Italian
English
Understanding synth types becomes much easier once you separate four ideas that are often mixed together: how the sound is generated, whether the engine is analog or digital, how the instrument is built and routed, and how many notes it can articulate at once. An analog synth is not automatically the same thing as subtractive synthesis; a modular synth is not a synthesis method; and a polysynth can be analog, digital, or hybrid.
This guide therefore uses a practical classification: first the main synthesis methods, then the architectures and product labels that describe the rest of the instrument.
The most useful map: four different ways to classify a synth
| Question | What it describes | Examples |
|---|---|---|
| How is the timbre created? | Synthesis method | subtractive, FM, wavetable, granular, additive |
| Where does the processing happen? | Implementation technology | analog, digital, virtual analog, hybrid |
| How is the instrument organized? | Architecture and format | standalone, semi-modular, modular, software |
| How many notes can it articulate? | Voice structure | monophonic, paraphonic, polyphonic |
This matters because the labels can overlap. A synth can be digital, wavetable, polyphonic, and desktop at the same time. Another can be analog, subtractive, monophonic, and semi-modular. Perfect Circuit explicitly separates terms such as analog, digital, virtual analog, and hybrid from “modular,” which primarily describes architecture and form factor.
The main synthesis types and what actually changes
| Synthesis type | Core principle | Character and strengths | What the player needs to understand |
|---|---|---|---|
| Subtractive | starts with a harmonically rich source and uses filters and amplifiers to sculpt it | basses, leads, pads, immediate and readable sound design | oscillators, filter, envelopes, LFOs |
| Additive | builds a timbre by summing many partials, often sine waves | precise spectral control, organic or abstract timbres | large harmonic structures or macro controls |
| FM | one oscillator modulates the frequency of another | bright, metallic, bell-like, bass and complex evolving tones | operator ratios, levels and envelopes |
| Wavetable | scans or interpolates through organized single-cycle waveforms | moving, modern timbres and continuous transformation | wavetable position and modulation |
| Sample-based | uses recorded audio as the sound source | realism, breadth of timbres, layering | mapping, pitch, filters, envelopes and sample manipulation |
| Granular | splits audio into tiny grains and reorganizes them | textures, drones, creative time-stretching, fragmented atmospheres | grain size, density, position, pitch and distribution |
| Physical modeling | mathematically simulates the behavior of an acoustic object | expressive strings, membranes, tubes and physically impossible but coherent sounds | exciter, resonator and model properties |
| Phase distortion / waveshaping | deforms phase or waveform shape to create richer spectra | sharp digital timbres and strong harmonic generation | waveform transformation and modulation |
| Vector synthesis | blends multiple sources in a two-dimensional space | fluid morphing among different timbres | trajectories, joystick or vector envelopes |
| Wave sequencing | sequences different waves or PCM samples over time | rhythmic, evolving and layered sounds | order, duration, pitch and transitions |
| Spectral / resynthesis | analyzes a sound spectrum and reconstructs it synthetically | deep timbral transformations and morphing | partials, spectral analysis and less conventional controls |
There is no mathematically closed list. More specialized approaches exist — including concatenative synthesis, scanned synthesis, VOSIM/formant methods, pulsar techniques, chaotic systems and many hybrids — but the families above are the most useful for understanding most hardware and software synths musicians are likely to encounter.
Subtractive synthesis: the most readable starting point
Subtractive synthesis normally begins with a waveform that already contains plenty of harmonics and shapes it primarily with filtering. In a classic signal path, oscillator, filter, amplifier, envelopes, and LFOs work together to define tone and dynamics. It is strongly associated with historic analog instruments, but the same structure can be implemented digitally.
Its main advantage is readability: closing a low-pass filter immediately makes the spectrum darker; changing the amplifier envelope clearly changes attack and decay. For learning sound design, it is often the most direct framework.
Additive synthesis: building a sound partial by partial
Additive synthesis works in the opposite direction. Rather than starting with a complex spectrum and removing parts, it constructs the timbre by summing sinusoidal components or partials. In principle this gives extremely detailed control over the spectrum; in practice it can become complex because an interesting sound may require coordinated control over many partials.
Modern additive instruments often solve this with macro controls, grouped partials, morphing systems, or analysis tools rather than expecting the user to edit every component manually.
FM synthesis: complexity through frequency modulation
In FM synthesis, a carrier oscillator produces the audible signal while one or more modulators alter its frequency at audio rate. The result can range from stable harmonic timbres to much more inharmonic spectra. Frequency relationships, modulation depth, and operator envelopes are central to the result.
FM is particularly effective for bright attacks, metallic tones, bells, precise basses, and sounds that evolve strongly over time. Compared with subtractive synthesis, however, the relationship between a parameter change and the perceived result can feel less immediate.
Wavetable synthesis: moving through the timbre
Wavetable synthesis uses a collection of single-cycle waveforms and lets the instrument move among them, often with interpolation. Wavetable position therefore becomes a modulation target much like filter cutoff: an LFO or envelope can continuously change the spectrum while pitch remains under separate control.
It should not be confused with sampling. In wavetable synthesis, the source is normally organized as cyclic waveforms intended to function as an oscillator; in sample-based synthesis, the source can be a recording of arbitrary duration.
Sampling, sample-based synthesis and romplers
Sample-based synthesis starts from recorded audio. A sample can be transposed, filtered, shortened, looped, time-stretched, or modulated. This is ideal when you want to begin with a piano, voice, field recording, acoustic instrument, or any other real-world sound and turn it into playable material.
A sampler generally lets you record or import material and manipulate it. A rompler is oriented more toward playback and transformation of an internal sound library. These labels describe workflow more than a fundamentally different acoustic principle.
Granular synthesis: turning a sample into a cloud of micro-events
Granular synthesis uses extremely short pieces of audio called grains. By changing grain size, density, position, pitch, and temporal distribution, a sound can be frozen, stretched, scattered, or turned into a completely new texture.
It is a natural fit for soundscapes, ambient music, cinematic sound design, and radical transformations of recordings, but it is not necessarily the fastest route to a conventional bass or lead.
Physical modeling: simulate behavior rather than replay a recording
Physical modeling does not simply play back a sample of a string or tube. It mathematically describes how a physical system responds to excitation. Many models can be understood as a relationship between an exciter and a resonator: something injects energy and a virtual structure reacts.
Its strength is expressiveness. Changing excitation strength, virtual dimensions, stiffness, or other model parameters can create variations reminiscent of acoustic behavior while also allowing instruments that could not physically exist.
Phase distortion and waveshaping: generate harmonics by deforming a wave
Phase distortion, famously associated with Casio's CZ line, changes the phase behavior of a waveform to create a more complex spectrum. It is related conceptually to some modulation techniques but is not the same thing as FM.
Waveshaping is a broader family of transformations in which a waveform passes through a nonlinear function. Wavefolding and other shaping techniques can produce major harmonic enrichment from a simple source. Modern instruments often combine them with filters, FM, or wavetable engines.
Vector synthesis and wave sequencing: two different ways to create evolution
Vector synthesis blends multiple sound sources across two axes. A player can move through that space with a joystick or let envelopes and modulation define the trajectory. The result is a morph among different sources, not simply a linear scan through one wavetable.
Wave sequencing instead places a succession of waves or samples in time. Each step can contribute to movement, rhythm, or timbral transformation. The two approaches can coexist in the same instrument, but they solve evolving sound in different ways.
Spectral synthesis and resynthesis
In resynthesis, a sound is analyzed to extract spectral components and then reconstructed with synthesis processes. This lets the instrument manipulate the internal structure of a timbre rather than treating it only as an audio file. It overlaps with additive ideas but can include much more advanced analysis and transformation.
It is powerful for sound design, morphing, and unusual transformations, but it is usually less immediate for someone who simply wants to program a conventional lead, pad, or bass.
