Reverb and delay place sounds in a space and give a mix depth. They're also the fastest way to make a mix muddy and distant. The craft is using enough to create space while keeping the important parts clear and close.
Sends, not inserts
Put reverb and delay on send/return (aux) channels, not directly on each track:
- Create an aux channel and put the reverb on it.
- Set the reverb's mix to 100% wet — the return carries only the effect, never the dry signal.
- On each track, use a send knob to feed some signal to that aux.
This has real advantages:
- Several parts share one space, so they sound like they're in the same room.
- You control the effect level with one fader and can EQ or compress the reverb itself.
- One reverb instead of twenty saves processing.
- You can automate the return level for a whole section.
Inserts still make sense for effects that are part of a sound — a guitar's spring reverb, a synth patch's built-in delay, a deliberately drenched special effect.
If the return isn't 100% wet, you're adding a second copy of the dry signal to the mix, which raises the part's level and can cause phasing. This is the most common send mistake.
Reverb controls
| Control | What it does | Starting point |
|---|---|---|
| Type | Room, hall, plate, chamber, spring | Plate for vocals, room for drums, hall for pads |
| Decay | How long the tail lasts | Rooms 0.4–1 s; plates 1.2–2.5 s (pop vocal); halls 2–4 s |
| Pre-delay | Gap between dry sound and the reverb starting | 20–40 ms for vocals |
| High / low cut | Filters the reverb return | High-pass 200–300 Hz, low-pass near 8 kHz |
Pre-delay is the most useful and least used control. A gap of 20–40 ms lets the dry voice speak clearly before the reverb arrives, so it stays at the front of the mix while still sounding spacious. Longer pre-delay keeps a sound closer; shorter makes it recede.
Filtering the return stops the reverb muddying the low mids and hissing in the top. Reverb tails with full low end are one of the main causes of a muddy mix.
Delay
A delay repeats the signal after a set time. Synced to tempo, it reinforces the groove; unsynced, it can blur it.
Delay times in milliseconds from tempo:
| Note value | Formula | At 120 BPM |
|---|---|---|
| Quarter note | 60000 ÷ BPM | 500 ms |
| Eighth note | 30000 ÷ BPM | 250 ms |
| Dotted eighth | 45000 ÷ BPM | 375 ms |
| Sixteenth | 15000 ÷ BPM | 125 ms |
Common uses:
- Slapback: a single repeat at 80–140 ms, no feedback. Thickens a vocal or guitar without obvious echo.
- Dotted-eighth delay: rhythmic repeats that fill gaps between phrases.
- Throws: automate a send up on the last word of a phrase so only that word echoes.
Filter delay returns too — roughly 300–500 Hz to 4–8 kHz — so repeats sit behind the dry signal. Ducking the delay (sidechain-compressing the return from the dry vocal) keeps repeats out of the way while the singer is singing and lets them bloom in the gaps.
How much
Listening tests on real mixes found that every mix whose reverb sat about 9 LU below the dry signal, or louder, was judged too wet. Around 14 LU below the dry signal was preferred, and listeners penalised too wet more than too dry.
In practice: set the reverb so you can hear it when you listen for it, then pull it down a little more. Mute the return now and then — if the mix falls apart, it was doing its job; if it sounds clearer, it was too loud.
Depth
Things further away are quieter, darker, wetter and narrower. To push a part back, lower it, roll off some top end and send more to reverb. To bring it forward, do the opposite: louder, brighter, drier, with pre-delay so any reverb it has sits behind it.
Sources
- Brecht De Man, Kirk McNally and Joshua D. Reiss (2017). Perceptual Evaluation and Analysis of Reverberation in Multitrack Music Production. Journal of the Audio Engineering Society 65(1/2).
- Roey Izhaki (2023). Mixing Audio: Concepts, Practices, and Tools (4th edition). Focal Press / Routledge.