Changing speed with an audio speed changer does not preserve pitch — the two move together because the tool resamples the file at a different playback rate. When you choose a faster or slower speed using the Audio Speed Changer's playback-rate rendering, the resulting WAV is shorter or longer and the voices, instruments, and ambient sounds are higher or lower than in the source. That behavior is the standard, predictable result of any AudioBufferSourceNode-based speed change, and it is the same trade-off you would get from a tape deck played at the wrong speed. If you need the original musical pitch kept intact while only the tempo changes, you need a separate pitch-preserving time-stretcher, not this browser tool. The Audio Speed Changer is built for the common cases where the pitch shift is acceptable — speeding up a lecture to skim it, slowing down a spoken clip so you can transcribe it word for word, or producing a quick reference copy to hand off as an uncompressed WAV. The output is a brand-new file with its own RIFF header — it does not patch the original in place, and the new pitch curve is fixed the moment the renderer finishes.
How Playback-Rate Rendering Reshapes the Signal
When you load an audio file into the Audio Speed Changer, the browser decodes the whole clip into a buffer of raw audio samples. The tool then routes that buffer through an OfflineAudioContext and plays it back through an AudioBufferSourceNode whose playbackRate property is set to your chosen speed. Because the same samples are consumed at a different rate, two things happen at once: the playback time shrinks or grows, and every tonal element inside the clip shifts up or down in frequency. A 1.5x setting makes the buffer play 50 percent faster, which by definition pushes every pitch up by several semitones — a noticeable change for music and a clear change for voices. A 0.75x setting does the inverse, dragging the duration out by a third and dropping the pitch by a comparable number of semitones.
This is the documented behavior of the Web Audio API playback-rate path, and it is the reason the FAQ for this tool answers "no" to the question of whether pitch is preserved. The same explanation is given in the Web Audio API specification, which describes playbackRate as a scaling factor on both the rate at which samples are read and the rate at which perceived pitch moves. Tools that keep pitch steady while changing tempo use a different signal-processing step — phase vocoder or granular time-stretching — that this browser speed changer does not perform.
Change Speed and Export a WAV in Your Browser
- Choose an audio file up to 25 MB that your browser can decode. Common formats such as MP3, WAV, OGG, FLAC, and M4A are supported where the browser codec is available; the file stays on your device.
- Wait for the local decoding step to finish. The decoded buffer is held in memory, and the tool will not let you pick a speed until decoding succeeds.
- Pick a speed between 0.5x and 2x. Lower values stretch the clip; higher values compress it. The page shows the result duration and file size before you commit so you can sanity-check the change.
- Select "Change speed and export WAV" to run the offline renderer. The buffer is re-played at the chosen rate through the OfflineAudioContext and serialized as interleaved little-endian PCM16.
- Preview the result inside the page to confirm the new speed, length, and pitch behavior match what you expected.
- Download the fresh PCM16 WAV file to your device. The output is a brand-new audio file — the original is not modified in place.
Duration Math at 0.5x and 2x
The relationship between source length and output length is a simple division: result duration equals source duration divided by the chosen speed. For a one-minute source clip, the calculation works out as follows:
- At 2x: 60 seconds ÷ 2 = 30 seconds, with pitch roughly an octave higher.
- At 1x: 60 seconds ÷ 1 = 60 seconds, with no change at all.
- At 0.5x: 60 seconds ÷ 0.5 = 120 seconds, with pitch roughly an octave lower.
These figures come directly from the tool's own renderer; the page displays the result duration before you download. Anything you compute in advance is a useful sanity check, but the exact value to trust is the one shown after the export step runs. Because the relationship is linear, the same formula works for longer or shorter source clips — a five-minute podcast at 1.25x becomes five minutes divided by 1.25, which is four minutes of output.
What the WAV Keeps and What It Drops
| Item | Status after using Audio Speed Changer |
|---|---|
| Audio samples | Resampled in time at the chosen speed; tonal content shifts with duration |
| Pitch | Changes with speed — not preserved |
| Duration | Shortens at higher speeds, lengthens at lower speeds |
| Output container | Always a fresh uncompressed PCM16 RIFF/WAVE file |
| Source codec (MP3, AAC, FLAC, OGG) | Not retained — the new file is always PCM16 WAV |
| ID3 tags, artwork, chapters, other metadata | Not retained |
| Processing location | Browser only — no upload, no server step |
The WAVEFORMATEX structure that defines a PCM16 WAVE file lists the format as uncompressed 16-bit integer samples, which is why the new file tends to be larger than the compressed source even when it is shorter. For the full structure, Microsoft's WAVEFORMATEX documentation covers the fields the tool writes into the RIFF header. Because the renderer only writes PCM16 frames, every output from this tool shares that format — there is no option to write MP3 or AAC and no preservation of whatever compressed format you started with.
When This Speed Changer Fits and When You Need Pitch Correction
The Audio Speed Changer is the right pick when you want a quick, local change and you do not mind the pitch moving along with the speed. Slowing a podcast to 0.75x so you can catch every word, lifting a lecture recording to 1.5x so a long session fits into a commute, or producing a slow version of a voice memo for note-taking are all cases where the pitch shift is harmless. The output is a clean, broadly readable PCM16 WAV that any local editor can reopen, and the whole operation stays inside your browser.
It is the wrong pick when you want the duration to change but the pitch to stay fixed. Speeding up a song so it plays in less time while sounding in the same key, slowing down a guitar solo to practice along with it without losing the original tuning, or trimming a voiceover while keeping the speaker's tone all need a pitch-preserving time-stretcher. If that is your task, the dedicated approach in Change Audio Pitch Without Losing Quality Locally outlines how that workflow differs and which signal-processing steps matter. For now, the only honest answer to "does changing speed preserve pitch" inside this tool is no — the pitch moves with the speed, every time.
File Size, Codec, and Browser Limits to Know Before You Download
The download is uncompressed PCM16 WAV, so a clip that was originally a 2 MB MP3 can easily grow to 10 MB or more as a WAV, even when the new clip is shorter than the original. Compressed formats encode redundancies that raw 16-bit samples do not, and that difference in storage is the trade-off for getting a universally readable local file. If keeping the source format matters more than the new file's portability, you would need to convert the WAV back to MP3 or AAC in a separate local step.
The page also enforces hard limits that protect your browser's memory. Inputs are capped at 25 MB, 15 minutes of audio, eight channels, 192 kHz sample rate, and 30 million rendered samples in the output. If the source file fails to decode because your browser does not include the right codec, the tool surfaces an error and no output file is created. Browser codec support is not uniform, so the same clip might load in one browser and refuse to load in another — there is no upload and no fallback decoder running on a server.
Every step — file selection, decoding, offline rendering, preview, and download — runs in your browser. No audio data is sent to a remote server, and the source file you picked remains untouched on disk. The output is a brand-new PCM16 WAV with a RIFF header and the rendered channel frames interleaved in little-endian order, as described by the WAVEFORMATEX format used by the tool.
If you're weighing options, Run an Audio Pitch Changer on Mac Without an App covers this in detail.