How to slow down a track on GarageBand

The tempo field sits in the display at the top of the window, showing a number in beats per minute. Drag it down or type a lower number.

For this to affect an imported audio file, that file must be set to follow the project tempo. Select the region, open the editor, and turn on Follow Tempo and Pitch.

GarageBand then time-stretches the audio to match, keeping the pitch where it was. Slowing a guitar part to half speed leaves it in the same key, which is exactly what you want when learning it.

This affects everything in the project that follows tempo, so it is the right approach when the whole thing should slow down together.

To slow one region without touching anything else, use Flex Time. Select the region, open the editor, and turn Flex on.

You can then drag the region's edge to stretch it, or place flex markers inside it and drag individual points to change the timing of one part while leaving the rest alone.

Choose the right Flex algorithm for the material. Slicing suits drums, Rhythmic suits most music, and Monophonic is the one to use on a single voice.

The wrong algorithm produces audible artefacts even at modest stretches, so this choice matters more than the amount of stretching.

Sometimes you want the old tape effect, where slowing the audio lowers the pitch with it. Turn Follow Tempo and Pitch off and use the region's pitch control instead, or export and re-import at a different sample rate.

GarageBand is less direct about this than dedicated tools, because its assumption is almost always that you want pitch preserved.

For a voice, a dropped pitch is rarely what you want. It sounds like a slowed recording rather than a person speaking more slowly, and listeners hear the effect rather than the words.

Small changes to a spoken recording are genuinely useful. Slowing by five or ten per cent can make a rushed delivery easier to follow, and with the Monophonic algorithm it is close to inaudible as processing.

Beyond about twenty per cent it starts to sound wrong. Vowels smear, consonants get a metallic edge, and the natural rhythm of speech falls apart.

The better tool for pace in speech is silence. Lengthening the pauses between sentences slows the perceived delivery without touching the words at all, and it is completely artefact-free because nothing is being stretched.

That is usually what someone means when they say a recording is too fast: not that the words are quick, but that there is no room between them.

Time-stretching artefacts hide in transients: the attack of a drum, the start of a hard consonant, a plucked string. Those are where to listen, not in sustained notes or vowels.

Compare against the original rather than judging the stretched version on its own. Duplicate the track, stretch one copy, and mute and unmute between them.

If you are slowing something down to learn it, artefacts do not matter at all. If it is going into a finished piece, they do, and a smaller stretch that sounds clean beats a larger one that does not.