Online Pitch Changer
Pitch and duration use a phase vocoder and band-limited resampling. Large changes, transients and stereo phase can produce audible artifacts. Files are decoded locally; unsupported or protected media cannot be processed. WAV export is 16-bit PCM, including when the input is MP3.
Related tools
- Audio Pitch Changer — Handle a supported audio file from selection through reviewed export.
- Online Key Changer — Apply a selected tonic-to-tonic shortest shift to audio.
- Video Pitch Changer — Change local-video audio pitch and choose audio versus visual output.
- Pitch and Tempo Changer — Plan and compare combined independent pitch/time changes.
Change the pitch of a local audio recording, adjust its tempo independently and download the processed result as WAV. Start with a short file, enter a semitone shift and keep the tempo multiplier at 1 if you want the original duration. Preview the result before using it for singing practice, rehearsal or a creative edit.
The main Online Pitch Changer works with the recording you select in your browser. It does not fetch a song from a pasted link, identify its key automatically or correct every sung note. Those are different tasks, and choosing the right workflow makes the result easier to understand.
Change pitch in a few steps
- Select a readable local audio file.
- Check the displayed duration and channel information.
- Enter Pitch change in semitones, from -12 to +12.
- Keep Tempo multiplier at 1 unless you also want a duration change.
- Choose Process audio and wait for completion.
- Compare the original and processed previews, then download WAV.
Positive semitones raise pitch; negative values lower it. Begin with a small change that answers your musical question. A large shift may be numerically correct while sounding less natural, especially on a dense mix or a voice with detailed articulation.
After editing a setting, process again. The earlier output is invalidated so it is not confused with the new requested values. Keep the original file as your starting point when comparing several versions.
Separate pitch, tempo and duration
Pitch describes how high or low a repeating sound is. Tempo describes the musical pace. Duration is the length of the recording. A basic playback-rate change links them, but this tool processes pitch and time separately.
At tempo 1, a pitch adjustment keeps the intended output duration equal to the input duration. At tempo 0.5, the output is approximately twice as long. At tempo 2, it is approximately half as long.
| Goal | Pitch | Tempo |
|---|---|---|
| Lower the key without slowing | A negative semitone value | 1 |
| Practice more slowly at the same pitch | 0 | Below 1 |
| Raise pitch and shorten the recording | A positive semitone value | Above 1 |
These are independent processing requests, not promises of artifact-free sound. Listen for changes to attacks, sustained tones and stereo detail before accepting an export.
Choose a semitone amount
One equal-tempered semitone is one chromatic half step. Twelve semitones make an octave. A shift of +12 doubles frequencies; -12 halves them.
The frequency multiplier is 2 raised to the semitone amount divided by twelve. For +3 semitones, the multiplier is approximately 1.189207. A 250 Hz component would move to approximately 297.302 Hz in an ideal uniform shift.
This example explains the requested interval, not a measured result from your recording. A full recording includes transients, noise and complex overlapping components that do not behave like one isolated test tone.
If you know an original and destination tonic, use the Semitone Calculator to check the interval or the Online Key Changer to apply a tonic-based shift to a file.
Pick a comfortable practice key
Identify the melody’s actual low and high notes in the version you intend to sing. Then consider whether a uniform shift moves both endpoints into a comfortable working area.
A key name alone does not show a melody’s register. A song in C can be arranged in different octaves, and sustained high passages may remain demanding even when the extreme notes fit.
The Vocal Range Test helps establish comfortable endpoints. Use that information as context rather than assuming a particular key always suits a particular voice category.
For a backing track, the Karaoke Pitch Changer explains arrangement and rehearsal checks. A uniform shift moves the entire recording; it does not isolate a singer from the accompaniment.
Use tempo as a multiplier
The tempo control accepts 0.5 to 2. It is a ratio, not a BPM entry or an automatically detected beat value. If the source tempo is known, multiply it by the selected ratio to estimate the new tempo.
For example, a known 100 BPM passage at 0.8 becomes 80 BPM. A sixty-second excerpt at 1.25 becomes approximately forty-eight seconds. Duration changes inversely with the multiplier.
If slowing is your only goal, the Change Speed Without Changing Pitch page keeps the task focused. For pitch-only work, use Change Pitch Without Changing Speed.
Prepare a file within the processing limits
The audio processor accepts a nonempty file up to 25 MB and ninety seconds. It supports mono or stereo input, with a cap of seventeen million decoded samples. File size and decoded size are different limits.
Browser codec support determines whether a file can be read. MP3, WAV, OGG, M4A and WebM are common input choices where their codecs are supported; a familiar extension is not a guarantee.
Choose a representative excerpt rather than assuming a long compressed song will fit. A short section containing the difficult phrase, attack or sustained note is often enough to compare the settings you need.
Protected, unsupported or unreadable media cannot be made usable by changing its filename. A decoding error asks for a different supported source, not a renamed extension.
Know what the download contains
Audio output is 16-bit PCM WAV, including when the input was MP3. Selecting MP3 does not turn the export into an MP3 encoder. The MP3 Pitch Changer explains that input/output distinction.
A WAV can be larger than the original compressed file. Its size depends on duration, sample rate and channel count. Saving as WAV does not restore detail already lost from a compressed source or remove processing artifacts.
The original file is not overwritten. Download creates a separate result. Keep clear filenames or your own notes for the semitone and tempo settings when comparing versions outside the browser.
Listen for processing limits
When a trial seems wrong, identify which part of the result failed before changing several settings. A correct interval can still be the wrong choice for the melody, and a readable download can still contain an unwanted processing texture.
| Observation | Next check |
|---|---|
| Pitch fits but the phrase is too slow | Review the tempo multiplier separately |
| The whole track moved, including a guide vocal | Check whether the source contains the parts you intended to process |
| The downloaded file is unexpectedly large | Check WAV duration and channels rather than assuming the shift failed |
| The key seems right but one melody note remains wrong | Remember that uniform transposition does not correct individual notes |
Save the settings that produced each useful trial. If you later compare versions, return to the same phrase and listening level. This makes a decision about register, pace or artifacts more informative than repeatedly changing everything until one version seems different.
The processor uses phase-vocoder time processing and band-limited resampling. These methods allow independent pitch and duration requests, but sharp transients, wide stereo material and larger adjustments can sound smeared or phasey.
Pitch shifting also affects perceived vocal character. This main tool does not offer a separate formant-preservation switch. The Formant Shifter provides a different approximate spectral-envelope effect rather than guaranteed natural voice conversion.
Use the same playback level for comparisons and inspect the same passage. A louder result can seem more impressive without actually being a better fit for the musical task.
Choose another input workflow when needed
The Video Pitch Changer processes a local video’s audio and can create browser-encoded WebM video. This homepage exports audio rather than a new video container.
The YouTube Pitch Changer uses explicit supported browser-tab audio sharing or a local file. A URL alone does not provide the page with a recording.
For written notes and chord symbols, use Transpose Music Online. For estimated tonal candidates from a file, use the Song Key Finder and check its alternatives before choosing a shift.
Control the session and keep the source local
Files are decoded and processed locally by the tool. It does not upload them for server processing or maintain a persistent results library. Download any result you want to retain.
Cancel discards the current processing job’s output. Stop audio / capture stops active playback or capture; Reset clears the working file and results. Native browser decoding may finish internally after cancellation, but a canceled job is not presented as a completed export.
These statements concern this tool’s media workflow. They do not describe every unrelated service that may be present elsewhere on the website.
Frequently asked questions
Can I change pitch without changing speed?
Yes. Keep the tempo multiplier at 1 and select the desired semitone shift. Preview the output for artifacts before downloading.
Does the tool detect the original key or BPM?
The main processor does not. It applies the values you enter. Use the separate key finder for an estimated tonal starting point.
Can I download the result as MP3?
This version exports 16-bit PCM WAV. MP3 is a supported input where the browser can decode it, not an output option.
Can I paste a streaming link?
The homepage works with a selected local file. The tab-capture page provides a separate permission-based workflow where supported.
Will it remove vocals or correct every note?
No. A uniform shift affects the whole recording. It does not separate stems or make note-by-note performance decisions.
Is processing lossless?
No such guarantee is made. Independent pitch/time processing can create audible changes, and WAV export does not make those changes disappear.