Saturday, March 17, 2012

Synth Chronicles 2: How To Make Sound

In the last part of this series, I talked about the basics of how analogue synths tend to function. Here, you will learn that there are SO MANY other ways to make sound electronically.

Additive - additive synthesis produces sound by adding sound waves together. These waves are usually Sine waves, but can be anything. This is just like that model you saw in school: if you have two waves that have matching peaks and troughs, the signal is twice as loud. If you move the second waveform just a little, it can have drastic effects on the output wave. How does this apply to music? consider other musical instruments. Their distinctive sound, or timbre, can be considered as the child of the of Fourier theory which states that sounds "consist of multiple harmonic or inharmonic partials," or overtones. Every partial used is a wave of a different pitch and amplitude. Additive synthesis produces sound by combining, or adding as the name suggests, the output of multiple wave generators, or oscillators.

Digital - digital waveguide synthesis is the synthesis of audio using a given template which mimics how acoustic waves are shaped. Digital waveguides are part of most modern synthesizers to produce realistic tones. You could think of it as a highly advanced filter which takes a normal waveform and makes it come to life. This can or can not also include the envelope that makes up how sounds naturally decay dynamically.

Distortion - Distortion synthesis is actually a group of synthesis techniques which changes existing sounds to produce more complex timbres. Before the discovery of distortion synthesis, complex sounds were created using many oscillators and modifying each oscillator's parameters. Distortion methods can form a complex frequency spectrum which is not only more efficient, but you can also separate and adjust parameters for various frequencies. Among others, a noteable example of distortion synthesis is FM or Frequency Modulation synthesis.

Frequency modulation - this form of synthesis distorts the 'carrier' frequency of an oscillator by modulating it with another signal, similar to how a vocoder works, if you are familiar with that. The 'distortion' of the first oscillator can be controlled in accordance with the amplitude of a modulating signal. This means that to alter the first signal, you must modulate it with another signal with a varying volume and pitch. For synthesizing 'pleasing' sounds, the modulating signal must have a harmonic relationship to the original carrier signal, meaning a pitch that would sound good with the first, assuming you played them as a chord. As the amount of modulation increases, the sound becomes more complex. To produce percussive sounds or noise, you can use non-harmonic modulators. Bell-like, dissonant (or chorused), and percussive sounds can easily be created by doing this.

Sample-based - this method of synthesis uses given, or seed, waveforms. These can be recordings of sounds or externally generated waveforms. You may ask how this is synthesis if it is just playing back sounds. Well, the synth using said sample interlaces the sample to cleanly loop, that is play repeatedly with no cracks or other disruptions in the audio. The synth can also have built-in filters to alter the sample.

Granular - Granular synthesis is a simpler method that uses the microsound time scale, and is based on the same principle as sampling. However, the samples are not played back normally. They are split into small, quick pieces, normally at a maximum of 100 milliseconds. These small pieces are called grains. Grains may be played as single parts like a sample, or layered to form a kind of additive synthesis. You can play grains at different speeds, phases, volumes, and pitches, too. A granular synthesizer can synthesize or remove parts of the sound to play and extremely slow or fast speeds respectively. At slower speeds, you get soundscape sounds, which are like drones or general background noises. At faster speeds, however, you can hear actual notes instead of noises. You can change the sample, envelope, overall length, panning, and density of the grains to produce some awesome sounds. Random neat thing; you can also sample a granularly synthesized sound and use that as a sample to re-synthesize.

Phase distortion - This form of synthesis is similar to FM synthesis, but this method dynamically changes the harmonics carrier waveform by applying another waveform, or modulator. This produces composite waveforms whose harmonics are both the sum and difference of the carrier and modulator waveforms. Fun fact: This was invented (or at least first used) by Casio. Yes, the same people who make calulators.

Subtractive - This method uses partials of an audio signal, usually of rich harmonic quality, and attenuates them with a filter to alter the timbre of the sound. This was pretty much what analogue synths of the 60s and 70s used.

Vector - This form of synthesis, invented by Sequential Circuits, namely the ProphetVS synthesizer, uses the fading between four sound waves using a single point on a vector plane as its synthesis technique. The mixing, if you will, was performed using a joystick, but could be automated using envelope generators or LFOs.

No comments:

Post a Comment