Sound, synthesis, rhythm & noise — making music with code since someone dared to ask "what if the interface was the instrument?"
Music apps, instruments, sequencers, and audio experiments. The ones where the output is something you can actually hear.
The threads of audio curiosity that keep weaving through every build on this page.
Oscillators, gain nodes, filters, convolution reverb — the browser has a real-time audio graph built in. Most people ignore it. We do not ignore it.
Additive, subtractive, FM, wavetable. Every synthesis method is a different answer to the question: what is a sound, and how do you manufacture one from nothing?
Step sequencers, Euclidean rhythms, polyrhythm generators. When you reduce music to a loop of on/off decisions, strange things become possible.
Music that writes itself. Markov chains, L-systems, probability tables, and neural models all produce sequences no human directly authored — some of them are even good.
FFT, spectrograms, and the view of sound as frequency content over time. Once you can see music, you hear it differently. And once you can edit the spectrum directly, you hear it differently again.
Karplus-Strong plucked strings, digital waveguide flutes, resonating bodies. Simulating the physics of acoustic instruments — close enough to fool the ear, different enough to be its own thing.
Smaller sonic contraptions. Built fast, played loud, sometimes broken in interesting ways.
A few things we keep coming back to when someone asks why bother building music tools from scratch.
A MIDI file is just numbers. An audio waveform is just an array of floats. The moment you accept that music is data, the entire toolkit of software engineering opens up for sound.
Unlike visual art, music exists entirely in time. Building music software forces you to think carefully about scheduling, latency, and the precise relationship between clock ticks and perceptual experience.
A QWERTY keyboard played as a piano sounds terrible — not because the notes are wrong but because the interface shapes what's playable. Every new interface implies a new musical vocabulary.
Audio bugs are often more interesting than audio features. Ring modulation, bitcrushing, buffer aliasing — half the vocabulary of electronic music started as someone's mistake that turned out to sound incredible.
4-bit samples. 3 oscillators. A 16-step grid. The constraints of early music hardware produced some of the most distinctive sounds in history, because limits force creativity.
Change a parameter, hear it instantly. Music software has the fastest feedback loop in all of programming. That's addictive — and it makes it one of the best domains for learning to build interactively.