Granular Synthesis Explained: Which Box to Buy

Granular synthesis cuts a recording into tiny slices, called grains, and plays thousands of them back at once. You control how long the slices are, how many overlap and where they come from, so one second of sampled voice can become a pad, a swarm or a stuttering drum (yes, it is as odd as it sounds).

On paper it is a physics lesson. In practice it is a kitchen blender for sound, the good kind, with a lid. This guide covers the four controls that explain every granular instrument, where the idea came from, how it differs from subtractive, FM and wavetable synthesis, and the six instruments we stock that actually do it, from €279 to €1,598.

Grains instead of waveforms

A normal oscillator draws a waveform and repeats it. A granular engine does something stranger: it takes a recording, picks a spot in it and fires off short slices, usually 1 to 100 milliseconds long. That is shorter than a snare hit, and short enough that your ear does not hear a single slice as a separate event.

Fire enough of them, overlapping, and you hear a cloud: one continuous sound made of pieces. Move the spot you read from and the cloud shifts. Freeze the spot and half a second of vowel becomes a held chord. Think of a paper shredder that plays the confetti back in tune. That is the whole trick, and it is why granular synthesis is built on sampling rather than on oscillators.

Four controls that explain everything

Brands label them differently, but nearly every granular instrument gives you the same four dials.

• Position. Where in the recording the grains are read from. Sweep it and you scrub through the sound. Hold it and you freeze one moment.
• Grain size. Short grains turn into buzzy, metallic texture. Long ones keep the character of the source, so a voice still sounds like a voice. The Texture Lab, for example, works from 2 ms to 1 second.
• Density. How many grains fire per second and how much they overlap. Sparse means stutter and glitch. Dense means a smooth cloud.
• Pitch and spread. Pitch transposes each grain. Spread, also called jitter or scatter, randomises position, pitch or panning slightly, which is what stops a cloud sounding like a loop.

Rule of thumb: if it sounds like mush, your grains are too long for the density you set. Shorten them or thin them out. If it sounds like a broken sprinkler, you have the opposite problem, so lengthen the grains or add density.

And if every patch sounds the same, check the source. Granular makes new sounds out of whatever you feed it. Feed it a plain sine and you get a plain sine, only more expensive.

Where the idea came from

The theory arrived before the technology. In 1947 the physicist Dennis Gabor proposed that any sound can be described as a series of elementary grains. In the late 1950s Iannis Xenakis built granular pieces the hard way, with analogue generators and cut tape. Curtis Roads ran granular synthesis on a computer in the mid-1970s, and Barry Truax managed it in real time in 1986.

Today it runs on a battery-powered box (the Texture Lab, below, takes six AA cells). That is a long way from slicing tape with a razor blade, and nobody misses the razor blade.

How it differs from the synths you already know

Method	What makes the sound	Typical result
Subtractive	An oscillator through a filter	Bass, leads, classic pads
FM	Oscillators modulating each other	Bells, electric pianos, metallic tones
Wavetable	Scanning through stored waveform cycles	Evolving, digital timbres
Sampling	A recording played back as one piece	Drums, loops, instruments
Granular	Thousands of tiny slices of a recording	Clouds, freezes, textures, glitches

Granular sits closest to sampling. If sampling is new to you, start with what a sampler actually does, then come back. Compared with subtractive synthesis, the character comes from the recording, and the filter shapes the result instead of defining it.

Several instruments below mix methods. The Lemondrop pairs grains with a wavetable oscillator, and the MiniFreak and Iridium Core put granular next to FM, wavetable and classic waveforms.

What it is good for

• Pads and ambient beds from short recordings. A two-second sample can hold a pad for minutes.
• Freeze and time-stretch. Hold a moment, or slow a sound down without changing its pitch.
• Glitch and rhythmic texture. Sparse density and short grains give you chopped, stuttering rhythms.
• Sound design from found sound. A voice, a field recording, a knock on a table. A recording of your own fridge counts.

Skip granular if you need a bass line by Friday. A subtractive synth gets you there faster, and a bass made of clouds tends to arrive late and slightly drunk.

Your first patch in six steps

These steps work on every instrument below, whatever the maker calls the knobs.

1. Pick a source with something happening in it: a vowel, a chord, a key strike. Two to four seconds is plenty.
2. Set grain size in the middle of its range. Save the extremes for later.
3. Set a medium density, so the sound is continuous but not blurred.
4. Sweep position slowly and listen for the interesting spots in your source.
5. Add a little spread. The cloud stops sounding like a loop.
6. Freeze the position and play a chord. This is usually where the addiction starts.

Which box for which way of working

Every instrument below does granular. They split into two camps: boxes built around grains, and instruments where granular is one mode among many. The first camp teaches you faster. The second suits you if you already have the habit and want one more colour. Prices are what is on our shelf today.

Built around grains

Sonicware LIVEN Texture Lab, €279. The cheapest way to learn the four controls. It slices samples into grains from 2 ms to 1 second, runs four voices and has a 128-step sequencer with parameter locks. It also works as an effects processor, freezing live input for up to 6 seconds. Samples top out at 6 seconds, and it runs on six AA batteries or a power adapter, with a built-in speaker. It is not a studio centrepiece and does not try to be. Marked as available in our showroom.

1010music Nanobox Lemondrop, €415. Four voices, and each voice combines two granular oscillators with one wavetable oscillator, then runs through two filters, two envelopes and two LFOs. The 2-inch touchscreen gives you X/Y control, which is the right interface for dragging a cloud around. It ships with 153 factory presets, you load your own WAV files from a microSD card, and each oscillator holds up to 30 seconds of sample. Also on our showroom floor.

Torso Electronics S4, €835. The grown-up option. Four stereo tracks, each with a five-stage chain: Material, Granular, Filter, Color and Space. The granular stage works from a 4-second buffer and up to 128 grains, with control over size, distribution, contour and pitch. It records at 24-bit/48 kHz, stores 128 scenes and doubles as a USB-C audio interface. If you have decided grains are your instrument, this is the one that will not run out of room.

Granular as one more mode

1010music Blackbox, €584. A sampler first. Its four playback modes are Sample, Clip, Slicer and Granular, with 16-voice polyphony, a touchscreen, microSD storage and USB power. You get an ordinary sampler that happens to have a granular mode, which is a sensible way to try the idea without giving a whole box to it. It is in our showroom.

Arturia MiniFreak, €540. A six-voice hybrid with two digital engines and 31 oscillator modes, granular among them, feeding polyphonic analog filters. MiniFreak 5.0 imports your own WAV, AIFF and FLAC files (up to six seconds each), so the sample and granular engines can chew on your recordings. It has a 37-key slim keyboard and MiniFreak V software included. Grains sit beside virtual analog, FM and wavetable here, so choose it if you want one synth for everything.

Waldorf Iridium Core, €1,598. A 12-voice desktop synth with three oscillators per voice and five synthesis modes: Wavetable, Waveform, Particle, Resonator and Kernels. Particle is the granular sampling mode. The rest of the spec sheet (a touchscreen, a 40-slot modulation matrix, five effect slots per layer) is the reason to buy it. It is in our showroom too.

Here is my honest take, with numbers. If grains are the only reason you are looking at the Iridium Core, don't. The Texture Lab does the grain part for €279, less than a fifth of the price. Buy the €1,598 box for its wavetable engine, its filters and its modulation, and treat Particle as the bonus. Nine times out of ten, your first granular instrument should be the cheap one.

Instrument	Price	How it does granular	Best if
Sonicware LIVEN Texture Lab	€279	Granular synth and processor, grains 2 ms to 1 s	You are starting out or travelling
1010music Nanobox Lemondrop	€415	2 granular + 1 wavetable oscillator per voice	You want clouds with touch control
Torso Electronics S4	€835	4 s buffer, up to 128 grains, 4 tracks	Granular is your main instrument
1010music Blackbox	€584	Granular mode next to Sample, Clip and Slicer	You want a sampler with grains
Arturia MiniFreak	€540	Granular engine among 31 oscillator modes	You want one synth for everything
Waldorf Iridium Core	€1,598	Particle mode among five	You want the whole engine

Plugin or box first? A plugin is the cheapest way to hear what grains do, and there is no shame in starting there. Hardware earns its place when you want knobs under your fingers and no screen between you and the cloud, or when you want to play live without a laptop.

Hear the clouds in Warsaw

Grains are hard to judge from a spec sheet or a video. Four of these six instruments are marked as available in our showroom at Nowogrodzka 6A/102 in Warsaw: the Texture Lab, Lemondrop, Blackbox and Iridium Core. Bring a sample you know well (a vocal line, a snare, a field recording) and push the density knob until it sounds wrong, then right. Call +48 573 580 840 before you come to confirm a specific model is on the floor. Every product carries a 36-month warranty and a 30-day return window, and we ship across the EU. Ask about the free 30-day Test-Drive on eligible gear.

Straight answers

Feed a granular box one second of anything you love and you will not run out of material. Fair warning: you may run out of sleep.

About the author

Michał Kaniuszkiewicz is a content manager and SEO specialist at Wired Tunes. He creates and optimizes content that helps readers find useful information about music, gear, and sound. Beyond his editorial work, he has been making music for 8 years in hip-hop, pop, and electronic genres — giving him insight into the subject not just from an SEO perspective, but from the inside, as a practicing musician and producer. This experience allows him to create content that's equally useful and clear for both beginners and experienced listeners and music creators.

Back to top ↑