How the Audio Reverb Effect Works
AnyTool adds reverb entirely in the browser with the Web Audio API. A persistent graph splits the signal into a dry path and a wet path — pre-delay (DelayNode) into a ConvolverNode loaded with the chosen impulse response, then damping and tone biquads, wet gain and an optional echo (feedback or ping-pong delay) — summed through a stereo-width stage. Impulse responses are a small bundled pack of measured-style spaces (studio/hall/plate/cathedral/spring/chamber) plus a synthesized Moorer-style fallback, and you can drop in a custom .wav IR. Every control ramps live with setTargetAtTime, and a one-click dry-vs-wet A/B is gapless. Export rebuilds the graph in an OfflineAudioContext sized for the reverb tail, writes WAV, and a self-hosted single-threaded ffmpeg.wasm engine (bundled core, no CDN) encodes MP3/M4A/OGG/FLAC with an anti-clip alimiter/loudnorm pass.
- Runs client-side — the audio is never uploaded
- Web Audio ConvolverNode reverb with a bundled IR pack
- Live wet/dry, decay, pre-delay, damping and stereo width
- Multi-tap / ping-pong echo and custom-IR (.wav) upload
- Self-hosted ffmpeg core (bundled) — no CDN
- Export WAV/MP3/M4A/OGG/FLAC with an anti-clip pass
