Delay is an audio effect that stores a signal and plays it back after a set time, usually mixed with the original. Among audio effects, it is the basic time-based process: depending on the delay time, feedback and filtering, it produces anything from subtle thickening and slapback to rhythmic echoes that repeat in time with the music.
Delay is also the building block of other effects. Chorus, flanger and many reverb algorithms are made from short delays that are modulated or combined in networks, so understanding delay explains much of the time-based effects family.
How a delay effect works
A digital delay writes incoming samples into a buffer and reads them out a fixed time later. The delayed output is blended with the dry signal by a mix (wet/dry) control. Sending part of the output back to the input creates feedback, which produces a series of repeats that decay gradually.
Each repeat passes through the feedback path again, so any filter in the loop shapes every later echo more strongly. Analog tape and bucket-brigade delays darken and degrade their repeats naturally, which is why many digital delays add low-pass filtering, saturation and slight modulation to imitate that character.
| Parameter | Typical range | Effect on the sound |
|---|---|---|
| Delay time | 1 ms to 2 s or more | Distance between the original and the echo |
| Feedback | 0 to about 95 percent | Number of repeats and length of the tail |
| Mix | 0 to 100 percent wet | Balance of echo against dry signal |
| Filter (high-cut, low-cut) | About 1 kHz to 10 kHz high-cut | Keeps repeats behind the dry sound |
| Modulation | Low rate, small depth | Pitch movement and analog-style warmth |
Delay time ranges and what they sound like
Human hearing treats short and long delays very differently. Below about 30 to 40 ms, the repeat fuses with the original and is heard as a change in tone or width; above that, it becomes a distinct echo.
| Delay time | Perceived result | Typical application |
|---|---|---|
| Under 1 ms to 10 ms | Comb filtering, tonal coloration | Flanger, phasing-like effects |
| 10 ms to 35 ms | Doubling and widening | Chorus, double tracking, Haas widening |
| 60 ms to 140 ms | Slapback echo | Rockabilly vocals, guitars |
| Tempo-synced (eighth, quarter note) | Rhythmic echo | Vocal throws, guitar parts, synthesizers |
| 500 ms and longer | Spacious, discrete repeats | Ambient textures, dub effects |
For tempo-synced delay, the time of a quarter note in milliseconds is 60,000 divided by the tempo in BPM; at 120 BPM it is 500 ms. Dotted and triplet values create rhythmic counterpoint, as in the dotted-eighth guitar delays common in rock and pop.
Types of delay
Delay units differ in how they generate and route repeats. Each type has a distinct use in mixing and sound design.
| Type | How it works | Character |
|---|---|---|
| Digital delay | Sample buffer with clean playback | Precise, unchanged repeats |
| Tape delay | Record and playback heads on moving tape | Warm, saturated, slightly wavering repeats |
| Analog (bucket-brigade) delay | Charge passed along capacitor chain | Dark, noisy, quickly degrading repeats |
| Ping-pong delay | Repeats alternate between left and right | Wide, moving stereo echoes |
| Multi-tap delay | Several read points in one buffer | Complex rhythmic patterns |
| Granular or pitch delay | Repeats processed by pitch shifting | Shimmering, evolving textures |
Delay compared with reverb and modulation effects
Delay, reverb, chorus and flanger all use time offsets, but they differ in density, length and modulation. A delay produces a small number of clearly separated repeats, while reverb produces thousands of reflections that blur into a continuous tail.
| Effect | Delay structure | Perceived result |
|---|---|---|
| Delay | One or few repeats, set time | Discrete echoes |
| Reverb | Dense network of many short delays | Room or space impression |
| Chorus | 10 to 35 ms delay with slow modulation | Ensemble-like thickening |
| Flanger | 1 to 10 ms delay with modulation and feedback | Sweeping jet-like comb filter |
In practice, a delay often replaces some reverb on vocals, because echoes add space without the dense tail that can cloud the words and push the singer back in the mix.
Using delay in a mix without clutter
Delay is usually placed on an auxiliary send so that several sources can share it and the return can be filtered and balanced independently. Filtering the return with a high-pass around 150 to 300 Hz and a high-cut around 5 to 8 kHz keeps the echoes from muddying the low end or competing with the dry sound.
- Too much feedback builds up runaway repeats that clutter the arrangement or self-oscillate.
- Unsynced delay times create rhythmic conflicts unless they are intentionally offset.
- Echoes over dense passages mask the lyrics; automating the send for phrase endings, known as delay throws, avoids this.
- Ducking the delay with a compressor keyed from the dry vocal keeps repeats quiet while the singer sings and lets them bloom in the gaps.
- Haas widening with very short delays can collapse in mono and should be checked for comb filtering.
A plain-language summary
A delay is an artificial echo. It remembers what was just played and plays it again a moment later, like shouting across a valley and hearing your voice come back.
Short echoes make a sound feel thicker or wider, while longer echoes repeat in time with the music. The trick is to use just enough echo to add space and movement without making the song sound crowded.
An example from practice
A mixing engineer is working on a ballad at 75 BPM in which the lead vocal sounds dry and detached from the band. Adding more reverb makes the words harder to understand.
The engineer sets up a delay on a send with a quarter-note time of 800 ms, 25 percent feedback and a high-cut at 4 kHz, then places a compressor after it, keyed from the vocal, so the echoes duck by about 6 dB while the singer is active. At the end of the chorus, the send is automated up for a single delay throw on the last word.
The vocal now sounds spacious yet stays clear, and the repeating final word fills the gap before the next verse without masking any lyrics.