Audio Time Stretcher

Time Stretching vs Changing Playback Speed

Changing playback speed by simple resampling links time and pitch. Play samples at half their original rate and the result lasts twice as long while musical pitch drops; play them twice as fast and duration halves while pitch rises. A true pitch-preserving time stretcher instead rearranges or transforms audio content so the timeline changes while periodic pitch is kept approximately stable.

This distinction matters for musicians, language learners, and transcription. A time stretcher with pitch preservation can slow a difficult passage without the large systematic pitch move produced by tape-style behavior, although it cannot reconstruct information perfectly. Turn preservation off deliberately when a tape-machine or sample-rate effect is the result you want.

How to Slow Down Audio Without Changing Pitch

Choose a value below 1× and keep Preserve Pitch selected. A 0.75× setting makes the output approximately one-third longer, 0.5× doubles duration, and 0.25× creates an extreme four-times expansion. The time stretcher preview asks the browser’s media engine to preserve pitch, while WAV export uses local WSOLA processing so the saved duration is not just a playback instruction.

Start the time stretcher at 0.75× or 0.5× before using the most extreme setting. Moderate ratios usually retain speech articulation and musical attacks more naturally. At 0.25×, repeated texture, softened transients, and phase-like coloration are expected. Loop a short passage to decide whether the slower version remains useful before spending time on a full export.

How to Speed Up Audio While Preserving Pitch

Choose a value above 1× with preservation enabled. At 1.25× the file is about 20 percent shorter; 1.5× reduces it to two-thirds of the original duration; and 2× halves it. A time stretcher can make speech review or rehearsal more efficient when intelligibility remains acceptable, but it cannot restore clarity lost when events are compressed too tightly.

Faster time stretcher processing removes or overlaps temporal detail. Percussive material may feel crowded, breaths may shorten, and reverberation can become less natural. Compare the preview at 1× and your target, then select the least extreme tempo that serves the task. Export uses the same numeric ratio shown beside the control.

Tempo vs Pitch: What Changes?

Tempo describes event rate and overall timing; pitch describes perceived frequency organization. They are connected in raw sample playback but can be manipulated approximately independently with DSP. This time stretcher exposes tempo from 0.25× through 2× and a Preserve Pitch switch. It deliberately omits a semitone control because the current WSOLA path is designed for time scaling, not independent high-quality pitch transposition.

With Preserve Pitch off, the time stretcher linearly resamples exported samples. A 2× result is one octave higher for steady tonal material in the idealized relationship, while 0.5× is one octave lower. With preservation on, the algorithm retains short segments at the original sample rate and changes their placement. Local alignment reduces discontinuities but does not make the process transparent for every signal.

For a changed key at the original duration, use the Pitch Shifter instead.

Using Loop Selection for Practice and Transcription

Enable Loop selection, enter start and end on the stretched timeline, and press Play. The time stretcher maps those visible times back to the source while accounting for tempo. It validates that the start is non-negative, the end is within the processed duration, and the region is at least 0.05 seconds long. Click the downsampled waveform or use the seek slider to move quickly.

A musician can repeat one phrase, a language learner can revisit a sentence, and a transcription user can isolate a difficult word. The time stretcher shades that waveform area as a timeline selection, not a frequency spectrum. Looping affects preview only; export is clearly labeled as the complete file so the interface never silently writes a different region than expected.

Common Uses for a Time Stretcher

Musicians use a time stretcher to practice fast passages at a manageable tempo. Language learners slow pronunciation, transcription workers inspect brief details, dancers rehearse movement at incremental speeds, and educators demonstrate the difference between rate and pitch. Podcast or lecture listeners may speed up familiar material, while creators can explore rhythmic texture on audio they are allowed to use.

Time stretcher processing does not change copyright or permission. Use recordings you own, created, or are authorized to process, especially before publishing or redistributing an exported file. The tool is not designed for voice-authentication evasion or copyright circumvention. It is a local practice, education, and creative audio utility.

Local Browser Processing and Privacy

The file picker and drop zone read bytes with browser File APIs. Decoding uses AudioContext, preview uses a temporary object URL, and export passes copied PCM channel buffers directly to a Web Worker in this page. The time stretcher contains no file-upload request, cloud conversion fallback, or code that adds filenames, waveform samples, or audio contents to product analytics.

The sanitized filename is inserted as text, never evaluated as HTML. Replacing or removing a file pauses playback, revokes its object URL, clears decoded-buffer references, and terminates active processing. Closing or leaving the page performs the same cleanup. General page analytics, if configured by the site, receive only coarse tool actions such as tempo and mode—not user-file content.

Audio Quality and Stretching Artifacts

The time stretcher WSOLA path searches near an expected source position for a short segment that aligns with material already placed in the output, then crossfades overlapping windows. This can preserve periodic pitch better than naive rate change and works reasonably for many speech and music examples. It can still produce flamming on transients, repetition, rough joins, phase-like color, or smearing, especially at extreme ratios.

Speech, sustained music, dense mixes, and percussion stress the algorithm differently. The browser preview may also sound different from export because browser vendors use their own pitch-preservation implementation. This time stretcher is a practical client-side tool, not a promise of studio-grade restoration. Always audition the downloaded WAV before relying on it.

Supported Formats and Browser Limitations

Format support follows the browser decoder. WAV and MP3 are broadly available in current browsers, while AAC/M4A and OGG behavior varies by browser, operating system, and codec build. The time stretcher therefore accepts common audio file types but does not claim to support every format. A clear error appears when decoding fails or the file is corrupt.

Time stretcher input is limited to 100 MB as a compressed-file guard, 30 minutes decoded duration, and 256 MB of decoded PCM. Export also checks a 512 MB working-memory budget because a slow result can contain many more samples than its source. These are browser-memory constraints, not server upload limits. Mobile browsers may need much shorter files than desktop systems despite passing the stated guards.

Exporting the Processed Audio

Press Export WAV after choosing tempo and preservation mode. A page-specific Worker performs the DSP away from the main interface and sends genuine progress updates as it advances through frames. Cancel terminates that worker. The result is 16-bit PCM WAV at the decoded sample rate, mono or stereo, with boundary fades and peak protection when necessary.

The safe output name combines a cleaned version of the original base name with the tempo, such as originalname-time-stretched-075x.wav. WAV is the only export format because it can be encoded reliably without a large third-party codec bundle. The time stretcher does not advertise MP3 or AAC output that it cannot create locally.