Time-Stretching and Granular Synthesis: Turning Audio into Infinite Textures

What Is Granular Synthesis?

Granular synthesis breaks audio into tiny fragments — called grains — typically between 1ms and 100ms long. These grains are then reassembled, overlapped, scattered, and manipulated to create textures that range from subtle time-stretching artifacts to completely abstract soundscapes.

Unlike traditional synthesis (which generates waveforms mathematically), granular synthesis works with real audio as its source material. Any recorded sound — a voice, a piano chord, field recording, even noise — can be transformed into something entirely different.

Three Scales of Time

Audio processing at different time scales produces fundamentally different results:

Micro Scale (5-50ms)

At the granular level, individual grains are too short to perceive as discrete sounds. Instead, they blur into texture. Manipulation at this scale creates:

• Shimmering, crystalline timbres

• Chorus-like thickening

• Formant-shifted versions of the source

• Granular "clouds" from pitched material

Meso Scale (50-500ms)

At this scale, you're working with musical gestures — transients, note attacks, rhythmic elements. Processing here is transient-aware: the algorithm can detect where attacks happen and preserve or reshape them. Results include:

• Smooth time-stretching without artifacts

• Rhythmic manipulation (stutters, holds, accelerations)

• Pitch-shifted versions that maintain timing

• Motion and movement within sustained sounds

Macro Scale (0.5-8 seconds)

At the phrase level, you capture complete musical ideas — a chord progression, a vocal phrase, an atmospheric moment. The key technique here is freeze: capturing a buffer and sustaining it indefinitely to create:

• Infinite drones from any source material

• Atmospheric pads from single notes

• Static textures from dynamic material

• Evolving soundscapes via slow modulation of the frozen buffer

Practical Applications

Ambient and Atmospheric Music

Granular processing is the foundation of modern ambient production. A single field recording of rain, processed through granular synthesis, can become an hour-long evolving texture. A piano chord, frozen and pitch-shifted across three octaves, becomes a pad that sounds nothing like a piano.

Film and Game Scoring

Composers use granular synthesis to create unique sonic textures for specific scenes or environments. Freeze a vocal sample to create an ethereal drone for a dream sequence. Time-stretch a metallic impact to build tension in a thriller.

Sound Design

In game audio and commercial sound design, granular processing generates unique UI sounds, transitions, and ambient loops. The source material determines the character — process a door slam, a whisper, or a synth stab and get completely different results from the same algorithm.

Multi-Lens Temporal Processing

Lens by Paraflex Audio takes a unique approach by processing audio through three synchronized lenses simultaneously:

• Micro Lens (5-50ms) — granular texture via streaming granular processing

• Meso Lens (50-500ms) — transient-aware processing with pitch independence

• Macro Lens (0.5-8s) — phrase capture with optional freeze for drones and pads

All three lenses read from the same audio buffer via a shared lookahead system, ensuring phase-coherent output. Each lens can shift pitch independently by ±24 semitones and has its own EQ shaping.

The result is a single plugin that handles everything from subtle granular shimmer to complete audio transformation — with all three time scales running in parallel and blended at the output.

Getting Started

The best way to explore granular and temporal processing:

1. Start with a familiar source — use a sound you know well so you can hear what the processing adds 2. Begin with one time scale — master micro-level textures before combining scales 3. Use freeze sparingly — a frozen moment is powerful because it's unexpected; overuse reduces impact 4. Pitch-shift the layers — setting each lens to different pitch offsets creates harmonic depth 5. Blend with dry signal — parallel processing maintains connection to the original while adding the texture layer

Explore Lens or browse all Paraflex Audio plugins.