🔒 Your files stay on your device — core audio processing runs locally. Privacy details →
Free - real FFT capture - runs locally

Online Spectral Freeze

Capture the frequency fingerprint of one instant and turn it into a sustained stereo texture. Shape its focus, motion, brightness and decay, then export an exact-duration freeze.

Before-and-after illustration of Spectral Freeze: Chosen audio moment becomes Sustained texture through Hold spectral bins.
Before → process → afterSee the signal story, then hear it in the workspace.

SPECTRAL MAP

Frozen Spectrum Frame

See where tone changes across the frequency range. Read the curve from low to high frequency, then connect every visible move to something you can actually hear in the source.

Before-and-after illustration of Spectral Freeze: Chosen audio moment becomes Sustained texture through Hold spectral bins.
Chosen audio momentHold spectral binsSustained texture
A frequency-domain view: the source profile, the selected spectral move and the resulting tonal balance share one coordinate system.
Open the full signal breakdownBest source, processing logic, limits and three-pass listening check
  1. Start with the right material

    Holding one harmonic colour from a pad, vocal or instrument to make drones, transitions and sound-design layers without looping a visible waveform segment.

  2. Understand the transformation

    A short-time Fourier transform captures one bounded frame, smooths and shapes its magnitude spectrum, then overlap-add synthesis evolves phase while preserving the selected spectral envelope.

  3. Verify the usable result

    Source, Frozen and Difference monitoring plus a custom-duration WAV or MP3 render made from the captured spectrum and current motion settings.

Frozen-bin motion

Where this workflow stops

The output sustains spectral energy rather than preserving words, rhythm or transient timing. Small FFT sizes blur pitch detail, while large sizes can smear fast changes and expose metallic phase texture.

THREE-PASS CHECK

Make the final decision by ear

  • Sweep or solo only to locate a problem; make the final decision in context.
  • Watch broad tonal balance as well as narrow resonances.
  • Recheck level after any boost so the comparison stays honest.
Capture engine
Hann-windowed radix-2 FFT at a selected source time, from 1024 to 8192 samples.
Synthesis
Bounded phase-continuous oscillator-bank resynthesis from captured magnitudes - not a repeated PCM loop.
Output contract
Stereo at the decoded sample rate and exactly the requested duration after frame rounding; WAV or MP3 export.

How to freeze a moment of audio

  1. Choose mono or stereo audio, or generate the deterministic harmonic demo.
  2. Move the freeze scrubber or enter a precise time, then choose FFT size and a preset.
  3. Shape duration, blur, phase motion, decay, brightness, stereo spread and source attack.
  4. Render locally, switch between Source and Frozen, and export the complete texture.

A frequency-domain hold, not an audio loop

Spectral Freeze takes one windowed FFT frame and stores the amplitude and starting phase of its bins. It then builds a new continuous signal from those bins. Spectral blur smooths neighboring magnitudes, brightness tilts them around 1 kHz, and phase motion adds slow deterministic movement while every oscillator remains continuous.

The optional source attack uses only a bounded slice immediately before the selected moment and crossfades it into synthesis. It does not repeat that slice. Set the attack to zero for a fully synthesized start.

Conventional DSP limits

Privacy: selected audio is decoded, analyzed, synthesized and encoded in this browser. It is not sent to AudioWrench. Normal page assets can still be requested when the page loads.

Spectral Freeze: quick answer and technical limits

Quick answer: A local spectral effect that captures the frequency magnitudes at one selected moment and resynthesizes them as a sustained, evolving texture of a chosen duration.

Best for
Holding one harmonic colour from a pad, vocal or instrument to make drones, transitions and sound-design layers without looping a visible waveform segment.
How it works
A short-time Fourier transform captures one bounded frame, smooths and shapes its magnitude spectrum, then overlap-add synthesis evolves phase while preserving the selected spectral envelope.
What you get
Source, Frozen and Difference monitoring plus a custom-duration WAV or MP3 render made from the captured spectrum and current motion settings.

Know before you use it: The output sustains spectral energy rather than preserving words, rhythm or transient timing. Small FFT sizes blur pitch detail, while large sizes can smear fast changes and expose metallic phase texture.

Privacy: Selected audio is processed in this browser and is not uploaded to AudioWrench. Normal page assets can still be requested as described in the privacy notice. Privacy details →

FAQ

How does Spectral Freeze sustain one moment of audio?
It takes a Hann-windowed FFT at the selected time, stores the magnitude and phase of its frequency bins, and resynthesizes selected bins with phase-continuous oscillators for the requested duration. The source PCM is not repeated as a hidden loop.
How is Spectral Freeze different from Spectrogram, Granular and Reverb?
Spectrogram visualizes frequency energy but does not create a sustained output. Granular works with short time-domain slices, while Reverb creates decaying reflections. Spectral Freeze captures one frequency-domain frame and synthesizes a new stationary or slowly moving texture from its bins.
What are the limits of this browser spectral freeze?
A single FFT frame cannot preserve changing words, rhythm or transients. Larger FFT sizes improve frequency resolution but average a longer instant, strong blur spreads energy into neighboring bins, and oscillator-bank resynthesis can sound synthetic or phasey. It is a creative effect, not forensic restoration.
Is my audio uploaded, and what do WAV and MP3 contain?
The selected audio is decoded, analyzed, synthesized and encoded on this device and is not uploaded to AudioWrench. WAV and MP3 contain the complete Frozen stereo output at the decoded sample rate and the exact requested duration after frame rounding. MP3 is a new lossy encode.

Related spectral and texture tools