2 Signal processing

2.7.2 Time

A lot of different effects can be obtained by recording a signal, storing it for a period of time, and then playing it back. The recording time can vary from .001 seconds to a few seconds, or actually virtually infinite, in which case you have a recording device - hardly a sound effect. With very short recording times, time based effects begin to overlap with frequency based effects.


Delay is best described as "echo". Back in the early days, analog delay machines consisted of a tape recorder with a circular (i.o.w. endless) tape. The signal was recorded onto the tape, spun round and played back just before being recorded again. Delay time could be varied by changing the tape speed or length. This type of delay gave a characteristic sound, mainly due to wear and saturation (overdriving) of the tape, comparable to tube saturation. For some time, solid state analog delays have been around. They apply a socalled bucket brigade delay (BBD). An analog memory array of up to a few thousand cells is constantly cycled. Some compact chorus and flanger pedals still use this technology.

Repeating, fading echoes can be created by feeding the attenuated output of the delay back into the input ("feedback").

A delay is the most basic application of a digital processor. The first fully digital sound processors were digital delays (Lexicon were the first). A simple digital delay doesn't require a DSP (digital signal processor), which were incredibly expensive at that time. Due to the nature of digital systems, storing and recalling information without loss is a piece of cake. If we simply continuously record, store, and playback, we have an outstanding delay. Maybe too good. That's why nowadays, modern delays have some sort of built-in signal degradation to make them sound more like good-ol' tape machines, only 10 times as cheap, and with much less noise smiley


A reverb creates the sense of room acoustics and reverberation. The spring reverbs of old are still quite popular, because of their low cost and often excellent results. More modern reverbs use digital technology: the signal is input in a mathematically created virtual room, and the sound of the "room" is sent out. Before life-like reverbs like these were possible, the only way to get the desired reverb was to actually record in a room that had that desired sound. Many close harmony groups recorded their LPs in bathrooms.

Digital reverberation requires very complex mathematical functions to be performed by the DSP. In section 2.5 2.5 Analog - digital, there was mention of "a trade-off between sound quality and capability". Because reverb requires a fast and accurate DSP, only costly reverb units that meet those requirements, provide life-like reverb. A good example is the legendary Lexicon PCM-90, which uses all of its processing power, just to create reverb. Inexpensive units, and especially most multi effect units, may sound artificial and cold, and may lack definition.

Reverb is rarely used for bass.

Chorus and flanger

A chorus is a short delay (5-50 milliseconds) with its delay time slightly varied over time ("modulation"). This creates a light shimmering effect, because the pitch of the delayed signal is constantly changing. The delay creates a doubling effect, while the modulation makes it seem as if two different instruments are playing the same notes, instead of just one original and an exact copy.

A flanger is essentially the same device, but part of the delay output can be fed back into the input (just like a normal delay), causing a phaser like sound. At extreme settings, a flanger can sound like you're playing in a pool or a rotating tunnel. When the feedback control of a flanger is set to 0, you have a chorus device.

Flangers used to be created by applying finger pressure to the flange of the reel of a tape delay machine. Hence the name "flanger".

© Joris van den Heuvel 2001-2009