How to slow down audio without changing pitch
- Step 1Drop the recording onto the time-stretcher — Add your audio (mp3, wav, flac, m4a, ogg, opus, aiff). It loads into FFmpeg WASM in your browser — nothing is uploaded, which matters for private interviews.
- Step 2Pick a slow-down factor below 1 — For slowing down you want a factor under 1. Good starting points: 0.85 (gentle), 0.75 (transcription), 0.6 (note-by-note music practice), 0.5 (half speed).
- Step 3Understand the duration it produces — Output length = original ÷ factor. At 0.5 the clip becomes twice as long; at 0.75 it becomes 1.33× longer. Plan around the longer duration if you're near a tier limit.
- Step 4Enter the factor (steps of 0.05) — Type your factor into the Factor input. It accepts 0.25–4 in 0.05 steps. Stay above 0.5 for transparent results; below 0.5 the sound gets intentionally smeary.
- Step 5Render the slowed file — FFmpeg applies
atempo=<factor>and re-encodes in your input format. Pitch is preserved by the algorithm, so the key doesn't change. - Step 6Download and loop the tricky part — Save the
-stretchedfile. For practice, cut the hard passage with audio-trimmer first so you only stretch (and loop) the bars you're drilling.
Slow-down presets for practice and transcription
Suggested factors by task. All keep pitch constant; the right-hand column shows how much longer the output becomes (output = original ÷ factor).
| Task | Suggested factor | Speed | Output length |
|---|---|---|---|
| Light slow-down for clarity | 0.85 | 85% | ~1.18× longer |
| Transcription of fast speech | 0.75 | 75% | ~1.33× longer |
| Language learning / dictation | 0.70 | 70% | ~1.43× longer |
| Music practice (riffs/solos) | 0.60 | 60% | ~1.67× longer |
| Note-by-note ear training | 0.50 | 50% | 2× longer |
| Extreme / ambient texture | 0.25 | 25% | 4× longer (two atempo stages) |
Factor → speed → duration
Full factor reference. For slowing down, only the rows with factor below 1 apply; the higher rows are shown for context.
| Factor | Playback speed | Output duration | atempo stages |
|---|---|---|---|
| 0.25 | Quarter speed (slowest) | 4× longer | Two (0.5 × 0.5) |
| 0.5 | Half speed | 2× longer | One |
| 0.75 | 75% speed | 1.33× longer | One |
| 1.0 | Unchanged (default) | Same length | One (no-op) |
| 1.25 | 125% speed | 0.8× (20% shorter) | One |
| 1.5 | 150% speed | 0.667× (33% shorter) | One |
| 2.0 | Double speed | Half length | One |
| 4.0 | Quadruple speed (fastest) | Quarter length | One |
Tier limits for the time-stretcher
Audio-family limits read from the code (lib/tier-limits.ts). The per-file DURATION cap is separate from the file-size cap — a 35-minute voice memo well under 50 MB still exceeds the Free 30-minute limit. The time-stretcher is a Pro-tier tool (minTier: pro).
| Tier | Max file size | Max duration per file | Files per batch |
|---|---|---|---|
| Free | 50 MB | 30 min | 1 (time-stretcher itself starts at Pro) |
| Pro | 200 MB | 120 min | 10 |
| Pro-media | 100 GB (streaming) | Unlimited | 100 |
| Developer | 100 GB | Unlimited | Unlimited |
Cookbook
Real slow-down jobs for practice, transcription, and study. Factors round to the 0.05 step the UI uses.
Slow a guitar solo to 60% to learn it
A fast solo at full speed is impossible to copy by ear. Stretch to 0.6× and every note is the same pitch, just slow enough to catch.
Input : solo.wav (28 s) Factor : 0.60 Output : solo-stretched.wav (28 / 0.60 = 46.7 s) Key : unchanged -- you can play along in the same tuning
Transcribe a 45-minute interview at 0.75×
A fast-talking interviewee is hard to type along to. 0.75× slows the speech enough to keep up while staying natural-sounding.
Input : interview.mp3 (45 min)
Factor : 0.75
Output : ~60 min (45 / 0.75)
Tier note: 60 min output is fine on Pro (120 min cap);
on Free (30 min) the SOURCE already exceeds the cap.Language clip slowed to 0.7× for dictation
A 90-second news clip in a second language goes by too fast. 0.7× gives time to parse each phrase without detuning the speaker.
Input : clip.m4a (90 s) Factor : 0.70 Output : clip-stretched.m4a (90 / 0.70 = 128.6 s) Encoder: aac (m4a in -> m4a out)
Half-speed drum loop for groove study
Slowing a break to half speed reveals the ghost notes. atempo at 0.5 is a single stage and stays reasonably clean for percussion.
Input : break.flac Factor : 0.50 (single atempo stage) Output : break-stretched.flac (2x longer, lossless FLAC)
Loop just the hard bars, then slow them
Don't stretch a whole 4-minute song to drill 8 bars. Trim the bars first, then slow only that snippet — smaller file, faster render, tighter loop.
Step 1: audio-trimmer -> bars-44-52.wav (6 s) Step 2: time-stretcher factor 0.60 -> 10 s practice loop Then loop bars-44-52-stretched.wav in your player.
Edge cases and what actually happens
Audio got faster instead of slower
By designYou entered a factor above 1. For a slow-down you need a factor BELOW 1 (e.g. 0.75). Factor is a speed multiplier: numbers under 1 slow down and lengthen, numbers over 1 speed up and shorten.
Slowed audio sounds watery or smeared below 0.5×
Algorithm limitatempo is transparent down to about 0.5×; past that it chains stages (0.25 → two 0.5 passes) and the time-domain stretch introduces a watery/smeary character. For practice, 0.5–0.75× is the sweet spot.
Source longer than the tier duration cap
RejectedA 45-minute interview already exceeds the Free 30-minute cap before any stretch. Slowing it only makes the output longer. Use Pro (120 min) or Pro-media (unlimited). Remember the duration cap is separate from file size.
Output much longer than expected
ExpectedSlowing multiplies length: 0.5× doubles it, 0.25× quadruples it. A long source plus a deep slow-down can blow past a duration cap or balloon a WAV's size. Trim to the section you need first.
MP3 in, MP3 out for transcription
Re-encode generationStretching always re-encodes (the samples change), so an MP3 picks up one extra lossy generation. For transcription that's irrelevant — intelligibility is unaffected. For music practice you may prefer a WAV/FLAC source.
Want it slower than 0.25×
Below UI minimumThe Factor input bottoms out at 0.25 (quarter speed). For an even deeper stretch, run the 0.25× output back through the tool a second time (0.25 × 0.25 = 0.0625× total). Quality degrades further each pass.
Pitch seems slightly off after a deep stretch
Artefact, not detuneatempo holds the pitch centre constant; what you may hear at very low factors is time-stretch ringing, not a key change. Verify against the original at 1× — the fundamental is the same.
FLAC won't load in Safari
Browser limitSome Safari versions can't decode FLAC/Opus in WASM. Use Chrome/Edge/Firefox, or convert the source to WAV first and slow the WAV.
Frequently asked questions
Can I slow audio down without it sounding lower-pitched?
Yes — that's exactly what this tool does. FFmpeg's atempo filter slows the tempo while holding pitch constant, so a slowed song stays in its original key and a slowed voice keeps its natural timbre.
What factor should I use to slow down?
Any factor below 1. Common choices: 0.85 for light clarity, 0.75 for transcription, 0.6 for music practice, 0.5 for half speed. Output length = original ÷ factor.
How much longer will my file get?
Length scales as 1 ÷ factor. At 0.75 it's ~1.33× longer, at 0.6 it's ~1.67× longer, at 0.5 it's twice as long, at 0.25 it's four times as long.
Is 0.5× still good quality?
Half speed (a single atempo stage) is usable for practice and transcription. Below 0.5× the tool chains stages and the sound gets noticeably smeary, so keep deep stretches for creative effects rather than careful study.
Can I slow down just one section of a song?
Yes — trim the section first with audio-trimmer, then stretch only that snippet. You get a tighter practice loop and a smaller, faster render.
Does the slowed file stay in the same format?
Yes. There's no format option; the output keeps the input extension and is re-encoded with that format's encoder. An MP3 stays MP3, a FLAC stays FLAC.
Will my interview audio be uploaded?
No. The stretch runs in your browser via FFmpeg WebAssembly. Confidential interviews and unreleased demos are processed locally and never sent to a server.
Can I go slower than 25%?
The Factor input's minimum is 0.25. To go deeper, run the 0.25× result through the tool again — but expect heavier artefacts each pass.
Why does slowing change my MP3's file size?
Stretching re-encodes the audio (the sample stream changes), so the file is regenerated. A longer, slowed MP3 at the same bitrate is simply a bigger file because it plays for longer.
What's the longest recording I can slow down?
Pro allows 120 minutes per file; Pro-media and Developer remove the duration cap (up to 100 GB). Note the SOURCE duration is checked, and slowing makes the output even longer.
Is this better than my media player's speed control?
Players can pitch-correct too, but they don't give you a saved file. This produces a downloadable slowed copy you can loop, share, or import into a DAW or transcription tool.
Should I use tempo-changer instead?
If you prefer to think in percentages (e.g. 70%), tempo-changer takes a percent value and does the same pitch-preserving slow-down under the hood.
Privacy first
Every JAD Audio tool runs entirely in your browser via FFmpeg (WebAssembly) and RNNoise. Your audio files never leave your device — verified by zero outbound network requests during processing.