When we talk about delay as an effect, it isn't just a matter of delaying a signal. The delayed signal is added to the original, normally at a lower level. So you get the original signal followed by a quieter echo, a fraction of a second later.
But you don't have to use it. Why not just keep the vocal nice and clean, as it was recorded?
There are three reasons to use a delay…
- You want to imitate a sound you have heard already – the 1950s 'Elvis' sound for instance.
- The vocal needs to be richer
- The vocal needs to be more interesting.
Using delay to imitate the 'slapback' sound of the 1950s is probably a little too straightforward to require explanation. Wanting a vocal to sound richer is commonplace. The best solution for a richer vocal is a richer-sounding singer, but too often we don't have that option. A subtle delay will warm and enhance the vocal, hopefully without sounding too obvious.
A richer vocal sound might also be useful in helping the vocal blend with the track. If you find that the vocal seems to sit on top of the track, then a delay at a low level might work wonders. Worth a try.
If the vocal needs to be more interesting, then a delay might just do that. But you don't always need the delay to go the whole way through the track. You might just switch it in every now and then. In the 1980s it was quite common to switch in the delay only at the ends of certain lines. (Hint – switch the send, not the return.)
There's a lot more to delay than this, but knowing when to use it is a start.