How to Slow Down a Song Without Changing Pitch

To slow down a song without changing its pitch, use a speed changer that applies pitch-preserving time stretching. Start with a short passage at 0.75× speed. The music takes longer, while its notes are intended to stay at their original pitches. Prepare and listen to the exported file: changing a player's listening speed alone does not necessarily create a slower download.

For a known steady tempo, divide the target BPM by the original BPM. Going from 120 BPM to 90 BPM means 90 ÷ 120 = 0.750×. You do not need to know the song's key to make this time change. If your goal is moving the notes into a different singing range at the same pace, use the separate song transposition workflow.

Choose the phrase before choosing the speed

Find the difficult run, lyric, rhythm, or transition in the original. Keep its lead-in and ending so you can hear where it falls in the phrase. A cut directly at a guitar pick attack can remove information you were trying to study; cutting a final vowel or ringing note short can make the slower copy sound broken before processing begins.

Use the free audio trimmer to save a useful excerpt if needed. It retains the sounds in the selected time range. It does not separate an instrument from the other sounds playing with it. For repeated practice, retain the original excerpt so each speed version starts from the same source.

If the instrument is hard to follow, first try the full mix at a modest slowdown. Original session tracks are another useful source when available. The stems and audio isolation guide explains the alternative of estimating one part from a finished song. Isolation can leave bleed or change detail, so it is a separate choice rather than a prerequisite for slowing audio.

Calculate a speed factor from BPM

Speed factor = target BPM ÷ original BPM. A number below 1 slows the recording; a number above 1 speeds it up. Duration uses the inverse relationship: nominal new duration = original duration ÷ speed factor.

Illustrative changes for a steady 120 BPM, 30-second excerpt

Target tempo	Speed factor	Tempo change	Nominal new duration
60 BPM	0.500×	−50%	60 seconds
90 BPM	0.750×	−25%	40 seconds
96 BPM	0.800×	−20%	37.5 seconds
108 BPM	0.900×	−10%	About 33.33 seconds
120 BPM	1.000×	0%	30 seconds

These are arithmetic examples, not measured exports of a song. A 25% tempo reduction makes the recording one-third longer, not 25% longer. Processing tails and MP3 encoder padding can also affect the saved duration. Check the actual result instead of using the table as a sample-accurate placement instruction.

The BPM-to-speed calculator handles known tempos and shows the nominal result after rounding to the speed control's 0.001× steps. If you do not know the tempo, tap along to a steady passage, then verify the pulse across several bars. Tap tempo measures your tapping. Counting twice per beat or once every two beats can give a double- or half-tempo estimate.

A fixed speed factor preserves changes in the performance's timing; it does not make an uneven drummer follow a constant grid. If a track moves between tempos, use a local passage's tempo for that passage rather than claiming the full song now has one exact BPM.

Make and save the slower recording

1. Open the audio speed changer. Select a mono or stereo MP3 or WAV up to five minutes and 50 MiB. This utility works locally in the browser; it does not accept a streaming page URL or directly edit an MP4.
2. Set a practice pace. Try 0.750×, or enter known original and target BPM in the calculator and apply its suggestion. The available range is 0.5×–2×. A smaller change may make the part clear enough with fewer distracting artifacts.
3. Choose WAV or MP3 and prepare the export. The prepared preview plays the same file offered for download. Listen before saving, including the start and decay.
4. Download a labeled copy. A name such as “chorus-practice-075x.wav” distinguishes it from the original. Download before changing settings, reloading, or leaving; the tool does not keep a cloud library.

Both source and processed output must fit the five-minute limit. A four-minute recording at 0.5× would need about eight minutes and will be rejected. WAV must also fit within 50 MiB. Trim a shorter phrase rather than expecting a long result to be silently cropped.

To try the calculation without a song, download the original 120 BPM click reference and open that saved WAV in the speed changer. It contains 16 clicks spaced 0.5 seconds apart, with a higher tone every fourth click. At 0.750×, the nominal pulse becomes 90 BPM. The listening comparison above includes an actual 0.750× export of this source. It demonstrates one synthetic input, not how natural a vocal or musical arrangement will sound.

Listen for useful detail, not just a lower tempo

Compare the original excerpt with the slowed version at a similar listening level. Check picked notes and drum hits for softened or doubled attacks, singing for unstable vowels, and sustained chords for a grainy or swirling texture. A slower file is useful for practice only if the details you need remain understandable.

Return to the original for each new setting. Repeatedly slowing an already transformed MP3 adds processing and another lossy encode. Try 0.9× before 0.75×, or break the passage into smaller musical phrases, if the larger change obscures the articulation.

The local tool decodes at 48 kHz and exports 16-bit PCM WAV or MP3 at 128, 192, or 320 kbps. It does not retain source tags, artwork, or original bit depth, and no automatic normalization or limiter is applied. Pitch preservation does not mean lossless processing. See the MP3 and WAV guide when choosing a working copy.

Build a repeated practice file with gaps

When you want the same phrase several times with room to answer or play it back, prepare the slower phrase first. Download that file, open it in the free audio joiner, and use Duplicate until the list contains the total number of copies you need. Choose a pause between copies, then preview and save the sequence. There can be two to eight entries, within the joiner's input and output limits.

The order matters if the gap length is important. Consider a ten-second original phrase, three total copies, one-second gaps and a speed factor of 0.750×:

Nominal timing examples, before processing tails or codec padding

Order	Gap in the final sequence	Total duration
Slow the phrase first, then join three copies with gaps	1 second	3 × (10 ÷ 0.75) + 2 = 42 seconds
Join three original copies with gaps, then slow everything	About 1.333 seconds	(3 × 10 + 2) ÷ 0.75 ≈ 42.667 seconds

These are arithmetic examples, not measured exports. Check the actual durations in both tools and listen at each repeat boundary. Choose a pause rather than a crossfade when you need a distinct gap. A repeated file has a fixed ending; it is not endless playback, and duplicating a clip does not automatically find a seamless musical loop.

Slow a song in Audacity instead

For a desktop workflow, select the passage and use Effect → Pitch and Tempo → Change Tempo. Enter known starting and target BPM, or a percent change: 120 to 90 BPM is −25%. Audacity's Change Tempo documentation describes pitch-preserving stretching and its quality and boundary limitations. Check the controls in your installed version and listen to the exported selection.

Change Speed and Pitch is a different effect. Its manual explains that reducing speed also lowers the frequencies. Choose Change Tempo when you want to keep the notes at their original pitches.

Common questions

Can I slow down just the guitar without changing the backing?

A speed change affects every sound in the selected file. Use a separate original guitar track where available. Slowing only that part changes its timing relative to the unchanged backing, so it will not remain aligned automatically.

Can I put the slower audio back under the original video?

Changing duration changes sync with unchanged picture. Use an editor to plan the corresponding picture change or placement, then check the final export. Simply replacing the soundtrack at its original start does not keep the rest aligned.

Does slowing a song identify its notes or chords?

No. It can give you more listening time, but this tool does not transcribe music, identify a key, or verify that you have heard a passage correctly.

Is the speed changer free?

The local speed changer requires no account or credits. Saturalabs' separate AI source-isolation downloads use an account and paid credits. You can slow a supported saved recording without running an AI isolation job.