Premiere Pro Club

Change the microphone in Premiere Pro and check the input

Choose the microphone in Audio Hardware, then confirm the recording track hears the right channel. Make a short take before committing the full narration.

Boom microphone with a furred windshield over a rain-wet night street, with amber shop lights behind it

On Windows, open Edit > Preferences > Audio Hardware and select your microphone under Default Input. On macOS, use Premiere > Settings > Audio Hardware. That changes the source for new recordings in Premiere; Default Output chooses where you hear playback.

I'd verify a short, unprocessed take before recording the narration: the device can be selected correctly while the recording track listens to the wrong interface channel. Changing the input also won't replace speech already recorded in an existing clip.

Choose the input device and a separate playback output

  1. Connect the microphone or audio interface and confirm that the operating system can see it.
  2. Open Audio Hardware through the Windows or macOS route above. Adobe's current macOS menu uses Settings; older tutorials may show Preferences.
  3. Check Device Class, which selects the hardware driver. Windows offers ASIO and MME; macOS uses CoreAudio. Select the driver appropriate to the connected hardware.
  4. Choose the microphone or interface under Default Input. If it is an interface, select its actual microphone channel on the recording track as well.
  5. Set Default Output to the headphones or monitoring device you intend to hear.

The distinction comes from the Audio Hardware controls: Default Input feeds recordings, while Default Output feeds monitoring. An interface's external settings depend on its driver; the Settings button opens the hardware dialog for ASIO and CoreAudio devices.

Audio Hardware preferences with the Default Input row highlighted and its device dropdown visible
Select the recording device in Default Input. An interface's individual recording channel is selected on the track.

Open the full-size Default Input control

Record a short voice-over on its own track

Let the picture timing settle before the full narration, so a late trim doesn't move the words away from the shots. For an interview, a useful example stack is V1 picture; A1 existing dialogue; A2 new voice-over. Keep the new recording on its intended audio track.

Put on headphones and turn off the computer speakers before recording, so playback stays out of the microphone.

For one monaural microphone, use a mono recording track. If you need another track, right-click a timeline track and choose Add Track; set the audio count, placement and Track Type in Add Tracks. Match the track to the channels you plan to record.

  1. Select the recording track. If its microphone button is missing, open Timeline Display Settings > Customize Audio Header > Button Editor, drag the Microphone button into that audio track's header and choose OK.
  2. Mark the desired recording range with In and Out points. The default shortcuts are I and O on both macOS and Windows; custom shortcuts can differ.
  3. Click that track's Voice-Over Record button. The voice-over tutorial identifies the button that starts recording; setting In and Out points by itself isn't the start command.
  4. Use the pre-roll countdown to prepare, then speak through the marked range. The timeline recording workflow starts at In and ends at Out. Stop the recording sooner if you need to end the take.
  5. Play the new clip back. Listen for the intended microphone, a complete beginning and end, and any distortion. Park the playhead before its entrance and play across that edit with the picture.

The recorded audio becomes a clip in the sequence and a source item in the Project panel. Listen to it before applying cleanup: processing a silent take or the laptop microphone won't recover the intended recording.

Choose an interface channel in the Audio Track Mixer

A device and a channel are separate choices. If your microphone is connected to an interface, use the Audio Track Mixer recording workflow to choose the channel explicitly.

Microphone and audio interface connect through Default Input in Audio Hardware and Track Input Channel in Audio Track Mixer to a recording track
Choose the device in Default Input, then select its microphone channel on the recording track. These are separate choices.

Open the full-size device and channel map

  1. Open Audio Track Mixer. The default shortcut is Shift + 6 on both platforms.
  2. On the intended recording track, enable R and choose the actual input from Track Input Channel.
  3. Choose Meter Input(s) Only from the mixer panel menu, then select the bottom Record button to enter record mode.
  4. Speak into the microphone and check the input level. Adjust the level on the input device so it doesn't clip.
  5. Select the Play and Stop toggle to begin recording. Select Stop to end the take.

Use a recording track, rather than a submix or master: those receive audio from the sequence and don't offer recording inputs. The input device controls the recorded level; the track fader adjusts the monitoring level. A quieter fader isn't a repair for clipping at the microphone input.

If the microphone is missing or the take is silent

Check permissions and the system input before resetting Premiere. If the operating system can't receive the microphone, choosing another timeline track won't fix that earlier part of the path.

On macOS, open System Settings > Privacy & Security > Microphone and allow Premiere to use the microphone. Then check System Settings > Sound > Input, select the actual device and inspect its input level.

On Windows 11, open Settings > Privacy & security > Microphone. Turn on Microphone access, Let apps access your microphone and Let desktop apps access your microphone. Desktop apps are controlled as a group and may not all appear in the list, so an absent Premiere entry isn't proof that access is off. Check Settings > System > Sound > Input and use the microphone test for the selected device.

Windows 11 microphone permission control with Let desktop apps access your microphone set to On
In Windows 11, allow desktop apps to use the microphone. Premiere may not appear as an individual app toggle.

Open the full-size Windows microphone permission control

Where the problem appearsNext check
The recorded source clip is already wrong or distortedCheck the selected input, interface channel and input-device gain before another take.
The source plays, but the intended sequence doesn'tCheck that sequence's track and monitoring route; compare with a simple sequence before changing device settings.
The same input works in a new projectCompare the original project's recording track and channel assignment. Keep the original project intact.
The operating system cannot receive the microphoneCheck its connection, compatible jack, permissions and system input first.

For an analog microphone, check the microphone and jack specification. Adobe documents channel loss, noise or silence from mismatched mono and stereo connections. Don't treat every silent input as a preferences problem.

Separate echo and delay from a failed recording

Use headphones or turn off the computer speakers to keep playback out of the microphone. In Edit > Preferences > Audio on Windows or Premiere > Settings > Audio on macOS, Mute input during timeline recording suppresses input monitoring to help prevent feedback and echo.

A delayed monitor return is a different issue from a silent file. Adobe names Latency on Windows and I/O Buffer Size on macOS in Audio Hardware. Smaller settings reduce delay; increase them if playback or recording drops out. There isn't one buffer value that fits every connected device.

Check the saved clip and keep its media

Find the new recording in the Project panel. In List View, Media File Path shows its disk location. Use Metadata Display in the panel menu if you need to expose that column; check the actual path instead of assuming every project saves recordings to the same folder.

Keep that audio file with the project's media. Play across the narration's entrance and exit with the picture, confirm the intended voice and monitoring route, then record the full take with the same checked input and channel.