Edit audio on a waveform or as text

Audio recordings open on a waveform. Beneath it is the region track, and to the right are left and right level meters. Editing is non-destructive. Each region is a slice of the original recording, so splits, cuts, fades and crossfades never alter the source file. Your changes are saved automatically. A recording with more than one track (your mic and system audio, say) is mixed together for the waveform, playback and export.

Toolbar

Button	Shortcut	What it does
Select	V	Select a range on the waveform
Split	S	Split at the playhead
Join	–	Rejoin a split: click a highlighted seam (greyed out when there’s nothing to join)
Delete	Delete	Delete the selection
Auto	–	Clean up the whole recording in one step: shorten pauses (for voice) and set the loudness
Pauses	P	Preview and shorten long pauses
Loudness	–	Set the whole recording’s loudness to a LUFS target
Reset	–	Restore the original audio and discard every edit
Script	–	Edit the audio as text: delete words to cut them, type to correct them
Video	–	Make a 9:16 video of the recording, with its transcript as captions
Volume slider	–	Output volume, 0–200%; double-click to reset

Undo and redo are at the top right. Reset also clears the undo history, so use it only when you really want to start over.

Transport

Control	Shortcut
Play / pause	Space
Step backward / forward	← / →
Go to start / end	Home / End
Playback speed (0.5×, 1×, 1.5×, 2×)	Speed button, left
Loop	Loop button, right

A step moves one second at 1× speed. The scrub bar spans only the trimmed part of the recording.

Clean up a recording in one click

Click Auto. What it does depends on what the recording is, shown under the button as Auto: voice, Auto: music or Auto: mixed:

Treat as	What Auto does
Voice	Shortens long pauses and sets the loudness to −16 LUFS
Music	Keeps the pauses and sets the loudness to −14 LUFS
Voice and music	Keeps the pauses and sets the loudness to −16 LUFS

SuperCapture works out which it is when the recording opens. A recording made with the mic set to Music is treated as music. If it guessed wrong, click the caption under Auto and choose under Treat as. Auto is one undo step, and a message says what it changed.

Shorten long pauses

1. Click Pauses or press P. The long pauses are hatched red on the waveform. Nothing has changed yet: it’s a preview.
2. Adjust what counts as a pause: Pauses longer than (300 ms to 3 s) become a shorter gap (0 to 1.5 s), and anything quieter than Silence below counts as silence.
3. Click a hatched pause to keep it as it is.
4. Click Play (or press Space) to hear the result: playback skips the red.
5. Click Shorten to make the cuts (the button counts them, as in Shorten 14 pauses), or Cancel (Esc) to leave the recording alone.

The cuts are made at zero crossings, so they don’t click, and they’re one undo step.

Set the loudness

Click Loudness and pick a target:

Target	Use it for
−14 LUFS	Streaming and video
−16 LUFS	Podcasts and voice
−23 LUFS	Broadcast (EBU R128)

Loudness is measured over the whole recording the way streaming services and broadcasters measure it (ITU-R BS.1770), and peaks are kept below −1 dBTP so the export doesn’t clip.

Fades and crossfades

Select or hover over a region to see a handle at each of its top corners.

• Fade in or out: drag a corner handle inwards. The label shows the length as you drag.
• Change the shape: right-click the fade and choose Fade in shape or Fade out shape: Equal power (the default), Linear, Fast or Slow.
• Remove a fade: drag the handle back to the edge, double-click it, or choose Remove fade.
• Region volume: drag the region’s gain line up or down. It moves in whole dB; hold Shift for tenths. Double-click it to reset.

To blend two regions, drag a region left on the region track so it overlaps its neighbour. The overlap is drawn as a striped zone, and the two regions crossfade across it. Drag right to shorten or remove the crossfade.

Edit by text in Script mode

Click Script and the recording becomes a page of text you can edit. Each word sits under the stretch of waveform that speaks it, pauses show as …, and sound with no words (music, or speech the transcription missed) shows as waveform over a blank.

The first time, click Transcribe. Speech is transcribed on your Mac: on macOS 26 and later with Apple’s built-in speech recognition, and otherwise with a speech model that downloads once, about 60 MB. Pick the language in the header, or leave it to be detected. Supported languages include English, Spanish, French, German, Italian, Portuguese, Dutch, Swedish, Polish, Russian, Turkish, Japanese, Korean, Chinese, Hindi and Arabic.

To	Do this
Select words	Click, Shift-click to extend, or step with the arrow keys
Cut them from the audio	Delete
Silence them and keep the timing	Shift-Delete
Fix a word the transcription misheard	Type over it. Only the text changes; the audio stays as recorded
Cut a pause	Select its … and press Delete

Cuts land on the quietest point between words and snap to a zero crossing, so they don’t click. Each edit is one undo step, and the original recording is kept, as with every edit in the editor.

Remove filler words

Fillers such as “um” and “uh”, and repeated words such as “the the”, are tinted in the text. Clean Up in the header counts them and offers Select Next to review them one at a time, Remove Fillers to cut them all, or Silence Fillers to mute them and keep the timing. Filler lists cover English, German, French, Spanish, Italian, Portuguese and Dutch.

Copy or save the text

Export in the Script header offers Copy Text, Save as Text (.txt), Save as Subtitles (.srt) and Save as Web Subtitles (.vtt).

Transcribe again makes a fresh transcript from the recording, and replaces any corrections you typed. The transcript is searchable from the gallery, so you can find a recording by something that was said in it. With Auto-caption When Mic Audio Is Present on in Settings, recordings with a microphone are transcribed as soon as they’re saved, except those recorded with the mic set to Music.

While Script mode is showing, the waveform tools (Select, Split, Join, Delete and Pauses) are off. Auto, Loudness and the volume slider still work.

Turn a recording into a video

Click Video to make a video of the recording for social posts: 9:16, with the transcript burned in as karaoke-style captions and a visualizer that moves with the sound. It opens in the video editor, where you can change the visualizer’s style and colours, the background and the frame shape, then export or share it like any video.

The video is a new item next to the recording, at 1080×1920 and 30 fps. It’s a copy, so later edits to the audio don’t change it. Transcribe the recording first if you want captions in it.

Other common tasks

Trim the start and end

Drag the trim handles at the left and right edges of the waveform. When you let go, the handle snaps to the nearest zero crossing, which avoids a click at the cut.

Remove a section

1. Press V and drag across the part you don’t want.
2. Press Delete.

Split and rearrange

1. Put the playhead where you want the cut and press S.
2. Click a region to select it.
3. Use ⌘C, ⌘V and ⌘D to copy, paste or duplicate it, or Delete to remove it.

Zoom the waveform

⌘-scroll zooms. A plain scroll pans.

Export

Choose Export from the top bar to open the Export Audio window:

Format	Use it for
WAV (Lossless)	Further editing, archiving
MP3 (Compressed)	Maximum compatibility
M4A (AAC)	Small files that sound good

Sample rate keeps the recording’s own rate by default, or converts to a rate from 48 kHz down to 8 kHz. The export includes your trim, cuts, fades, crossfades, region volumes, loudness and output volume, and shows the duration and progress. The save dialog also offers SuperCapture Archive, which keeps the project so it can be edited again. Dragging an audio item out of the gallery produces an M4A.

Noise removal and the compressor

Noise removal and the record-time compressor work while you record, set by Optimize mic audio for in the recorder. In the editor, Auto and Loudness even out the level after the fact. See Screen and audio recording.

Troubleshooting

• Join is greyed out. No seam can be rejoined. Split a region first.
• The exported file is quieter or louder than the preview. Check the output volume slider. It applies to the export too.
• Auto shortened pauses in music. Click the caption under Auto, choose Music under Treat as, then undo and run Auto again.
• The transcript has a word wrong. In Script mode, type over it. The correction changes the text, captions and search, not the audio.
• A pause I wanted was cut. Undo, open Pauses again and click that pause to keep it before you shorten the rest.

For video recordings with audio tracks, use the video editor’s Audio page instead.