, their own in-jokes, their own jargon... and even their own form of music! It's made by people in the community for the community and it misses the point to judge it by "objective" standards from the outside, because it's not about trying to make "interesting music" or write the best song of all time. The music made in that context is not being made as an end in itself, or even as the focus of the event, it's just a mechanism to enable a quirky subculture to hang out and bond in a way that's fun for them. I see this kind of live coded music as fulfilling a similar role in a different subculture. Maybe it's not for you, but that's fine.

Fair point, and that's the challenge in both the software's abilities and the creator's skills.

If you see it as yet another instrument you have to master, then you can go pretty far. I'm finding myself exploring rhythms and sounds in ways I could never do in a DAW so fast, but at the same time I do find limiting a lot of factors, especially sequencing.

So far I haven't gotten beyond a good sounding loop, hence the name "loopmaster", and maybe that's the limit, which is why I made a 2 deck "dual" mode in the editor, so that it can be played as a DJ set where you don't really need that much progression.

That said, it's quite fun to play with it and experiment with sounds, and whenever you make something you enjoy, you can export a certain length and use it as a track in your mix.

My goal is certainly to be able to create full length tracks with nuances and variations as you say, just not entirely sure how to integrate this into the format right now.

Feedback[0] is appreciated!

[0]: https://loopmaster.featurebase.app/

My wife's (and many other music-lovers') test for whether something counts as "real music" is whether they can perform it live (and sound as good as the recording). Music which is programmed doesn't count, as there's a lot of nuance that a skilled musician with an actual instrument can put into a performance in a split-second as they play.

here's a whole opera from a Star Trek episode I coded in Supercollider - can indeed code things other than EDM... (its a screen grab - being synthesized in real time)

https://vimeo.com/944533415?fl=ip&fe=ec

I've seen a couple of TikToks with someone doing live coding with this same tool and it was really cool to watch because they really knew it well, but like you said it was bog-standard house/techno.

I find trackers to be in the same category you put live coding into, probably DAWs as well, but many people do some amazing things with all three. In the more academic computer music world there is a fair amount of diversity in live coding where it is generally combined with algorithmic/generative techniques to rapidly create complexity and nuance. SuperCollider seems to have the most interesting output here, for me at least; I have seen little that really grabs me but they do show the capabilities of the process and I find that quite interesting. Improvisation and jamming is just not my thing, so live coding falls a bit short for me.

It's not gonna add anything to your repertoire. It will appear so after some time but it's really just an approach for people who have bad hand-eye coordination and ability to hold a rhythm or a hard time acquiring these skills, or tinkering with DAWs, which have a weirdly annoying first hour use time/learning curve

Often an article posted to hn will cause a mini-trend as users who are engaging with the subject discover and share more related resources.

Computer music is as old as computers, live coding is pretty old too. (I posted this in the strudel discussion too: https://toplap.org/wiki/HistoricalPerformances) Maybe everyone doing live streams during the pandemic helped get visibility for live coding? It's interesting to see it kind of becoming popular now.

Live coding music/visuals/art has been a fairly major subculture for over 15 years: https://blog.toplap.org/ Prior to that there was plenty of live/interactive code-based music going on within the computer music scene, HMSL (FORTH based)[1] and CLM (Lisp based)[2] come to mind.

Real-time sound synthesis was tough to live-code, or to run in real-time at all, prior to the faster personal computers of the early 90s. (The tracker scene obviously pre-dates this, but in that case the actual sound synthesis algorithms weren't live coded.) In fact, code-to-music dates back to 1951[3], or 1957[4], depending on your definitions. There is a large history of development by many computer musicians following on from Max Matthews' MUSIC-N. The Computer Music Tutorial[5] is a good source for the academic/research institutions/serious composers part of the picture.

[1] https://en.wikipedia.org/wiki/Hierarchical_Music_Specificati...

[2] https://ccrma.stanford.edu/software/clm/

[3] https://cis.unimelb.edu.au/about/history/csirac/music

[4] https://en.wikipedia.org/wiki/MUSIC-N

[5] https://mitpress.mit.edu/9780262044912/the-computer-music-tu...

CSOUND is the oldest code-2-music framework I know of, and that's been here since the 80's, so the concept is not new

The tools/frameworks have become more plentiful, approachable, and mature over the past 10-15 years, to the point where you can just go to strudel.cc and start coding music right from your browser.

I think the creator of the video you've linked has had a couple videos go viral, which may be part of it

My guess is we are either at the top or rising to the top of cyclical curve of the trend.

Thanks! I tried to make it as familiar as possible, inspired by JS. It's not yet open-source, mainly because the source is a bit of a mess, but it will be once I tidy things up. Follow me on GitHub[0] for updates. Also that sounds to me like Tech-House/Electro-House :D Very nice!

[0]: https://github.com/stagas

Nice, that's the optimized version - sounds actually a little different than the original one it's derived from. (actually better which I didn't expect) Original: https://dittytoy.net/ditty/24373308b4

I like how music recognition flags it as the original Jarre piece.

I first did stuff like this when I was a teen using a 6502 machine and a synth card - using white noise to make tshhh snares etc. All coded in 6502. The bible was Hal Chamberlin's Musical Application of Microprocessors.

Then of course we had games abusing the SID etc to make fantastic tunes and then came very procedural music in size coded PC and Amiga demo coding that underneath the hood were doing tiny synth work and sequencing very much like dittytoy etc.

Shadertoy even has procedural audio but it doesn't get used enough.

Fantastic to experience all of this!

So many tweaks but here is another:

change line 12 from 8000 to 800

you can also highlight one or multiple lines and press (cmd + /) to toggle if it is commented out or not to turn off a layer!

I'm considering a VST version but for now there is an Export Audio feature you can use to get a perfect audio loop to use with Ableton (or any other DAW) with oversampling up to 8x for great quality.

Shoutout to PACE who banned scripting in the JUCE 8 license terms so if you wanted to make this using the leading framework, you can't.

I route midi generated by strudel.cc in to my DAW.

And SuperCollider, Faust, Extempore:

https://supercollider.github.io/

https://faust.grame.fr/

https://github.com/digego/extempore

And numerous others

Thanks! For anyone trying this, it's being HN crushed right now and hitting rate-limits, you should try again in a bit if you see an error.

Also, there is an AI DJ mode[0] where you set the mood/genre and the AI generates and plays music for you infinitely.

[0]: https://loopmaster.xyz/editor?aidj

My understanding is that because it's a very small sample, it's basically a combination of a subset of sine waves, and because we're very sensitive to the nuances of high-pitched sounds, even small changes in that space make a lot of difference. Every RNG seed produces a different sounding hihat, and if you don't reset it, it continues producing different hihats, which is unnatural. Another explanation is also to resemble sample-based audio, but perhaps it's all of these things combined.

Yes, there is a `sidechain` function designed specifically for this. I wanted to keep this tutorial simple so I skipped a lot of mixing techniques or were left as an exercise to the reader, but I will try to cover those as well in future tutorials. Sign up to get notified for when they arrive!

For now you can see how it's done here[0] on line 139. I pretty much use it on every other track I've made as well.

[0]: https://loopmaster.xyz/loop/6221a807-9658-4ea0-bfec-8925ccf8...

If it was clipping it would show as red in the amplitude visualizer, however the bass does have a lot of energy and that might be the cause for the clipping you're hearing. You can multiply it by a factor to reduce its amplitude.

I think it's more of a red flag that they chose a name that's one letter away from a well-known site that sells music samples: https://www.loopmasters.com/

Not like a fringe unknown one, but one with over 20 years of history and now-owned by Beatport.