# Room Reverb for WAV Files

> Add room, hall, cathedral or cave reverb to a WAV file with a known reverberation time. Applies a Freeverb-style stereo reverb with room presets (small room, living room, studio, concert hall, cathedral, cave, open forest) to an audio file. The reverberation time (RT60) is set exactly per preset or by you and measured on the impulse response; the reverb tail is kept by default.

Part of usefulweb.site (https://usefulweb.site/). Part of [Soundscape Atlas](https://soundscape.usefulweb.site/) (ownerApp: soundscape). Results are deterministic calculations from the inputs and the listed assumptions; the API returns exactly what the web page shows.

## Use it
- Human interface: https://soundscapereverb.usefulweb.site/ (interactive tool at https://soundscapereverb.usefulweb.site/tool)
- JSON API: POST https://soundscapereverb.usefulweb.site/api/v1/calculate with a JSON object of inputs (schema: https://soundscapereverb.usefulweb.site/api/v1/tool)
- OpenAPI: https://soundscapereverb.usefulweb.site/openapi.json
- MCP (Streamable HTTP): https://soundscapereverb.usefulweb.site/mcp - tool `soundscape_reverb`

## Inputs
- file (file, required): Audio file - WAV directly, other formats are decoded with ffmpeg; up to 120 seconds, 8–96 kHz. The WAV file is kept on the server for about 2 hours so you can download it, then deleted.
- preset (select): Room
- rt60 (number): Reverberation time (optional) - Empty = the preset value. Time for the reverb to fall by 60 dB.
- wet (number): Reverb level (optional, 0–1) - Empty = the preset value.
- dry (number): Original level (optional, 0–1) - Empty = the preset value; 0 gives only the reverb.
- keepTail (boolean): Keep the reverb tail (the file gets longer)
- encoding (select): File format

## Method
- Freeverb structure (Jezar, public domain): per channel eight parallel comb filters with a low-pass in the feedback loop (damping) followed by four series all-pass filters; the right channel’s delays are 23 samples longer for a wide stereo image. Delay lengths are scaled to the sample rate.
- Unlike the original “room size” knob, each comb’s feedback gain is computed from the reverberation time: g = 10^(−3·delay / (sample rate × RT60)), so every comb decays by 60 dB in RT60 seconds at low and middle frequencies; damping makes high frequencies die away sooner, as in real rooms.
- The RT60 is verified by measuring the reverb’s impulse response with Schroeder backward integration (T20: the −5 to −25 dB part of the decay, extrapolated to −60 dB).
- Mono input becomes stereo. With “keep the tail” the file is extended by about 1.2 × RT60 plus the pre-delay so the reverb can fade out. If long reverbs on sustained sounds build up above full scale, the whole result is turned down to a −1 dBFS peak instead of clipping (reported as gain reduction).

## Limits
- Room presets give typical reverberation times for such spaces; they do not simulate a specific real room’s geometry or early reflections.