Quick answer
Create a project with an audio track, choose your microphone or interface as the input, and monitor on headphones. Set the gain so the track meter peaks around -12 to -6 dB, then record. Cycle mode keeps each pass as a take in a take folder, so you can choose the best one afterward. Add a vocal patch or your own EQ and compression, keep the dry take, and export.
Before you start
- Give GarageBand microphone access. Go to System Settings > Privacy & Security > Microphone and tick GarageBand.
- Connect your input. A USB microphone appears as an input automatically. For an XLR microphone, connect an audio interface.
- Choose your devices. Go to GarageBand > Settings (Preferences in older versions) > Audio/MIDI, and select your input and output.
- Set the buffer. For recording, a buffer size around 128 samples keeps latency low. Use a larger buffer when you mix to ease the load on your computer.
Create the project and a vocal track
- Choose Empty Project, then select the Audio option with the microphone icon.
- Pick your input in the setup window and click Create.
- In the track’s Smart Controls, open Recording Settings and select the mono input channel you’re using, such as Input 1.
Set levels
- Adjust the gain on your interface, or on the microphone, while you speak in your normal recording voice.
- Watch the track meter. Aim for peaks around -12 to -6 dB. Red means clipping.
- Record a short test and listen back before you start the real take.
A quiet, clean take is easier to work with than a clipped one. Keep the level comfortably below the top of the meter.
Monitoring without feedback
Click the input monitoring icon (the arcs icon on the track header) to hear your voice while you record. Use headphones. Monitoring through speakers can feed back or bleed into the microphone.
Recording takes
- Count-in and metronome. Turn them on in the control bar before you start.
- Cycle mode. Press C to loop-record. Each pass creates a take in a take folder, and you choose the best one afterward.
- Punch in and out. Turn on Autopunch to re-record only a selected section, instead of the whole take.
Press R to start recording and Space to stop.
Vocal sound
GarageBand includes vocal patches in the Voice category. They add EQ, compression and sometimes reverb, and their names vary by version. Built-in plugins include Channel EQ, Compressor and Noise Gate. For starting values, see vocal compression settings that work.
In most versions, recording with a patch keeps the dry take, but check your version’s settings if you need an unprocessed recording to re-mix later.
Pitch correction (optional)
Recent versions include pitch correction in the Track Editor, with a slider and a Limit to Key option. Use it lightly. Heavy correction is easy to hear.
Export
Choose Share > Export Song to Disk. The export formats include AAC, MP3, AIFF and WAV. For a master you may edit again later, export WAV or AIFF.
Sample rate
GarageBand projects typically run at 44.1 kHz. For how that compares with 48 kHz, see 44.1 kHz vs 48 kHz. GarageBand has no sample rate setting, so 48 kHz or 96 kHz recording needs a different DAW, such as Logic Pro. If you need 48 kHz for video work, plan for that before you start the project.
Common mistakes
- Monitoring through speakers, which causes bleed and feedback.
- Recording with peaks in the red. Clipping can’t be fixed later.
- Keeping only the processed take. Keep the dry take if you may re-mix.
- Using the built-in Mac microphone for vocals. It works, but it sounds thin and roomy. A USB or XLR microphone is a clear step up.
Related guides
If you’re recording a voiceover without GarageBand, see how to record a voiceover at home on a budget. For capturing a Mac’s system sound alongside your voice, see how to record internal audio on a Mac.