Understanding the Basics of Sound Synthesis
Many of those who use synthesizers in the modern technological world are not well versed in the basics of different kinds of synthesis. With the ease of computer-based synthesis, any synthesis program can be opened and fiddled around with by ear until something “cool” comes out. To break the mold in the use of synthesizers in the modern music world, you have to be educated or lucky. This is exactly why it is more important now than ever before to understand the fundamentals of different synthesis methods.
Subtractive Synthesis and Its Related Synthesis Methods
This is the most common method that gave birth to the concept of sound-synthesis.
Subtractive Synthesis is a very simple signal chain of an oscillator (sound source) running through a filter (EQ curve) which is then sent to an amplifier for gain staging and ADSR control. This method is very easy to achieve in both analog and digital realms and can be used to create numerous (possibly infinite) instruments, effects, and sounds.
The main principle behind Subtractive Synthesis is that any harmonic character can be constructed by an oscillator, or the combination of multiple oscillators. Then, by running these oscillators through various filters, and controlling the envelope response, the harmonics present within the oscillators tones can be whittled into harmonic structures that mirror those of actual instruments.
The analog subtractive synthesizer was initially designed for this purpose–as an alternative to hiring musicians to play on recordings, however, it quickly morphed into its own instrument, creating various sounds never before made by any acoustic instrument.
Additive Synthesis (and Resynthesis)
Additive Synthesis is trying to achieve the same result as Subtractive Synthesis, but approaches the method from a constructive philosophy, rather than carving. Rather than presenting a wall of harmonics and carving out the harmonic structure desired (Subtractive Synthesis), in additive synthesis multiple sine waves of varying levels and frequencies are combined together to build the harmonic structure desired. Simply put, instead of starting with everything you need and throwing away what you don’t need, you start with nothing and build harmonic structures from scratch.
The very connected process of Resynthesis is highly connected to Additive Synthesis. In essence, Resynthesis involves analyzing the harmonic structure of a sampled sound, and trying to recreate that structure. Additive Synthesis is essentially Resynthesis, excluding the fact that Resynthesis is the recreation of a specific existing sound, not a general instrument tone. Given this link, additive synthesis is quite often used in Resynthesis processes.
Component (Physical) Modeling Synthesis
Physical Modeling Synthesis is mathematical, and uses set algorithms to define the harmonic and acoustic characteristics of the sound being generated. This method is mostly used for creating real-sounding instruments, as it is programmed to make characteristic distinctions between various aspects of the instrument being created. For instance, the materials that make up the instrument, the size, the stiffness of a membrane, the volume of a reverberant object (in order to reproduce its resonant frequency), and many other fine details are factored into the algorithm that generates each sound’s different qualities using various forms of synthesis (dependent upon manufacturer).
Wavetable Synthesis employs the use of a table with various switchable frequencies played in certain orders (wavetables). As a key is pressed, the sound moves in order through the wavetable, not spontaneously changing the waveform, but smoothly changing its shape into the various waves in the table.
This method produces sounds that can evolve really quickly and smoothly. The method was intended to create digital sounding noises, so it is not used for instrument replication very often, but is an effective way to create pads or harsh-sounding tones like bells or digital sounds.
Vector Synthesis is almost exactly the same as Wavetable Synthesis, only it employs a two-dimensional grid, through which wavetables can be made even smoother and works with sequences as well as wavetables.
LA (Linear Arithmetic) Synthesis
LA Synthesis was created by Roland as an attempt to utilize Wavetable Synthesis to create real-sounding instrument patches. They achieved this by cutting the waves on the wavetables in half and combining the complicated attack (first half) wave patterns with simple release (second half) wave patterns, thus emulating more of an acoustic environment.
Phase Distortion Synthesis
Phase distortion synthesizers are subtractive synthesizers with one difference – they employ the waveform flexibility of wavetable synthesis in the oscillator. So, instead of having set waves to choose from at the oscillator, you are given full control over the shape of the waveform between all set shapes – in other words, variable waveform control.
Synthesis Methods That Do Not Employ Subtractive Synthesis
FM (Frequency Modulation) Synthesis
FM Synthesis is an alternative approach to altering the harmonic character of a generated wave through the use of filters. FM Synthesis, instead, employs a modulator oscillator that varies the frequency of the sound signal, thus producing new harmonics.
FM synthesis is not very effective at recreating acoustics instruments, but is great for electronic and digital sounds, as well as harsh bell or cymbal-type sounds.
Sample-Based (PCM) (S&S) Synthesis
The main way Sample-Based Synthesis differs from other forms of synthesis is that it does not employ the use of oscillators. In their place, recorded samples are the sound source. These samples are lined up on a keyboard in different zones. Each sample is pitch-shifted (which changes its speed) to span about 5 notes until a new sample is needed (otherwise the sample begins to sound noticeably distorted).
This method is meant to emulate real instruments by recalling actual samples of those instruments upon striking the keyboard. These types of synthesizers can take up a lot of processing power due to the storage and instant recall of samples.
Granular Synthesis works on the same principle as wavetable synthesis, except it pulls from a multitude of samples, rather than oscillator waveforms. It then plays the samples back in 10-50 ms “grains,” rather than playing back entire samples.
The philosophy behind granular synthesis is that sound can be broken down into very small fragments and combined together to create varying timbres and thus highly complex harmonically structured waveforms. These types of synthesizers take up even more processing power than sample-based synthesizers due to the micro sound-scale on which they operate.
In the modern computer-era, all of these forms of synthesis are present in countless programs. Many of these synthesis methods are combined or layered within single programs to accommodate the creation of unique synthesizers. In understanding these methods, experimenting, and combining them, you become the creative force behind sound-synthesis!