Delay (audio effect)

A delay stores incoming audio briefly and then reads it back later. A digital processor uses a memory buffer; tape and analogue circuits achieve related effects through different physical mechanisms. The delay time controls the gap between the original sound and its copy. Short gaps can merge perceptually with the source, while longer gaps produce recognisable echoes. The exact impression also depends on the source, playback level and surrounding arrangement.

Feedback returns some delayed output to the delay input, producing further repeats. Increasing feedback usually makes the echo tail last longer, but the relationship depends on the processor. Filtering or saturation inside that loop can change each successive repeat. Some settings can build toward self-oscillation, so begin with conservative levels and keep a clear way to bypass or mute the return.

The wet signal is the processed echo, and the dry signal is the original. On a parallel effects return, the delay is usually set fully wet because the source already reaches the mix through its own channel. An insert may instead blend both internally. Follow the actual signal flow rather than assuming that every preset suits both arrangements.

Delay can make a sparse recording feel spacious without placing every note inside a dense reverberant wash. A single repeat can answer a vocal phrase, while quiet short echoes can add apparent thickness. Because repeats remain recognisable, they also occupy rhythmic and lyrical space. A setting that flatters an isolated singer may obscure the next line once the whole arrangement plays.

Creative timing is only part of the decision. Echo level, stereo placement and frequency balance determine whether the listener hears an obvious effect or simply a sense of depth. Filtering low frequencies from repeats can reduce buildup, while softening their high frequencies can keep consonants from competing with the lead. These choices should serve the song rather than imitate a particular hardware label.

Tempo-synchronised delay expresses time as note values. A quarter-note delay in milliseconds is 60,000 divided by the tempo in beats per minute when the beat unit is a quarter note. Dotted and triplet settings create different relationships to the pulse. In music with tempo changes, confirm whether the processor follows the tempo map and whether changing time produces pitch movement or another audible transition.

A slapback effect usually uses a relatively short delay with little feedback, giving a distinct reflection without a long trail. Ping-pong routing alternates or moves repeats between stereo sides, depending on the design. Neither description guarantees a specific timing value. Choose the gap by listening to articulation, groove and the distance between phrases rather than applying one universal number.

Delay and reverb overlap creatively but organise reflections differently. Reverb normally produces a dense field of decaying energy; delay often preserves identifiable repetitions. Combining them can work, although processing order changes the result. Reverberating an echo creates a different envelope from delaying an already reverberant signal. Compare both quietly before committing to a complicated effects chain.

In a Dubai recording session, a singer leaves a deliberate pause after the final word of each chorus line. The engineer wants that space to feel animated without making the verse less intimate. Rather than leaving a loud echo active throughout the song, the engineer creates a dedicated delay return and automates selected phrase endings into it.

• Start with the full arrangement playing and choose a repeat interval that complements the groove without landing distractingly on the next lyric.
• Set feedback low enough that the tail clears before the following phrase, then raise it only where the arrangement leaves room.
• Filter the return while listening to the lead vocal, preserving the intended texture without duplicating excessive sibilance or low-frequency energy.
• Automate the send for selected words and check whether the effect receives audio before, during or after any channel level changes.
• Listen in stereo and mono, then play the transition into the next section to check that the final repeats do not obscure its entrance.

When preparing the session for another mixer, label the return and keep its automation. If printing a wet reference, also retain the unprocessed vocal and describe whether any echo is considered part of the arrangement. This gives the next person a usable starting point without forcing an irreversible creative choice.

• Increasing feedback while monitoring loudly can create an unexpectedly persistent or rising signal. Test extreme settings at reduced level and watch the return meter.
• Adding more wet level to hear an echo in solo can make it dominate the chorus. Judge the balance in the section where the effect actually operates.
• Assuming tempo sync guarantees musical timing ignores swing, phrasing and empty space. A small timing offset may suit the performance better.
• Leaving a partly dry parallel return can duplicate the source and change its level. Confirm the wet control before adjusting the send.
• Confusing creative delay with monitoring latency leads to the wrong remedy. An intentional echo is not a reason to accept a distracting recording-monitoring path.

Treat an echo return as a mix element with its own level, tone and automation. Before adding another processor, briefly mute it and ask what disappears. If the answer is only clutter, simplify the timing or reduce the send. A purposeful single repeat can communicate more clearly than a long chain of competing effects.