Browser sound models

Before sound starts

Nemisindo models do not autoplay. Choose Play or one of the model’s named sound actions to create or resume the browser audio system. Start at a comfortable device volume; the site uses a conservative output level and peak limiting, but the character of each model can still change as you adjust it.

Choose a starting point

Use Built-in preset to load a curated starting state. A model link can also name a preset in its query string, such as ?preset=Bad-Radio. If the requested preset is unavailable, the page reports that clearly and leaves the default state ready.

Randomize controls creates a new valid state inside every control’s supported range. Moving a control or randomizing clears an old preset query from the address so the URL does not describe a state that is no longer active. Double-click a model or post-effect slider to restore that parameter to the value from the last selected built-in preset. Global gain and Panning remain visible below the Post effects rack and are independent from presets; double-click either one to restore its neutral default.

Play and shape the model

Continuous models keep sounding until you choose Pause. Triggered models play a new event each time you choose Play or one of their named sound actions; Play remains ready while earlier bounded events finish, so supported voices can overlap. A model may expose more than one action, such as a single contact and a scheduled sequence. Every input has a visible label and is available to keyboard users. When controls belong to different actions, named sections keep those relationships clear. Hybrid models retain Pause for their continuous layer and active events. Browser interruptions cancel triggered events instead of leaving a sequence queued for later playback.

Try a continuous texture, a repeating tone, a triggered impact, or a multi-action mechanism to compare the behaviours.

Record and download

For continuous models, start the sound and then choose Record. For triggered models, you can choose Record before the first action so its opening transient is captured. The enabled recording control and elapsed timer show that capture is active; capture starts immediately, so trigger the sound when you are ready and expect any pause to be included. Pause affects playback without discarding captured audio. Choose the recording control again to finish the WAV; recording also stops at the displayed time limit.

The recorder captures the site’s own Web Audio output—not the microphone—and encodes a genuine WAV inside the browser. When encoding finishes, Download WAV becomes available. The model, preset, recording, and WAV data are not uploaded and no account is required.

If recording is unavailable

Recording requires browser support for the site’s AudioWorklet and Worker-based capture path. When a required capability is missing, the recording control stays disabled and the page explains why. Sound playback can still be available independently.