What is a Synthesizer?
Synthesizers are known primarily as the electronic keyboards that can produce a wide variety of sounds from the real world (e.g. piano, trumpet, bongo drums) and “beyond” (e.g. bleeps, sweeps, and UFO landings!). Synthesizers are actually a collection of sound creation and processing “modules” that can be interconnected in an infinite number of ways. Some of these sound manipulation modules can make sounds that are “impossible” in the real world, expanding the sonic palette for musicians, composers, and audio technicians.
By definition, a synthesizer is creating and modifying analog electronic sounds from scratch, as contrasted to “samplers”, which actually record small audio files that are, in turn, triggered by the keyboard.
Synthesizer "modules" don’t include a keyboard, and are triggered via MIDI
(Musical Instrument Digital Interface).
"Soft synths" are computer versions producing digital sound that can be used with computer audio and sequencing programs.
This is how the synthesizer creates a basic sound. The “oscillators” (sometimes called “tone generators”) output a "pulsing" (positive and negative) electrical signal to create a “waveform”. The "frequency" of the pulses (how many times a second it goes positive, negative, and then returns) determines the "pitch" of the sound; 440 pulses (or "beats") per second ("bps") would create an "A" note (above "middle C" on the piano). A simple oscillator can be set to create basic waveforms - the sawtooth-wave, triangle-wave, or square-wave, which help emulate different "timbres"; for instance, the difference between a trumpet and a xylophone. Most synthesizers can also create a completely random waveform - a noise wave. Naturally occurring “acoustic” sounds generally derive from smooth sine-waves which are harder to create from scratch. Sine-waves (think "flute") can be simulated with some extra manipulation, by mathematically “smoothing” a triangle-wave.
To be able to vary the basic waveforms to some extent, most synthesizers use “filters”. A filter is an electronic circuit, which works by selectively increasing or decreasing specific frequency regions, effectively smoothing out the “edges” of the original waveform. If you listen to a sawtooth wave, you will find it rather “harsh” - because the waveform has a lot of sharp “edges”. These edges are in the acoustic terminology called “overtones”. Now, if we apply a filter to the waveform, some overtones can be removed - and the waveform will become more “rounded”. The more overtones are being removed, the more “rounded” or “dull” the sound will become. Such a tone without any overtones is called a sine-wave. A sine-wave is the simplest naturally-occurring waveform: it has no edges and sounds like a flute.
To control the volume of a sound in the synthesizer, the signal is passed through an “amplifier” circuit. The amplifier can raise, or lower the height of the waveform, thereby raising or lowering the sound volume. So basically what happens when the user presses a key on the keyboard is that the oscillator creates a waveform. This waveform is filtered through the filter and is finally amplified by the amplifier. Also seen on synthesizers as “VCA” (voltage controlled amplifier), it can be applied to other synthesizer elements in varying degrees; e.g. to layer oscillator and filter volumes.
Now we have the minimal requirements to create a synthesizer voice. But we still need to make this sound playable as a music instrument voice. Natural sounds don’t normally switch simply on and off. Sounds are hardly ever “static” and change their character through time. A real life sound has a fade-in and fade-out period. To take an example, a drum hit begins very sharply as the drumstick hits the skin and also fades away quite fast. The sound volume of a note on the piano will also rise rather quickly, but will dampen much more slowly. Many sounds also change their pitch between the onset and the fade-out, like how the pitch of a train whistle drops as the train passes by. This behavior is called the “envelope” of the sound. To simulate this effect on a synthesizer, we must be able to control the oscillator, the filter and the amplifier in a much more detailed way than just our basic on/off function. We want to be able to control the “envelope” for each synthesizer circuit. The envelope of the oscillator controls how the pitch changes through the duration of the sound and the envelope of the amplifier controls how the volume changes over time. The filter has also its own envelope, which controls the changes in the “brilliance” or “dullness” of the sound. The simplest synthesizers envelopes have four stages. This means that they control four different parts of the envelope. These different parts of the envelope curve are termed “ADSR”:
All these different stages of the envelope can be set individually. The envelope generators of some synthesizers have many envelope stages, which makes it possible to create some very complex envelopes and interesting sounds.
Frequency Oscillator (Modulation)
Most synthesizers have a wheel controller to the left of the keyboard labeled, "Modulation". This enables the performer to mix in a very slow oscillation to a sound, creating “vibrato” or "filter sweep" effects.
Well, now we have a fully usable synthesizer voice. But we still have only one voice we can’t play chords yet. The first synthesizers were all like this - they are “monophonic” synthesizers. If you wanted to play chords, you would have to buy more than one synthesizer - one for each tone.
By adding more oscillators, filters, and amplifiers in the instrument, the synthesizer can play more than one sound at once. A synthesizer with eight voices can play eight tones at the same time. This may seem enough, but a modern synthesizer is usually equipped with 32, 64, or even 128 voices. One could ask, why anybody would need an instrument which can play 64 voices at once - nobody in the right mind would want to play (and listen to) huge 64-note chords.
Well, first of all, voices can be stacked on top of each other to create a more complex sound. A single key depression could for instance play a piano voice, a string voice and a choir voice at the same time. It’s easy to see that the available voices are quickly gobbled up when playing with such a complex voice stack. But the main reason is the ability to let the synthesizer play more than one part at once in a musical piece. Nearly all modern synthesizers are “multi-timbral” instruments, meaning that the synthesizer can play several different sounds at once. If the synthesizer is connected to a computer, it can for instance play drums, strings, brass, bass and guitar parts - all at once, like a big one-man-band.
Some hybrid synthesizer/samplers also contain banks of audio samples or “waveforms”, which are digital audio sound clips that may be used instead of a simple oscillator tone. The synthesizer modules (filters, envelopes, LFOs, etc.) may then be applied to create an infinite variety of sounds. Samplers can also record any natural sound with an attached microphone, which may then be converted and used as a waveform/sample. Wavesample voices and oscillator voices can also be combined to create amazing, complex sounds.
Appendix: Where to Find Synthesizer Stuff
Here is a list of synthesizer-related links:
"The Synth Museum" http://www.synthmuseum.com/
"Keyboard Player Magazine" http://archive.keyboardonline.com/index_old.htm
"Shareware Music Machine" site: www.hitsquad.com/smm/ .
© 2002 by Paul Alan Smith