Sound Synthesis is the process of generating sound from a network of one or more analog or digital "circuits". Each circuit produces or modifies an audio signal which eventually reaches the end of the network where it is played and/or recorded.

The rest of this section describes the basic types of circuits used in sound synthesis.

Oscillators are usually one of the first circuits in a sound synthesis network because they are used to generate a raw repeating signal/waveform. Different types of oscillators produce a variety of signals including sine, triangle, sawtooth, square and random noise waves. Click on the illustrated waveforms below to hear what they sound like.

Sine	Triangle	Saw	Square	Noise

Basic oscillator waveforms represented visually and audibly.

Filters modify an incoming signal by cutting/removing a specific frequency band. They can be used to change a sounds thickness or brightness. Filters use what is called a cut-off frequency to define which frequencies are removed from the signal. Also, some filters provide an effect called resonance which can be used to specify the amount to increase the the frequencies just higher and lower than the cut-off frequency.

There are two main types of filters: low-pass and high-pass filters. Combining these basic filters together creates more complex filters: band-pass, band-elimination and peaking filters.

• A low-pass filter (LPF) removes frequencies higher than the cut-off frequency, allowing the lower frequencies to pass through.
• A high-pass filter (HPF) removes frequencies lower than the cut-off frequency, allowing the higher frequencies to pass through.
• A band-pass filter (BPF) uses a combination of both a low-pass and high-pass filter to allow only frequencies near the cut-off frequency to pass through.
• A band-elimination filter uses a combination of both a low-pass and high-pass filter to remove only frequencies near the cut-off frequency.
• A peaking filter emphasizes frequencies near the cut-off frequency which emphasizes a specific frequency band.

Envelopes are used to alter an attribute of the audio signal over time, such as the volume. They control the evolution of the sound from the time it is started through all of it's stages until the sound finally stops. One common type of envelope is called an ADSR envelope, which gets it's name from the segments of time it represents: Attack, Decay, Sustain and Release.

A visual representation of an ADSR envelope.

Low Frequency Oscillators, also known as LFOs, are used to modulate attributes of sound, such as volume or pitch to produce effects including vibrato, tremelo and wow. They work by modifiying other circuit's parameters, such as an oscillator's amplitude or a filter's cut-off frequency in a cyclic pattern. The cyclic pattern may be similar to the waveforms produced by oscillators, only using a much lower frequency/slower speed.

Now that you understand the basic elements of sound synthesis. You can choose from the following guides to learn about different types of synthesis that use some or all of these elements.

Books: Computer Sound Synthesis for the Electronic Musician, Sound Synthesis and Sampling, The Computer Music Tutorial