The Primary Types Of Synthesis - Sound Design
Many sounds that we hear today in songs, films, soundtracks and other media are synthesised using either analogue synthesizers or digital plugins. Broadly, there are 5 basic kinds of synthesis, and several other types are derived from these.
Subtractive synthesis is based on the idea of subtracting frequencies from a rich, harmonic sound to achieve the desired tone. This is achieved with the help of two key components, an oscillator and a filter. The oscillator is routed through the filter, and the filter cuts the frequencies based on its type (LP, HP, BP, BR). Subtractive synthesis does not apply to sine waveforms, as they do not contain any harmonics to attenuate.
It is still widely used for designing bass sounds, leads, plucks, pads, etc. By modulating the oscillators and filters with different envelopes, a simple subtractive synth can generate a vast array of sounds. Though initially designed with just one oscillator and one filter, many modern synthesizers now employ a combination of subtractive and other types of synthesis with the help of multiple oscillators and filters.
Additive synthesis achieves comparable results as subtractive synthesis through a very different approach. Unlike subtractive, additive synthesis relies on adding several simple waveforms to make up a more complex waveform. In this case, instead of using one oscillator and one filter, multiple oscillators are used to generate the waveforms. Usually, all the oscillators play just sine waves at different frequencies to generate a complex sound.
This principle was originally applied through huge organs, where pulling a lever meant adding more pipes which add harmonics. This concept is difficult to implement on analog synthesizers and isn’t cost-effective, as each new sine wave requires an oscillator of its own. However, using digital software synths, additive synthesis can be used to generate a wide spectrum of sounds.
Frequency Modulation (FM)
FM synthesis, or frequency modulation synthesis is known for its unique, metallic, gritty sounds. Producers like Skrillex use FM synthesis to make their bass and wobble textures in their tracks. This is possible because of the way FM works. All the sounds are made up with just sine waves most of the time. There are two components to an FM synth, a carrier and a modulator. The carrier signal is generated by an oscillator (also known as operator in FM synths) and it is modulated by another sine wave like an LFO. This modulator controls the pitch of the carrier signal. However, the frequency of the modulator is within the human hearing range, and this produces a bright metallic sound. We cannot hear the modulator on its own. We only hear its effect on the carrier.
FM synthesis is based on the ratio of the modulator signal to the carrier signal. Simple integer ratios (2:1, 4:1, etc) produce a harmonic source and non-integer ratios produce inharmonic sounds (2.73:1, etc).
Sample-based synthesis, as its name suggests, relies on previously-recorded audio samples, rather than generating sound through oscillators. By doing away with oscillators entirely, sample-based synthesizers
require much less processing power and can arrive at complex sounds based on the original sample. The downside to this is the tedious amount of effort required to effectively sample any real instrument.
Pianos can be sampled by recording different keys at different velocities. Assuming that an 88-key piano is being sampled with 128 possible velocities for each key, sampling the piano would have meant recording over 11,000 samples! Fortunately, all keys at all velocities aren’t necessary to be sampled, as adjacent keys and values can be interpolated using complex algorithms. However, some instruments are far more challenging to properly sample, such as strings. The different articulations along with velocity and other parameters makes it very cumbersome.
This is where hybrid synthesizers come into play. By combining samplers and oscillators into one unit, these powerful synthesizers can get very close to emulating real instruments with a human feel to them
Most modern synthesizers have oscillators which can play either sine, square, saw, triangle and so on. They are not capable of transitioning from one waveform to another in a smooth fashion. This is overcome with the help of wavetable synthesis. By storing different samples or snapshots of waveforms, a wavetable synthesizer can go through all the snapshots in a rotary fashion which can result in a smooth transition.
Many sounds in genres of music like dubstep and electro are not static sounds. They always have movement associated with them. Wavetable synthesis helps in emphasizing that movement. It is extensively used in bass design, pads, leads, plucks and other textured sounds.