Delay or echo is one of the oldest artificial effects. A wide variety of musical styles use delay in many different ways. At it simplest it is one or more repeats of a sound played back later in time. We are all familiar with echo in the natural world. It occurs in large spaces enclosed by hard surfaces such as canyons or cathedrals. Perhaps for this reason it has a nature association with the epic and ethereal. It can bring this and much more besides to a musical production and has become an integral part of some musical styles such as dub, trip hop and trance.
Setting a musically relevant delay time
In a musical context, delays are generally required to synchronize in some form to the tempo of the piece so that they enhance rather than clutter the mix. By adding delays one is effectively adding notes to the music but at a later point in time. So you need to take care to ensure that delays that continue over subsequent chord changes don't create unwelcome clashing notes.
If the music has a fixed tempo it is easy enough to calculate the delay time. There are 60 000 milliseconds in a minute Sc you know the number of beats (ie quarter notes) per minute you divide 60 000 by the tempo to find the duration in milliseconds of one beat. If you use this as your delay time you will have a quarter note delay. You can divide this by 2 ft an 8th note delay time, by 4 for a 16th note delay time and so on.
Here are the figures for 120 bpm.
1/2 note delay - 1000
1/4 note delay - 500
1/4 note triplet delay - 333.33
1/8 note delay - 250
1/8 note triplet delay - 166.67
16th note delay - 125
These days however most delay or echo plug-ins do the hard work for you. They are able to synchronize to the tempo of the session (set in the software) and allow you to se a delay time from a range of note values. Be aware that the longer the delay time you choose the more likely you are to create clashes with subsequent musical phrases.
Artificial delay first became possible with the advent of tape recording. As noted in the previous chapter, a . machine with a play head after the record head could be used to create a slap back delay which was used to emulate if early reflections present in reverberation. Feedback or "spin" was created by sending the return from the play head bad the record head mixed with the unaffected sound.
This was later refined with units such as the classic Roland RE-201 which first appeared in 1973 and worked by means of a constantly circulating tape loop. A sound would be recorded onto the tape played back from several spaced play heads and then erased when the tape came round again. This unit included a spring reverb and had controls for mode, repeat rate, intensity and wet/dry mix amongst others. Adjusting the repeat rate controlled the delay time, the mode controlled which play heads were active and the intensity controlled the feedback i.e. the number of repeats and the amount of self-regeneration. As well as being used as a basic delay this unit could create some pretty strange effects, it had a warm sound and could be rather unpredictable which, was also part of its charm. It can be heard on numerous classic recordings from the 70s, 80s and 90s from Bowie to Portishead via dub gae, The RE-201 is still much sought after even today. Universal Audio now do a software emulation of the RE-201 but ething like Logic's tape delay which is more readily accessible to many of us can get close to recreating many of its nds. The Pro Tools user can choose from some sophisticated plug-ins like the Line 6 Echofarm and the amazing hoboy by Soundtoys both of which emulate all the classic tape delays.
Tape delays generally tended to have a very poor frequency response which in fact helped to give them their warm sound. is side effect also helps to distinguish the delayed sound musically from the source sound and to some degree reflects at happens in nature. Real echoes are naturally duller and quieter than their source sounds since the sound waves lose ergy as they travel through the air and bounce off reflective surfaces. Psycho-acoustically speaking, the brain is more ely to recognize a delay as such (rather than as a new note in the wrong place) if it has some high frequency roll-off. So plying this principle during mixing can often help to ensure that delays don't make the mix sound cluttered. Tape delays re also infamous for their dodgy transport systems but again this was part of their charm. Provided they weren't too onounced slight variations in the playback speed would throw up interesting flutter irregularities in the sound. Logic's tape lay has a flutter section to emulate this aspect of tape delays and there is also a high cut filter to remove those offending pend frequencies. As you can see the user can sync the delay time to the session tempo and choose from a selection of te values. You can modify these to create dotted or triplet note values by moving the groove slider fully to one end or the er. If you move the feedback control higher than 50% the sound will start to self regenerate. Such controls are fairly typical of most delay type plug-ins.
Ping-Pong and Stereo Delays
A stereo delay is, in essence, two mono delays in parallel and panned hard left and righ In the case of a mono send to a stereo delay the input signal would be the same on both sides. Each side can send feedback to the other side as shown in the diagram below. This allows for the creation of some interesting effects since can set musically relevant but different delay times for the left and right channels and of course differing amounts of crossover feedback.
In a ping-pong delay the signal would arrive only at one input. Having passed through the delay it would then be fed acro to the other side and so on. If the delay times are the same this creates an even bouncing effect from side to side. If the delay times are different then a myriad of interesting configurations are possible. Logic Pro's Stereo Delay plug-in has a fairly typical set of parameters with separate wet/dry controls for the left and right channels, a high and low cut filter sectii the usual delay time selection options, crossover feedback for each side and internal feedback for each side.
The term multi-tap originates from vintage tape delays like the RE-201 which had more than one playback ad. A delay was considered to be "tapped off' from the signal at each playback head. This type of delay creates a huge rnber of possibilities for manipulating the sound. These fall into two categories:
- If each tap output is panned at even spacings across the stereo image and is set to a sequential and musically relevant delay time (1/2 notes, 3/8th notes, 114 notes 3/16th notes, 8th notes etc) interesting musical delay configurations can be created.
- More commonly though multi-tap delays are used with short and slightly random delay times (say 50ms-350ms) with the taps panned at alternating positions across the stereo image and gradually getting further from the centre. The ear then doesn't necessarily distinguish the taps as separate echoes and so you get a spreading, thickening and reinforcing type effect. If a little feedback is sent from each tap the repeats quickly build up in complexity creating a coarse reverb-type sound. This can sound really effective on solo electric guitar passages for example. Below is an example of this type of configuration with DSP FXs multi-tap delay plug-in.
Multi-tap delays are also often used as the basis for voice-doubling type effects. An 8 tap delay such as the one above Id work very well with extremely short delay times and panned voices to create thickened and widened vocal sound.
Since delay effects can easily make the mix sound cluttered and busy they make a good candidate for spot For example if there is a point where the music pauses it can sound great to have a delay on the last word of a vocal at tails off into the space. This can be achieved in a number of ways. One is to make a copy of the phrase or word to be eyed on a new channel and set up the delay as an insert on that channel, setting the mix to 100% wet. Another is to put bypassed version of the delay insert on the original channel (this time with a combination of wet and dry in the mix) and le the automation to un-mute it when you want it to activate. Another is to set up the delay as a send and automate the nd so that the delay only receives an input at the desired point in time.