Changing pitch versus changing speed

How to use this workflow

1. Choose the musical goal

   For a singing key, change pitch. To practice more slowly in the same key, change speed independently. For a rate-shifted effect, change linked playback rate.

2. Use the matching control

   The pitch changer has separate semitone and speed controls. Slowed + Reverb uses a linked playback-rate control and optional reverb.

3. Compare the same passage

   Keep the original file, try a short phrase and listen for changes in attacks, vocals and sustain. Generate a file after the settings are final.

Three operations, three different results

The examples below use an ideal 440 Hz tone lasting ten seconds. They are calculated descriptions of the intended transformations, not recordings or measured benchmark results. Real musical signals can also acquire processing artifacts.

Swipe or scroll sideways to read all columns.

The arithmetic behind a linked rate change

With a linked rate multiplier r, frequencies multiply by r and duration divides by r. The pitch interval in semitones is 12 × log₂(r). A rate of 0.8 therefore lowers pitch by approximately 3.86 semitones; it is not exactly a four-semitone transposition.

Independent time stretching and pitch shifting require additional processing to separate those changes. They can preserve the intended pitch or duration relationship, but they do not promise that every transient and vocal texture will sound identical to the source.

Choose a workflow by the result you need

A singer usually wants an integer semitone shift at the same tempo. A player learning a fast passage may want slower tempo at the same key. A slowed or sped-up effect often deliberately uses the pitch change caused by rate.

Be explicit when sharing settings: “80% linked rate” and “80% speed, 0 semitones” describe different sounds. The two tools also use different slider steps, so do not assume a setting available in one interface exists in the other.

Change the singing key and practice tempo together

Suppose a steady C-major recording is at 120 BPM and you want a D-major practice copy at 108 BPM. In the pitch changer, set Pitch to +2 st and Speed to 90%. These controls make two separate choices: +2 moves the notes to the new key, while 90% reduces the beat rate. Both settings are available in the current interface.

The table uses a hypothetical 120-second source. Its values are calculated expectations, not measured exports. Confirm the source key and beat interpretation before using the recipe, then listen to the saved result. Using 90% linked playback rate instead would lower the pitch by about 1.82 semitones; it would not perform the intended +2-semitone transposition.

Swipe or scroll sideways to read all columns.

Compare speed methods without letting reverb hide the difference

Use the same short source in both tools. In the pitch changer, choose 80% Speed and 0 st Pitch. In Slowed + Reverb, select Slowed Only, then manually set Playback speed and pitch to 80% and check that Enable reverb is off. Slowed Only initially selects 85%, so the preset alone does not match this comparison.

With reverb disabled, both versions should have the same intended duration change: a ten-second source becomes approximately 12.5 seconds. The independent-speed version is intended to keep the key, while the linked-rate version lowers pitch by about 3.86 semitones. Compare the same note entrance and sustained sound, with similar playback levels. Keeping the effect tail out of the comparison lets you judge the pitch relationship and processing texture rather than the extra space around the sound.

An octave ratio does not specify a volume change

For an equal-tempered pitch shift of n semitones, the intended frequency multiplier is 2^(n/12). At +12 st it is 2; at −12 st it is 0.5. A hypothetical 220 Hz note becomes 440 Hz or 110 Hz respectively with Speed at 100%. The factor describes the note frequency, not the peak sample amplitude, perceived loudness or an instruction to double your speaker volume.

By contrast, doubling a waveform’s amplitude represents about +6.02 dB of amplitude gain and can cause clipping if there is insufficient headroom. That gain operation does not by itself move a 220 Hz note to 440 Hz. These are mathematical distinctions, not measured amplitude results from this pitch shifter; processing and the playback system can change how loud a shifted recording seems.

Swipe or scroll sideways to read all columns.

Compare note height without rewarding the louder version

Use one short sustained-note passage. Save an original copy and a +12 st version at 100% Speed. Play them alternately at a comfortable level, reducing the louder-sounding playback if necessary. Listen for the interval and the changed timbre first. Adjusting the listening volume externally does not alter the downloaded file.

If you need a technical check, inspect frequency and peak level as separate measurements in an external audio editor using the same sustained segment. A chord, vibrato or noisy attack makes a single-frequency reading less straightforward. The site does not supply a loudness-match button or a measured octave-quality score. Keep any level observation separate from your semitone setting.

Two settings can name the same key in different octaves

If a hypothetical C-major recording needs a G-major label, both +7 st and −5 st reach that tonic pitch class. They differ by twelve semitones, so the complete +7 result sits an octave above the complete −5 result. The key name alone cannot tell a collaborator which register you chose. With Speed at 100%, neither setting intentionally changes the duration.

Use an authored C4–E4–G4 phrase to work through the distinction before choosing a setting. The table gives calculated note destinations, not measured exports. Try both settings on your own short passage, checking bass detail, vocal timbre and the highest note. Choose using the musical register and processing sound you need rather than selecting the first route to the key name.

Write the semitone setting beside the destination key, for example “G major, −5 st, speed 100%.” If you instead use linked playback rate to obtain the −5 st interval, its ideal multiplier is 2^(−5/12), approximately 0.7492, and the recording becomes longer. Independent pitch shifting is the appropriate choice when the original beat timing must remain. An interface with discrete rate steps may not offer that exact multiplier.

Swipe or scroll sideways to read all columns.

Calculate a reversal before processing an already changed file

When you load an already processed file, controls apply to that file rather than to the earlier original. Independent pitch intervals add, while speed multipliers multiply. In a calculated example, a first pass at Pitch +3 st and Speed 80% followed by +2 st and 125% has a net interval of +5 st and a net speed of 100% because 0.8 × 1.25 = 1. The second export may return to approximately the original duration while remaining in a different key.

To reverse the example’s first pass mathematically, use −3 st and 125% on its saved output. Repeating −3 st at 80% would instead slow it further to 64% overall. This arithmetic describes intended pitch and timing relationships; processing twice does not reconstruct bit-identical source samples or erase artifacts introduced during the first pass. Reopen the unchanged original whenever it is available and make the desired final version from that source.

Keep a small record with input filename, pitch interval, speed multiplier and resulting filename for each pass. If you cannot identify the settings used earlier, compare with the source rather than deriving them from a vague filename such as “slow.” Check a known note and both early and late timing cues in the final download. The interface does not recover a prior processing history automatically.

Swipe or scroll sideways to read all columns.

Common questions

Continue with your audio

• Use independent pitch and speed →
• Use linked rate →
• Calculate a BPM change →
• Choose a key using the whole melody →