How to granulate audio online in 4 steps
- 01
Load a source sound
Click Open file or drag audio into the drop zone, or press Try a demo to start from a built-in loop. Any sound works as source material — a single note, a field recording, or a full mix.
- 02
Shape the grains
Set grain size (10–500 ms) and density (1–100 grains per second) to decide how the texture is built, then add pitch shift, scatter, and stereo spread to move it away from the original sound.
- 03
Pick a starting point or preset
Try a preset like Cloud or Glitch, or set the position slider to choose where in the source audio grains are pulled from, and adjust reverse-grain probability for extra unpredictability.
- 04
Render and download
Click Render preview to hear the result, set an output length from 1 to 60 seconds, compare Original/Processed, then click Download audio once you are happy with the texture.
What you can do with this granular synthesizer
- Grain size from 10 to 500 milliseconds, the single control with the biggest effect on texture — short grains sound granular and buzzy, long grains sound smoother and closer to the source.
- Density from 1 to 100 grains per second, controlling how thick and continuous the resulting texture feels, from sparse clicks to a dense, overlapping wash of sound.
- Pitch shift from -24 to +24 semitones applied per grain, independent of the source recording's original pitch, for harmonizing or completely reworking a sound.
- Scatter from 0 to 100% randomizes where each grain is read from, and position sets the general playback point in the source, so you can focus on one moment or spread across the whole file.
- Stereo spread widens grains across the left/right field, and reverse-grain probability plays some grains backward for a less predictable, more organic texture.
- Output length from 1 to 60 seconds per render, and a seed value that makes a random-sounding result exactly reproducible whenever you reuse the same seed.
- Four built-in presets — Cloud, Shimmer, Glitch, Frozen — as starting points for common ambient and sound-design textures.
- Render preview to hear every change before committing, then Download audio once the texture sounds right.
What is granular synthesis?
Granular synthesis breaks a sound into hundreds of tiny fragments, called grains, each just a few milliseconds to half a second long, and plays them back at a controllable rate, pitch, and position. Instead of hearing the original recording, you hear a cloud built from its pieces — recognizable in texture but transformed in shape, closer to a wash of sound than a straightforward playback of the file you loaded.
This online granular synthesizer runs the whole process in your browser: load any sound, adjust a handful of grain parameters, and render the result as a new piece of audio. It is a genuinely different way of working with sound than an equalizer or a reverb effect — instead of shaping an existing signal, you rebuild it from small repeated fragments, which is why the same source can end up sounding like a shimmering pad, a stuttering glitch, or a wall of static depending on the settings.
How grain size, density, and position shape the sound
Grain size sets the length of each fragment. Very short grains, under 20 ms, tend to sound grainy and pitched on their own, closer to a buzz than a note; grains above 200 ms sound closer to the original recording, with more of its natural character intact. Most ambient textures sit somewhere in the middle, long enough to carry tone, short enough to blur together into a cloud.
Density controls how many grains play per second, so it decides how thick the resulting texture is — low density leaves gaps and a stuttering, granular character, while high density overlaps grains into a smooth, continuous wash. Position chooses where in the source file grains are drawn from, and scatter adds randomness around that point, so a tight scatter setting repeats a narrow moment while a wide one pulls material from across the whole file.
Grain synthesis parameters explained
Granular synthesis has more controls than a simple effect, but each one changes a specific, learnable aspect of the resulting texture rather than the whole thing at once. The table below summarizes what every parameter does and its typical range, so you can look something up while you experiment instead of guessing at what scatter, position, or reverse probability actually changes in the sound.
None of these settings need to be understood all at once. Start from a preset, change one parameter, and listen — grain size and density explain most of what you hear, while pitch, scatter, position, stereo spread, and reverse probability add finer shading once the basic texture already feels right. The seed value is worth knowing early: keep it fixed and a randomized-sounding patch renders identically every time.
| Parameter | What it does | Typical range |
|---|---|---|
| Grain size | Length of each individual grain | 10–500 ms |
| Density | How many grains play per second | 1–100 grains/s |
| Pitch | Semitone shift applied to each grain | -24 to +24 st |
| Scatter | Randomizes the read position of each grain | 0–100% |
| Position | General playback point within the source | 0–100% |
| Stereo spread | Spreads grains across the left/right field | 0–100% |
| Reverse probability | Chance that a given grain plays backward | 0–100% |
| Length | Duration of the rendered output | 1–60 s |
| Seed | Fixes the random elements so a result repeats exactly | any number |
Presets: Cloud, Shimmer, Glitch, Frozen
Cloud favors small grains, moderate density, and wide scatter for a soft, evolving pad that barely resembles the source material — a good starting point for ambient beds and background texture. Shimmer adds upward pitch shift and stereo spread on top of a similar grain size, producing a brighter, more harmonically rich version of the same idea, closer to a synthesizer pad than a texture.
Glitch uses short grains, high density, and a higher reverse-grain probability for a stuttering, broken-up character suited to rhythmic sound design and transitions. Frozen sets position and scatter to hold on a narrow moment in the source, effectively freezing a single instant into a sustained drone — useful for turning one chord, one word, or one breath into an indefinite pad.
Why use a granular synthesizer online for sound design?
This granular synthesizer online is a natural fit for ambient music and drone work, since it can turn a short recording into a texture that sustains far longer than the original without simply looping it. Feed in a single guitar chord, a field recording, or a vocal phrase, and the grain engine builds a pad that keeps some of the source's character while sounding genuinely new.
It also works well for sound design: glitchy transitions between scenes, granular risers, textured background beds under dialogue, or abstract source material for further processing in another tool. Because everything renders locally, you can experiment freely — try an extreme scatter and reverse-probability setting, render it, and use Continue in another tool to send the result to Add reverb or the Evaporator for an even more extreme stretch.
Tips for better results
- Start from a preset — Cloud, Shimmer, Glitch, or Frozen — then change one parameter at a time so you can hear exactly what it does.
- Lower density and increase grain size if a texture sounds harsh or buzzy; both push the result closer to the original recording.
- Fix the seed once you find a texture you like, so re-rendering or exporting produces the exact same result every time.
- Try very short source clips, even a single note or word — granular synthesis can stretch a second of audio into a full minute of texture.
- Use position with a narrow scatter to freeze on one moment for a sustained drone instead of moving through the whole file.
- Layer a rendered granular texture with reverb using the reverb tool for an even more spacious ambient pad.
- If a result sounds too chaotic, bring reverse probability and scatter down first — they add the most unpredictability of any setting.
Frequently asked questions
Is this granular synthesizer really free?
Yes. Every parameter, all four presets, and audio export are free, with no sign-up and no watermark. There is no limit on how many times you can render a texture in your browser, and nothing is saved to a server.
Does my audio get uploaded to a server?
No. The source sound you load and every grain you hear are processed entirely on your device. Nothing is uploaded, so you can granulate private recordings without them ever leaving your browser at any point.
What is granular synthesis, in simple terms?
It is a way of building new sounds from tiny fragments, or grains, of an existing recording. Instead of playing a file from start to finish, the engine plays hundreds of short slices at a controllable size, rate, pitch, and position, which is what creates the cloud-like or textured result.
What audio formats can I use as a source?
Anything your browser can decode, including MP3, WAV, FLAC, OGG, M4A, and AIFF. You can also press Try a demo to granulate a built-in loop without loading your own file first, which is a fast way to explore the controls.
Can I use this on my phone?
Yes, the controls and playback work in a mobile browser, though dense settings with a long output length ask more of your device's processor, so very long renders may be slower on an older phone.
What does the seed value actually do?
Several parameters, like scatter and reverse probability, involve randomness. The seed fixes that randomness so the same settings and the same seed always produce an identical result, which is useful when you want to reproduce or slightly tweak a texture you liked.
What is the difference between scatter and position?
Position sets the general point in the source audio that grains are read from. Scatter adds randomness around that point — a low scatter stays close to the chosen position, while a high scatter pulls grains from much further away, including from elsewhere in the file.
How long can the rendered output be?
From 1 to 60 seconds per render. For a longer ambient bed, render a few passes — perhaps with slightly different settings, seeds, or source positions — and combine them afterward with the audio joiner into one continuous track.
Do I need musical or production experience to use this?
No. The four presets are a complete starting point, and most people find the sound-design side of granular synthesis closer to exploring than composing — small parameter changes, not music theory, drive most of the result.
Can I export the result and keep editing it elsewhere?
Yes. Download audio saves the rendered texture in your chosen format, and Continue in another tool hands it directly to another Audio Lab tool, such as reverb or the Evaporator, without re-uploading anything or losing quality.
Why does a short recording sound so different after granulating it?
Granular synthesis does not play a file start to finish — it rebuilds it from short, reordered, and often overlapping fragments. That process removes most of the original timing and structure, which is exactly what turns a short clip into an extended, evolving texture.
How is this different from the Evaporator?
The Evaporator applies an extreme, spectral time-stretch to a whole track, preserving its original structure just at a vastly slower speed. This granular synthesizer rebuilds a sound from independent grains you control directly, which produces a more textured, less linear result than a straightforward stretch.