Skip to content

Free Audio Visualizer for Music & Podcasts

Any audio file becomes a clip you can post: a track, a podcast episode, an interview, a voice note. The styles below are ordered for spoken word, which is the harder case to make read well. Free, and it runs in your browser.

What it reads

Anything your browser can decode: MP3, WAV, M4A, AAC, OGG, FLAC. The file is opened locally, analysed locally, and never sent anywhere. That matters more than usual for audio that is not music, since unreleased interviews and client work are exactly the sort of thing you do not want sitting on somebody else's server.

The analysis pipeline is the same regardless of what the audio is: loudness over time for the waveform styles, frequency content across 256 logarithmically spaced bands for the spectrum styles, plus onset detection driving the beat reaction.

Styles that suit non musical audio

Speech sits in a much narrower band than music, so the styles built to show off a full mix have little to work with. The ones below were picked because they stay legible when the only thing moving is a voice.

Podcasts and interviews. The three band meter shows bass, mid and treble as labelled bars, which stays readable during speech and does not flail about in the pauses. A plain waveform is the other safe choice. Add the episode title in the text layer and export square or vertical for social clips.

Voiceovers and audiobooks. The minimal line is the quietest option: a single horizontal trace that moves with the voice and sits under a caption without competing with it. Turn level normalisation on, since spoken recordings are usually quieter than mastered music, and ease the smoothing up so the pauses between sentences settle rather than snapping flat.

Samples and loops. The oscilloscope shows the actual shape of the wave rather than an abstraction of it, which is useful when the character of the sound is the point. Turn smoothing down and detail up.

Ambient and field recordings. The particle and terrain styles both respond to gradual changes in level rather than to individual hits, so they suit material without a strong beat.

Sound design and effects. Radial styles with a trail give a sense of an impulse spreading out, which matches how single hits and impacts read.

Best presets for voice

Ordered for speech: the calm, legible ones first, then the styles that suit samples and atmospheres.

  • Minimal Line

    A single thin line that reacts to level rather than to individual frequencies. Restrained enough to sit under text.

  • Multi-band Spectrum

    Bass, mids and highs shown as three separate meters, which makes the mix easy to read at a glance.

  • Waveform

    The classic horizontal wave. Reads clearly at small sizes and sits well under a title.

  • Oscilloscope

    A hard-edged trace on an optional grid, in the style of lab test equipment.

  • Particles

    Points that are emitted on transients and drift outward. The louder the passage, the denser the field.

  • Mountain Spectrum

    The spectrum drawn as a ridge line with layered slopes behind it, like a landscape.

What people make with it

Podcast teasers. Pull the sharpest thirty seconds of an episode, put the show name and guest in the text layer, and export at 1080 by 1080. It reads in a feed without sound, which is how most people will meet it.

Audiograms for Shorts and Reels. Export vertical at 1080 by 1920. A quote on screen with a line moving underneath it gives the eye something to follow while the words do the work.

Voice notes and clips worth keeping. A message, a rehearsal take, a recording of someone you want to remember. The audio never leaves the machine, which matters more here than it does for a finished track.

Overlays for edited video. Export with a transparent background and drop the trace over your own footage, so a talking head gets a reactive element without a second render. See transparent export for the formats.

Questions

Does this work with speech, not just music?

Yes, though not every style suits it. Speech occupies a narrow frequency range compared to music, so a full spectrum has little to show. The three band meter, a waveform or the minimal line all read better. Turn level normalisation on, since spoken recordings are often quieter than mastered music.

Can I visualise a very short sample?

Yes. There is no minimum length. For anything under a couple of seconds, note that smoothing takes a moment to settle, so start your export slightly before the part you care about and trim afterwards.

What about mono recordings?

Fine. Channels are averaged into one before analysis, so mono and stereo behave the same.

Is there a limit on file size?

50 MB and ten minutes. Both are checked before decoding, so an oversized file is refused immediately rather than after a long wait.