Other Synthesis Methods

Sample-Based Synthesis

Sample-based synthesis, which is sometimes known as Pulse Code Modulation (PCM), or sampling and synthesis (S&S) synthesis, is differentiated from subtractive synthesis mainly by the use of samples in place of oscillator waveforms.

The samples—digital recordings of existing sounds—are mapped across the keyboard. Typically, each sample is mapped to a note in the center of a keyboard range that spans 5 or so notes that are unique to that sample. The reason for this range of 5 or so notes is that samples tend to sound much less like the source sound if played more than a few notes higher or lower than the original pitch—due to the relationship between the pitch and playback speed of samples.

The pitch of each sample isn’t changed with a frequency control, unlike the oscillator waveform of a synthesizer that is not sample based. Rather, a sample is played back at a faster or slower speed to alter its pitch, which has a corresponding impact on the sample playback time. For example, a sample played back at twice the speed requires half the time to play through.

The EXS24 mkII is a sample player that can be used much like a sample-based synthesizer, due to the subtractive synthesis facilities that it offers.

Popular instruments that use this synthesis approach include Korg’s M1, O1/W, and Triton; the Roland JV/XP instruments; Yamaha’s Motif series; and many others.

Frequency Modulation (FM) Synthesis

Put simply, FM synthesis involves the use of a modulator oscillator and a sine wave carrier oscillator. The modulator oscillator modulates the frequency of the carrier oscillator within the audio range, thus producing new harmonics. These harmonics are known as sidebands.

Typically, FM synthesizers don’t incorporate a filter. You can generate some subtractive synthesizer style sounds with FM synthesis, but it is difficult to recreate the sound of a resonant subtractive synthesizer filter with this method. FM synthesis is extremely good, however, at creating sounds that are difficult to achieve with subtractive synthesizers-sounds such as bell timbres, metallic tones, and the tine tones of electric pianos. Another strength of FM synthesis is punchy bass and synthetic brass sounds.

Logic Pro includes a simple FM synthesizer, the EFM1. Although it is minimalist, it is capable of producing many of the classic FM sounds made famous by Yamaha’s DX series of synthesizers (the DX7, sold from 1983 to 1986, remains the most commercially successful professional-level hardware synthesizer ever made).

The ES2 also features some FM techniques that allow you to modulate one oscillator with another. You can use these FM techniques to partially bridge the gap between the very digital sound of FM synthesis and the fat analog sound that the ES2 is noted for.

Component Modeling Synthesis

Also known as physical modeling, this synthesis method uses mathematical models to simulate instruments. Parameters are used to describe an instrument’s physical characteristics, such as the materials the instrument is made of, the dimensions of the instrument, and the environment it is played in—under water, in air. Equally important are descriptions of how the player would interact with the instrument—for example, whether it is played by plucking, bowing, or strumming strings; by hitting it with sticks; by placing fingers on sound holes; and so on.

To model a drum sound, for example, the following aspects would need to be taken into account. Of primary importance would be the actual drum strike—how hard it is and whether the drumhead is struck with a wooden stick, a mallet, a beater, and so on. The properties of the drumhead (the skin or membrane) would include the kind of material, its degree of stiffness, its density, its diameter, and the way it is attached to the shell of the drum. The volume of the drum cylinder itself, its material, and the resonance characteristics of all of the above would need to be mathematically described.

To model a violin, you’d need to take into account the bow against the string, the bow-width and material, the bow tension, the string material, the string density, the string tension, the resonance and damping behavior of the strings, the transfer of string vibrations through the bridge (materials, size, and shape of the bridge), and the materials, size and resonance characteristics of the violin body. Further considerations would include the environment that your modeled violin is played in, and the playing style-“hammering” or tapping with the bow as opposed to drawing it across the strings.

The Sculpture component modeling synthesizer is capable of producing convincing recreations of acoustic (and electronic) instruments. It is also exceptionally good at creating atmospheric, constantly evolving pad sounds. Other included instruments that incorporate physical modeling components and techniques are Ultrabeat, the EVP88, EVB3, and EVD6.

Wavetable, Vector, and Linear Arithmetic (LA) Synthesis

Wavetable synthesis uses a number of different single-cycle waveforms, laid out in what is known as a wavetable.

Playing a note on the keyboard triggers a predetermined sequence of waves. In general, this is not a stepped transition but rather a smooth blend from one waveform into another, resulting in a constantly evolving waveform. Multiple wavetables can also be used simultaneously—either played one after the other, or blended together—resulting in more harmonically complex waveforms.

A single wavetable can emulate filter cutoff with a series of bright, less bright, then dull-sounding waveforms played in sequence—which resembles a reduction of the filter cutoff frequency in a subtractive synthesizer.

Wavetable synthesis isn’t particularly successful at emulating acoustic instruments. It is, however, extremely successful at producing constantly evolving sounds; harsh and metallic, or bell-like sounds; punchy basses; and other digital tones.

Wavetable synthesis was championed by the PPG and Waldorf instruments. The ES2 also includes wavetable facilities.

Roland LA (Linear Arithmetic) synthesizers such as the D-50 work on a similar principle. In these synthesizers, however, complex sampled attack phases are combined with simple sustain or decay phases to create a sound. In essence, this is a simple wavetable that consists of two samples.

Where LA and wavetable synthesizers differ is that the latter were designed to create new, original, digital sounds. LA synthesizer designers, in contrast, wanted to emulate real instruments using a minimum of memory. To facilitate this, they combined samples of the attack phase—the crucial part of a sound—with appropriate decay and sustain phases.

Vector synthesis—used in the Sequential Circuits Prophet-VS and Korg’s Wavestation—allows you to move through wavetables and sequences arranged on a two-dimensional grid (two different vectors, or less technically, on the X or Y axis). The main benefit of this approach is that the balance between samples and waves is achieved in real time by moving a joystick. You can also use the ES2 to perform vector synthesis by modulating the Oscillator Mix (Triangle) parameter with the Vector Envelope.

Additive Synthesis

Additive synthesis could be considered the reverse approach to subtractive synthesis. See the beginning of this appendix, including the discussion about all sounds being a sum of various sine tones and harmonics, for background information to provide insight into additive synthesis.

In essence, you start out with nothing and then build up a sound by combining multiple sine waves of differing levels and frequencies. As more sine waves are combined, they begin to generate additional harmonics. In most additive synthesizers, each set of sine waves is viewed and used much like an oscillator.

Depending on the sophistication of the additive synthesizer you are using, you will be provided with individual envelope control over each sine wave, or you will be limited to envelope control over groups of sine waves—one envelope per sound and its harmonics, or all odd or all even harmonics, for example.

Logic Pro doesn’t offer a true additive synthesizer, but aspects of the additive synthesis approach are used in the EVB3 and all other drawbar organs. In the EVB3, you start off with a basic tone and add harmonics to it, to build up a richer sound. The level relationships between the fundamental tone and each harmonic are determined by how far you pull each drawbar out. Because there’s no envelope control over each harmonic, however, the EVB3 is limited to organ emulations.

Phase Distortion Synthesis

Phase distortion synthesis creates different waveforms by modifying the phase angle of a sine wave.

In essence, you can bend a sine wave until it becomes a sawtooth wave, a triangle wave, a square wave, and so on. The synthesizer engine beyond the waveform generation in general follows the standard subtractive method.

Phase distortion synthesis was commercially introduced in 1984’s Casio CZ series synthesizers.

Granular Synthesis

The basic premise behind granular synthesis is that a sound can be broken down into tiny particles, or grains. These sampled grains—usually no more than 10 to 50 ms long—can then be reorganized, or combined with grains from other sounds, to create new timbres.

In many respects, this is much like wavetable synthesis, but it works on a much finer scale. As you might expect, this method is ideal for creating constantly evolving sounds and truly unique tones.

The downside is that granular synthesis is extremely processor-intensive, and it wasn’t possible to do it in real time until relatively recently. For this reason, it has remained largely ignored by all but a few in academic institutions. Today’s computers, however, have sufficient processing power to make this synthesis method a practicality, and there are a number of commercial products now available.