Filter rack

Drive adds harmonics by pushing the signal into a curve. A little warms it; a lot breaks it up. Input gain sets how hard it hits.

Filter removes or emphasises a band of frequencies. Lowpass keeps the bottom, highpass keeps the top, bandpass keeps a window. Resonance boosts the frequencies right at the cutoff, which is what makes a filter sweep sound like a sweep.

Delay repeats the signal. Time sets the gap, feedback sets how many repeats, mix sets how loud they are against the original.

Output is level, after a compressor that catches peaks so a heavy setting does not clip.

The two sources are independent. Start a clip, then switch the microphone on and both play at once — sing or talk over the loop, and the rack processes the pair together. Each source has its own level so you can sit your voice on top of the music instead of under it.

Stopping one leaves the other running, and if you are recording, the take keeps going until both are off. Use headphones: a microphone in a room with speakers playing a delayed, resonant signal will feed back.

A preset is just the position of every control, written down as data. The block above shows the current settings in that form — the same nine numbers the tool reads when you click one of the chips.

Letting creators publish their own presets for other people to load is something we plan to add. It is not built yet, and there is no date on it.

Nowhere. Processing happens in your browser using the Web Audio API, the microphone feed is read through getUserMedia, and a recorded take is captured with MediaRecorder into your browser's memory and handed straight back to you as a download. No audio is transmitted to ZygApp and none is stored on our servers.

Because a take is held in memory, recording is capped at 60 seconds. The tool stops on its own and tells you why.

The demo clips are not files. They are generated by code in your browser each time you pick one, which is why they load instantly and why no creator is credited for them.