AUD · Audio tools

Audio Reverb

Place an audio file in a simulated room. Eight parallel comb filters feed four allpass stages, the arrangement used by Freeverb, with the comb lengths offset between channels so the stereo tail does not collapse to the centre.

Room size sets the decay, not the level

Room size raises the feedback inside the comb filters, which is what makes reflections persist. Larger values produce a longer tail.

Left alone, more feedback would also mean more energy, so a large room would simply be louder. The input feeding the reverb is scaled down in proportion, which holds the wet level steady across the range and leaves room size controlling decay time alone.

Damping

Damping puts a low-pass filter inside the feedback path, so each pass through the reverb loses more high frequency than the last. The tail starts bright and gets darker as it decays.

Real spaces behave this way. Soft furnishings, carpet, and curtains absorb treble faster than bass, which is why a bedroom sounds duller than a bathroom. Low damping suits tile and glass; high damping suits a furnished room.

Pre-delay

Pre-delay holds the reverb back by up to 500 ms while the dry signal passes through immediately. The gap between the two is what the ear reads as room size, and it keeps the source from being smeared by its own reflections.

On speech and vocals, somewhere between 20 ms and 50 ms usually keeps consonants readable while still putting the voice in a space. At 0 ms the reverb starts with the source and blurs the attack.

The tail is rendered past the source

A reverb outlives its input. When the source stops, the export keeps running the reverb on silence until the remaining energy drops below -60 dB, then finalises the file. That is why an exported file runs a few seconds longer than the one you loaded.

The preview works differently: it processes the source as it plays, so it ends when the source ends. Judge the decay length from the exported file rather than from playback.

Mix

Mix sets the balance between the dry signal and the reverb. At 0 the audio passes through untouched, and the reverb is skipped entirely rather than being mixed in at zero level.

Frequently Asked Questions

It adds the reflections of a simulated space to the audio, mixed against the original signal. It reads WAV, MP3, M4A, AAC, Ogg, Opus, FLAC, AIFF, and WebM audio, and exports in the format the source arrived in.

The decay has to finish somewhere. The export keeps rendering after the last sample of the source until the tail has fallen below -60 dB, which adds up to a few seconds depending on room size.

The preview follows the source, so it ends when the source does. The tail is added during export, which is why the downloaded file rings on and the preview does not.

No. The input feeding the reverb is scaled against the feedback amount, so room size changes how long the decay lasts without changing its level.

It holds the reverb back by up to 500 ms so the dry signal is heard first. On voice, 20 ms to 50 ms keeps words intelligible while still placing them in a space.

Explore Our Tools

Browse all tools