The Delay
This article comes from the wonderful Ziggy's sonorisation site. It is reproduced here, of course, with the kind permission of Ziggy himself ;o) We really advise you to go to this site, where you will find other articles that don't have their place here, especially concerning microphones, mixing boards, amplifiers, etc.
Voice, etc.
Special use
References
Delay or echo has to do with the spatialization
of sound. Indeed, reverb and delay are basically the same thing, i.e. sound
reflections. The main difference is that the delay (or echo) corresponds to
one or several precise and distinct reflections, whereas reverb is the result
of mixing many various reflections, which we cannot hear distinctly (see
the excellent report made by Vincent Burel
on this subject).
When mixing, delay often replaces
reverb because it basically creates the same effect, that is to say a
sound space around the main sound, but without adding superfluous sounds
(as it is the case with reverb).
Years ago, tape recorders were used to create echoes.
Later, heads were added to get more control over the number of repetitions and
the speed. Jimmy Page, Hendrix and Ritchie Blackmore have been innovators in this
domain. Nowadays, digital technology has replaced analog delays.
Settings that can be found on a delay processor are:
-
Type: gives a choice of several delays:
-
MONO-DELAY – produces one or several repetitions of the sound in MONO
-
STEREO-DELAY - produces one or several repetitions of the sound in Stereo. Most of the time, Time and Feedback parameters can be set independently for left and right channels.
-
PING-PONG – produces repetitions panned from left to right and vice versa.
-
TAP-DELAY – allows you to freely locate sound repetitions on a time basis as well as in the stereophonic field (PAN).
-
SAMPLING or HOLD records a phrase that you can then loop. Some machines even allow to play it backwards.
-
-
Time: sets the delay time. Generally, values are set in milliseconds (ms), but sometimes you can also choose other values as feet, meters, or BPM, which is very useful to set the delay on a given tempo.
-
Feedback: with this parameter you can set the number of repetitions.
-
Tap: by pushing this button following the beat the song, the delay time will automatically fit the tempo (sometimes a foot switch can be plugged in to set the tempo by foot).
-
MIX or LEVEL: enables to mix the initial sound (dry) with the delay (wet).
-
LOW-PASS or HIGH DAMPING: concerns the delay high frequencies. We use this parameter to simulate analog delays or tape-delays which were known to have difficulties to reproduce frequencies above 8kHz, but which gave warmth to the sound at the same time
-
RATE and DEPTH: many delay devices have a small modulation processor (chorus etc.), which colours the repetitions. With the Rate and Depth parameters, you can set the speed and intensity of the modulation.
Some examples of use
If the reverb is important to emphasize the colour of an instrument or to create a coherent sound space in a mix, it is easy to spoil a mix by adding too much reverb or by destroying the sound space with an incoherent reverb! Too much reverb makes listening tiring et blurs the mix by drowning the instruments in diffuse and omnipresent sound-trails. That's why it is often replaced by a delay.
General
A delay can be applied to all instruments
(be careful with rhythmic instruments, it remains very delicate). Generally,
the delay foolows the tempo (crotchet, half note, quaver etc.) of the
song in order to not disturb the beat.
If the device does not have a TAP function, or
a BPM setting, we will have to calculate these values. It is very simple if
we know the song tempo (rhythm box or MIDI): we know that one minute contains
60,000 milliseconds and the tempo is given in number of crotchets per minute.
We just have to divide 60,000 by the BPM to obtain a value for a crotchet in
milliseconds (ms).
Example: tempo = 120; the delay time will be of
500 ms to prop up on the crotchet, and 250 ms to prop up on the quaver (two
quavers are a crotchet). Otherwise, we will have to find the tempo with a rhythm
box or a metronome. It is always better to prop up the delay (even for small
values) on the song tempo. Everything will sound more coherent. Anyway, this
also applies on the modulations speed setting (phaser, flanger, etc.). Try more
"exotic" values too, such as triplets for example, which sometimes
give surprising results.
Always set the number of repetitions to the strict minimum to obtain the effect you want, because delays that trail in the mix quickly become rather noisy.
Voice
Voice takes nearly always need a delay. Most of
the time, it will be set in stereo spread, that is to say that, in addition to
the voice in the center of the mix, we will add two delays onto the left and onto the
right (for instance 20 and 30ms, or 40 and 50 ms) with a single repetition.
This will give a good stereophonic image.
Another famous effect is the John Lennon-like echo.
It is a 120 ms delay with about 4 repetitions, with are set
fairly loud.
A delay setting which simulates well a reverb but without sullying or drowning the sound is a mono or stereo 200-ms delay with three or four repetitions, and which will be mixed behind the original sound.
Guitar
Guitarists are those who use the most the
delay. For a lead sound, the delay is often set at about 300 ms with three
to five repetitions. It can also be set to the crotchet or half note,
which is better to support the beat of the song.
A very spectacular effect consists in setting the delay
to half-quaver, with a single repetition, and to play a riff or a
phrase in quavers ! (try the triplets too !). For a rockabilly sound, a slap-back
delay style is used, with a delay-time of about 120 ms.
A rather short (20 to 60 ms) mono or stereo
slap is also used to emphasize a rhythm guitar. You have to know that
a delay of less than 20 ms is not perceived as a distinct reflection, but
mixes with the original sound and gives the impression of a fatter sound.
A studio tip is to pan the original sound to 8 O'clock and a 20 or 40-ms delay sound to 4 O'clock, which gives an stereo impression.
Drums et percussions
On drums and percussions, we tend to use reverbs rather than delays, because the delay destroys rhythmic figures too easily. The exceptions are Reggae and Pop slow-motion songs, where a delay is often used on the snare drum. Repetitions are then set to the crotchet, or most of the time to the quaver (reggae) and feedback to three or four repetitions.
Special uses
A
delay can be used to coordinate several systems in big places (stadium, open
air etc.). We know that the sound covers 1080 feet per second, that is to say
a spectator which would be 1080 feet far from the scene would hear the sound
a second later than his friend who is in front of the scene. This phenomenon
is well known, when we can see the drummer hitting the snare drum and we hear
the sound later. If, in order to cover the whole stadium, a second system is
installed 1000 feet far from the scene, the spectator who is there will hear
the sound at the same time it is produced on the scene, but, a second later,
he will hear the same sound again (coming from the first system). Bloody shambles!
If, on the other hand, we delay the sound of the second system in relation to
the main system (one second in our example), everything goes normal and becomes
coherent. We often use a little delay (some milliseconds) for the return system.
This considerably reduces larsen risks, because the sound-micro-speaker loop
is interrupted by the delay.
A wide domain of use of the delay remains the modulation (flanger, chorus, phaser, etc.), because these effects are basically delays which are dynamically processed. For example, a chorus is produced by several very short delays which delay-time constantly changes.
The flanger is basically an echo too, which was produced by a tape magneto that was speeded up and down to create this turbo-space effect. You can also approach this effect by setting the delay-time very short (2 to 6 ms) and increasing feedback. Then you play on these two parameters at the same time.
Conclusion
Nowadays, there is a big polemic regarding digital delays in relation to analog ones. I remain sure that with a good digital one you can almost do everything - even simulate any analog or tape device and its so-called advantages (warmth of the sound etc.) of these analog devices do not justify anymore the problems that come with them (noise, frequencies reduction, tape echoes mechanical noises etc.)
Try the Line 6 DL-4 for the guitar and even Lenny Kravitz will freak out!
References
Today, all the digital delays are very good, and you will have to look for for a long time to find a crappy stuff. Nevertheless, small foot pedals and rack processors have to be distinguished. The first ones are generally made for the guitar, and the frequencies efficiency is not the same (trebles are less present). Indeed, there are less parameters than on rack processors, but they are half cheaper. For a home studio, a small pedal can be useful. For a serious work, buy a processor (specialized in delay, or multi effects!). Specialists such as the Line 6 DL-4 (guitar) or the TC ELECTRONICS D-Two (studio) integrate many modulation effects and many settings possibilities in relation to the multi-effects, where the delay is just an effect among others!