Audio Reverb that sounds like a place

Reverb is the sound of a space answering back. Clap in a stairwell and you hear dozens of returns of the same clap, each quieter and slightly different. A reverb effect builds that pattern artificially so a recording made in a small dry room can pretend to be somewhere else.

The tail tells you where you are

The length of the tail is the strongest cue. A short tail means a small room: the reflections die before the next syllable. A long tail means a large hall: the sound keeps returning while new words arrive. That is why reverb on speech is risky. The effect does not know which part is the message.

Three controls, in order

Space decides the character. Rooms are tight and neutral, halls are wide and dark, plates are dense and bright. Decay sets how long the tail survives. Mix decides how much of the wet copy you hear against the dry original.

A useful starting point for narration is a small room or a studio at 15–20 percent. For a sung vocal, a plate or a hall at 20–30 percent. For an instrument that should sit far away in a mix, more — but then it is no longer the focus and does not need to be understood word for word.

Pre-delay keeps the voice in front

Pre-delay is the gap between the dry sound and the first reflection. Without it the reverb seems to start inside the mouth, which sounds unnatural. With 15–30 ms the voice stays forward and the room sits behind it. Longer pre-delays create the impression of a bigger space than the decay alone suggests.

Common mistakes

• Adding reverb to an already wet recording. Two tails overlap and the source disappears into a wash. Fix the recording or accept it dry.
• Using reverb to hide noise. It makes hiss more audible, because the wet copy is mostly noise. Use the noise remover first.
• Judging on headphones only. A phone speaker in a real room already adds its own space. Check both.
• Confusing reverb with delay. A delay repeats the whole sound; reverb smears it. They solve different problems.

Reverb and loudness

Spreading a sound over time lowers its peak, so a wet file usually measures quieter than the dry one. That is normal. If the result has to match a catalogue or a podcast host, normalize it afterwards so the platform does not turn it up for you. The normalizer handles that in one pass.

Finally: a little goes a long way. If you can name the effect while listening to the content, it is probably too much.