Free tools for your next audio edit
Open a short MP3 or WAV to inspect its amplitude over time. See separate mono or stereo channels on a shared scale, then save a static SVG or PNG reference. Your audio stays in this browser; the image does not edit or separate sound.
This replaces a selected recording with an existing three-second synthesized test. It loads only when you press the button and does not autoplay. See the channel example method and listening checks. No AI separation is involved.
No recording selected. Audio stays in this browser.
Up to 1,200 min/max time bins per channel. The full recording is reduced to a visual envelope; this is not a sample editor, spectrogram, LUFS meter or automatic clipping diagnosis.
Make a visual reference
Open mono or stereo audio up to five minutes and 50 MiB. The browser decodes it locally at 48 kHz. Your original is not overwritten, uploaded or re-encoded.
Read the time labels from left to right. Stereo has separate left and right rows with one amplitude scale. Preview the original recording to connect a visible region with what you hear.
Choose a dark, light or transparent background. The SVG is the same graphic used in the preview. PNG renders that graphic at its stated dimensions. Keep your audio separately; an image is not an audio export.
Amplitude over time
A waveform plots sample amplitude against time. The horizontal axis here spans the complete decoded recording. Points above and below a channel’s center line represent positive and negative sample values, not high and low musical notes. A waveform is different from a spectrogram, which represents frequency content over time.
This viewer divides each decoded channel into up to 1,200 consecutive time bins. Each bin retains its lowest and highest sample, including the first and last source frames. It draws the range between those values. With a long recording, many samples occupy one bin: the result is an envelope for an overview, not a trace of every individual oscillation.
For a five-minute recording, 1,200 bins represent about 0.25 seconds each. A brief transient can affect a bin without revealing its precise sample position. Use a dedicated editor to locate an edit at sample level. The image itself cannot play sound or be used to reconstruct the original recording.
Both channels use the same vertical scale, at least ±1. If decoded samples extend beyond that range, the shared scale expands to include them and the image labels that scale. Quiet channels are not enlarged independently. This avoids making a quiet right channel look as strong as a loud left channel just because the plot was resized.
Look, then listen
Look for long quiet regions, broad changes in amplitude, or a prominent event. Listen to the corresponding passage before deciding what it contains. A quiet region can hold soft speech or room tone; a peak does not identify a bell, word or instrument.
Check whether the left and right streams have broadly different levels or activity. A visibly flat channel is a reason to listen and inspect, not proof that another channel contains an isolated speaker. Use the channel tool to review the original recording.
Save an image alongside notes about the source and range. For an extracted event, the sound-effect guide and cue sheets help retain its original source timing. The image does not automatically record your editor’s timeline offset.
Keep the measurement separate
A dense or flat-looking envelope can come from many different recordings. This chart does not detect clipping, determine whether a signal sounds distorted, identify words, measure integrated LUFS, or recover missing detail. Use the loudness meter for integrated loudness and sample peak, and compare what you hear with the original.
Browser decoding may resample the source to 48 kHz before plotting. The displayed rate is the working decoded rate, not a claim about the stored file header. The WAV inspector reads stored WAV properties without using the waveform to infer them.
The image is 1,200 × 300 pixels for mono or 1,200 × 500 for stereo. SVG stores vector paths and labels; PNG stores raster pixels at those dimensions. Transparent mode has dark lines intended for a light backdrop. Images include timing, channel count and amplitude scale with a Saturalabs credit. Filenames and playable source audio are not embedded. A waveform can still reveal characteristics of a recording, so choose deliberately what you share.
No audio normalization, channel mixing, separation, trimming or sound export happens here. To edit a passage, return to the original in the audio trimmer. To find frequencies and harmonics, use a spectrogram or spectrum analyzer in a suitable audio editor. This tool generates a static waveform reference.
A few useful answers
Choose a supported mono or stereo MP3 or WAV up to five minutes and 50 MiB. The browser decodes it locally and plots minimum and maximum samples in up to 1,200 time bins per channel. Choose a background, download the SVG, or prepare a PNG from the same graphic.
The local viewer and image downloads need no account, credits or audio upload. Your original stays unchanged. Images contain channel, timing and amplitude information plus a Saturalabs credit, but no filename or playable audio. Keep image exports only if you are comfortable sharing the information they show.
No. It shows a reduced amplitude envelope, not a transcript, spectrogram, source separator or automatic clipping diagnosis. A flat-looking envelope or a high peak alone does not establish recording damage. Use listening and the appropriate measurement tool.
A stereo recording has separate left and right sample streams. Both are plotted on one shared amplitude scale, so quieter channels stay visibly quieter. The channels are not summed, isolated or normalized. A silent channel can be checked with the separate channel tool.
Both use a 1,200-pixel-wide layout: 300 pixels high for mono and 500 for stereo. SVG retains vector paths for scaling; PNG is a raster copy at those dimensions. Transparent mode uses dark lines intended for a light backdrop.
The browser decodes at 48 kHz. The chart shows that decoded audio in time bins, not every source sample or its stored header. Browser decoding and resampling can affect values. Use the WAV inspector for stored format details and an audio editor for sample-level work.
After preparing your audio