Since bursting onto the music production scene in the late 90s, granular synthesis has become an indispensable weapon in the sonic arsenals of many producers. But where did it come from, why was it so late to the synthesis party and, most pertinently, what exactly is granular synthesis anyway?
What is granular synthesis?
Where did granular synthesis come from?
Like many synthesis methods, the theory underpinning granular synthesis existed long before the technology capable of realising the idea. It was first mooted by Hungarian-born physicist Dennis Gabor in 1947, shortly after gaining British citizenship following his 1933 escape from Nazi-occupied Europe.
Gabor posited that sounds could be crafted by chaining together very short pulses of other sounds, which he referred to as “grains”. Implementing this theory using the technology of the day seemed like a practical impossibility, but this didn’t stop composer Iannis Xenakis from making use of it in his 1959 composition Analogique A-B. The piece is very avant-garde and so not the most relaxing of things to listen to, but Xenakis’ analogue application of Gabor’s idea, created using tone generators and tape splicing, is quite an achievement, not to mention astonishingly futuristic for a piece of music produced some 65 years ago.
I shudder to think what Xenakis must have gone through to create even a second of Analogique A-B’s ethereal granular tones. Of all the sound synthesis methods, this is one that really needs the data-crunching talents of a computer. Unfortunately, it was’t until the late 90s that affordable computing power caught up with Gabor’s 50-year-old idea, allowing granular synthesis to finally escape the research lab and start to run wild in the wider world of music production. Lucky us!
How does granular synthesis work?
The titular grains of granular synthesis are nothing more than short snippets of audio extracted from a longer audio clip or sample. They are typically between 1ms and 100ms in length, but can be longer in some implementations.
The grain is extracted from a chosen point within the source audio and looped to create a steady timbre. In essence, this turns the grain’s snippet of audio into a cyclic oscillator waveform, albeit with some crossfading to ensure the waveform’s end blends smoothly with its start.
Short snippets of audio, referred to as grains, become analogous to the waveform created by an oscillator. Image: PressWhen you play a note, the synth engine calculates the frequency at which a grain needs to cycle to produce the requested pitch. Meanwhile, progressing or modulating the grain extraction position while voices are sounding causes the timbre of the grains to follow the changing harmonic makeup of the source audio.
What we have here, then, is a controllable separation of the usually intertwined pitch and time components of audio, allowing each to be manipulated independently of the other. Not only does this open up all sorts of possibilities for sound synthesis, but it also lies at the heart of audio time-stretch and pitch-shift processing.
Few granular synths are limited to producing and looping just a single static grain. Most can do so, but the resulting sound tends to be quite thin and flat. Rather, the synth will produce a cloud or cluster of grains taken from around a chosen position within the source audio clip, and blend these together before passing them on to the rest of the synth. The quantity of grains produced is typically referred to as the Grain Density.
Clusters of grains can be extracted from around a voice’s main extraction position. Image: PressIf all of these grains were extracted from the exact same position in the source audio, all you’d end up with is a louder grain. In order to add more timbral depth and colour to the sound, then, the position from which individual grains are extracted can be smeared around the main extraction position. This is normally referred to as the Grain Spread.
Small spreads give a thickening effect similar to an analogue synth’s unison mode, but wider spreads can reach a long way from the extraction point and so have the potential to create complex timbral layers and interactions within the resulting sound. Some synths also allow detuning of the individual grains, which, again, adds to the complexity and interest of the timbre being produced.
Many granular synthesis implementations allow all of these grain parameters to be jittered, injecting a random element into the parameter values to create a more varied and animated tone. And these parameters can all be modulated too, giving them even more motion and expressiveness.
Beyond the tone generator, granular synths operate on subtractive principles. Image: PressBeyond the granular tone generator itself, granular synths tend to fall back into familiar subtractive synth territory, with filters to further shape the tone, an amp envelope to sculpt the sound’s volume over time, and various other modulators for creating movement and expressive responsiveness – LFOs, envelopes, keyboard tracking and such.
Assuming you’re familiar with subtractive synthesis, then, once you’ve got your head around the granular part, you’re good to go with one of the most versatile and sonically distinct forms of synthesis there is.
- READ MORE: All you need to know about subtractive synthesis
What is granular synthesis good for?
As a specialised form of sampler, granular synths can turn their hand to pretty much anything you ask of them. Emulating acoustic instruments – the main driver behind the development of samplers – is super simple, because creating usable sustain loops is effortless, and any repetitiveness within that loop is masked.
The separation of time and pitch makes granular synths great tools for retiming drum and musical loops, and there’s little better for creating interactive tempo changes, and for emulating vinyl scratching, spin-down and spin-up effects.
And massive analogue-style synth tones are no problem either. Just load up a synth sample, set some loop points, dial in a bit of grain density and spread, and enjoy.
Where granular synths are particularly adept is in creating sustained and evolving sounds — pads, textures and special effects. In all of these cases, granular synthesis’ ability to focus on tiny timbral details within a source clip and to then abstract these details into completely new timbres means that practically any audio recording offers up a smorgasbord of new and original tones from which to build new instruments. Dial in a wide spread, and throw in some jittering and modulation, and creating huge, mysterious, evolving and scintillating pads and textures becomes very easy indeed
Some say that the downside of granular synthesis is that, when embarking on a spot of sound design, you often can’t predict where you will end up; this is fine if you’re up for a bit of sonic exploration, but isn’t helpful if you have a particular destination in mind. This criticism isn’t unreasonable, but whether or not it is applicable depends entirely upon what you choose as the source audio clip.
For example, if you’re creating a synth brass sound, as long as your source clip is pitched and has a brassy character, you’ll have no problem achieving what you set out to do. But if you load up something with less consistent tonal and/or dynamic characteristics, such as a passage of vocals or a found sound recording, then sure, there’s no way of knowing where you’re going to end up. You can, however, be confident there’ll be something interesting to be discovered and used.
Overall, granular synthesis pays out huge dividends without demanding much investment of effort. If you are comfortable with sampling and subtractive synthesis, then you will have no difficulties creating vast libraries of fascinating and completely original synth patches.
Granular synthesizers you should try
Beetlecrab Tempera
Beetlercab’s Tempera represents audio samples as columns on an 8×8 grid of touch-sensitive pads, with each cell acting as a source for any of four simultaneous granular synth engines. The pads can be interacted with whilst playing the instrument, with any lit pads contributing to the overall sound being produced.
Read our review of the Beetlecrab Tempera to learn more.
Arturia Pigments
With its flexible multi-engined architecture, there’s very little that Arturia’s uber-synth can’t do, and this includes granular synthesis, offered as an option within the instrument’s Sample engine. This then benefits from Pigments’ powerful modulation system, not to mention the ability to be layered with additional synth engines.
Read our review of Arturia Pigments 6 to learn more.
Native Instruments Kontakt
Although sold as a sampler, Kontakt has long been able to flip into a granular mode, although this is not as flexible as a dedicated granular synth (I miss you, Absynth!). This ability also serves as the basis for many of NI’s Kontakt-based instruments such as Pharlight and Ethereal Earth.
Steinberg HALion
HALion 7’s powerful and open-ended architecture allows a largely free choice of different sampling and synthesis elements to be combined within a single instrument. Its easy to use granular engine provides control over the randomness, spread and offset of all the important grain parameters, and is put to good use in HALion’s SkyLab instrument.
BLEASS SampleWiz
Developed in conjunction with legendary keyboardist Jordan Rudess, SampleWiz 2 has both a granular mode and an innovative “Cloud” mode that comes at things in a slightly different way. It’s also available for iPhone / iPad alongside the usual Mac and Windows versions.
Read our review of BLEASS SampleWiz 2 to learn more.
UVI Falcon
Falcon’s IRCAM Granular and IRCAM Multi Granular engines turn UVI’s powerhouse sampler-and-synthesiser into a fully fledged granular instrument. And, of course, you get all Falcon’s other sampler- and synthesis-based engines and tools too
Steinberg Padshop 2
Included in some versions of Cubase, and available for any VST-compatible host, Padshop 2 combines two contrasting engines – one of which is a powerful granular engine – that are ideal for creating pads and evolving sounds.