HOW DOES A SYNTHESIZER PRODUCE SOUND

@

A typical synthesizer consists of the following main components : an oscillator, a filter, an amplifier, two envelope generators (volume envelope and filter envelope) and a low-frequency oscillator.  Each of these components is actually an electronic circuit that preforms a specific function by either generating or modifying a sound.  Each of these component is voltage or digitally controlled.  Their relationship is shown as follows:

@

@

An oscillator is an electronic circuit that generates a constantly repeating waveform.  This can be a square wave, a sawtooth wave, a sine wave or any one or combination of different waveforms.  In a synthesizer  the oscillator is where the sound originates in the form of electrical vibrations (or waveforms).  It is the beginning of the audio path.  From here the signal travels to other components where it is filtered, modulated, and then transformed into a finished sound.

The filter is an electrical circuit that filters out or blocks certain portions of the signals generated by the oscillator.  With it you can affect the brightness and overall timbre of a sound.  The filter accomplishes this by preventing all frequencies above a certain point from passing through the circuit.  We refer to this point as the cutoff frequency or the filter cutoff point.  When you cut out the upper frequencies (harmonics) of a sound you are in effect making the sound "duller".  As you open the filter (or raise the filter cutoff point), you let more and more harmonics through; and the result is a brighter sound.  The oscillator acting as a sound source generates a rich, harmonic-filled waveform.  This raw material is then routed to the filter, which proceeds to filter out the unwanted portions of the waveform allowing you to drastically modify the sound's timbre.

There are normally two envelope generators in a typical synthesizer.  They are exactly the same.  The difference is in their application: one is used with the amplifier to affect volume, and the other is used with the filter to affect the timbre.

One important way in which the sounds of various instruments are differentiated is by the differences in the shapes of the notes they produce.  By shape we mean how the volume of the notes changes over time.  For instance, a note played with a bow on a violin starts very gradually whereas a plucked guitar string sound immediately.  A note played on a clarinet will end abruptly as soon as you stop blowing whereas a note played by striking a triangle will sound for a long time and then fade out gradually.  We refer to the shape of the note as the volume envelope.

Working together, the amplifier and the envelope generator determine a sound's volume envelope.  The  amplifier is a circuit capable of increasing and decreasing the amplitude of audio signal.  This is controlled by the envelope generator.  The typical envelope generator, an ADSR, has four stages.  Attack, decay, sustain and release are the four parameters.  When you press down a key, you are sending two messages: one tells the oscillator which pitch to produce; the other instructs the envelope generator to begin its four-stage cycle.  It starts with attack, then decay, then sustain.  It will remain at the sustain level until you lift your finger from the key.  It will then continue with the release portion of the envelope.

The attack portion of the volume envelope refers to the very beginning of a sound.  It is defined as the time it takes for a sound to go from zero amplitude (no sound) to peak amplitude.

Decay is the portion immediately following attack.  It is defined as the time it takes for a sound to go from peak amplitude to sustain amplitude.  Many sounds reach a momentary peak amplitude during the beginning of the volume envelope, and then level off and remain at a fixed amplitude.

Sustain is defined as the portion of the volume envelope that remains at a fixed level for as long as the note is held or, for a synthesizer, the key is depressed.

Release refers to the final portion of the volume envelope.  It is defined as the time it takes for a sound to go from the sustain level to zero amplitude.  It affects how the sound fades out when you lift your finger from the key.

Just as the volume envelope determines the shape of a sound and how the amplitude of a sound changes over time, a filter envelope determines how the effect of a filter changes over time.  A filter envelope has the same parameters as the volume envelope - attack, decay, sustain and release.  When you set a gradual attack on a volume envelope, the beginning of the sound will fade in gradually.  When you set a gradual attack on a filter envelope, the harmonics determined by the cutoff point will fade in gradually.  With a filter envelope, you can then control the arrival and disappearance of harmonics over time.  You can create a sound that starts "dark" but grows increasing brighter, or vice versa.  We are now in fact superimposing two sound shapes, one over the other - the shape of a sound's timbre over the shape of the sound's volume.  This is one important way we have of adding internal motion to a sound.

A low frequency oscillator (LFO) is technically the same as an ordinary oscillator except for two important features: it oscillates as a lower frequency and it is used not as a source but as a modifier.  It's function is to modulate (regulate or vary the frequency of) the other synthesizer components.

When the waveform generated by the LFO, which is usually in a very low frequency - a few cycles per second, is used to modulate the output of the amplifier, we hear a periodic fluctuation in amplitude.  We get tremolo effect.  When it is used to modulate the oscillator we get vibrato.  Vibrato is defined as a periodic fluctuation in pitch.  When it is used to modulate the filter we get a modulated filter, in other word, a periodic opening and closing of the filter.

So, the oscillator generates a waveform; the filter changes the timbre of the waveform; and the amplifier along with the envelope generator determines the shape of the waveform's volume and timbre.  Finally, with the LFO, we are able to impart a little vibrato or tremolo onto our waveform.  A balanced combination of these will produce a sound you are looking for.

A more detailed description of the components will follow.

@

Click here to return to previous page