Edit the story first and the pacing second. Build a version that preserves complete answers and their context. Save it before trimming hesitations or adjusting the joins between sentences. Separating these decisions helps you tighten an interview without changing what the speaker means.
This walkthrough uses CapCut online on Mac in a browser. The controls and example come from the web editor. If you use CapCut as your Macintosh video editor, keep the recording and a transcript beside your edit so you can return to the speaker's original answer.
Decide what the interview should say
Write one sentence describing the point your finished interview should make. Then identify the source passages that explain it. Keep the question when the answer depends on it, and retain qualifications that change the meaning of a claim.
For our example, we selected an interview published by The Green Interview featuring Jane Goodall. The short edit keeps the account of how Roots and Shoots began and the later explanation of why individual choices matter. We removed the intervening scale statistics so those two ideas sit together. We kept the original order of the selected passages.
The source interview Jane Goodall — The Green Interview is by The Green Interview and is available under CC BY 3.0. Our example selects and trims excerpts; the interview participants do not endorse this guide or CapCut. Use your own recording or material whose permission covers your intended use.
Make a short selection note before cutting:
- Context: What does the viewer need to understand this answer?
- Point: Which complete statement carries the interview's main idea?
- Example: Which explanation makes that statement concrete?
- Keep intact: Which qualification or meaningful pause must survive?
Pass one builds the structure
If you are learning how to edit a video on Mac, start with complete thoughts rather than individual filler words. Import your recording into CapCut online and place it on the timeline. Set the canvas to match the intended frame shape; our source and project use 16:9.
Use the transcript to find passages
Open Transcript, select the spoken language, select the video track and choose Transcribe. Our recording used English and Video track 1.
Read the generated text against the recording. In our sample, the transcript treated some opening music as words and misrecognized a few names. That made it useful for navigation but unsuitable as an unchecked quotation source. Correct your selection notes from the recording before deciding what to retain.
Click a word to navigate to its place in the interview. Treat that position as a place to inspect, rather than a reason to cut immediately. Check the surrounding sentence and the timeline boundary.
Split around complete answers
Select the timeline clip that contains the playhead, then use Split at the beginning and end of the passage you want to keep. After splitting, select the unwanted segment and remove it from the edit. Keep your original recording separately.
Selecting the correct clip matters. During our edit, the first split left the earlier segment selected. Moving the playhead into a later segment did not make that earlier selection the right one to split. We selected the segment under the new playhead before making the next cut.
Arrange the retained passages so each answer has enough context. You do not have to reverse their order to improve the structure. Removing a digression can bring two related ideas together while preserving the speaker's sequence.
Save the structure before polishing
Our structure version was about 79.6 seconds long. Its job was to preserve the explanation of the program's beginnings and the later point about everyday choices. We left the smaller pacing decisions for the next pass.
Name the project clearly and export a structure reference before continuing. Keep the source recording and a note of the selected source ranges. An exported reference preserves the version you reviewed; it does not replace an editable project.
Pass two refines the rhythm
Once the story works, review each join and each proposed word removal. Start playback before the edit and continue through the next complete phrase. Check the sound as well as the picture: a tidy-looking boundary can still clip a consonant or rush an answer.
Our fine cut removed a short “you know” from the later passage. It retained the pause after the speaker describes the students' reaction and the pause before the concluding point. The exported fine cut was about 79.2 seconds long. The modest duration change reflects a small pacing decision, rather than a goal of making the speaker as fast as possible.
Decide what a pause is doing
A pause may give an idea time to land, separate two thoughts or preserve the speaker's delivery. Leave it when removing it makes the answer feel hurried. Shorten or remove a pause when it interrupts understanding without adding meaning.
CapCut's transcript panel identifies speech gaps and offers removal controls. In our fine cut it still showed two gaps totaling 1.9 seconds. Those detections were review candidates, not an instruction to remove them. We kept both.
The native transcript view below shows the retained pause before the concluding statement. Use the surrounding words to judge its role rather than treating the displayed duration as a universal threshold.
Keep a small change log
A change log makes the two versions easy to compare. Record the passage, the decision and the reason. For this edit:
If a shortened join interrupts a word, restore enough of the adjacent clip edge to recover the whole sound. Replay the surrounding sentence before accepting the cut. If the join still sounds unnatural, keep more breathing space or use a different complete sentence boundary.
Export and check the actual file
Open Export, choose Download and inspect the export settings. Check them for each version: our export panel reopened with defaults rather than the settings selected for the preceding export. We used MP4 at 360p and 30fps for a source that was 640×360 at approximately 30fps. Increasing the export resolution would not restore detail absent from that source.
Open the downloaded file and check its content, duration, frame shape and sound. Compare the structure and fine cut using the same questions:
- 1
- Can you explain the speaker's main point without reading the transcript? 2
- Does each answer still make sense after the preceding cut? 3
- Are the first and last sounds of words intact at the joins? 4
- Do the retained pauses leave room for the meaning? 5
- Does the exported picture preserve the intended composition?
Keep both version files with the change log. If the story is unclear, return to the structure decisions. If the meaning is clear but one join feels abrupt, adjust that boundary rather than rebuilding the whole edit.
For another editing task, browse the CapCut video workflow guides. If your finished interview needs a separately planned illustrative shot, explore the CapCut AI video generator and keep generated material distinct from the interview recording.
Frequently asked questions
Can I follow this workflow in CapCut online on a Mac?
Yes. This example uses CapCut's web editor on a Mac. Its Transcript, Split and Download steps describe that surface. Do not assume the desktop application's panels or shortcuts are identical.
Should I remove every filler word and speech gap?
No. Review them in the full sentence. A hesitation can carry emphasis or preserve the speaker's delivery. Remove a word or shorten a pause only when the change improves clarity without changing meaning.
How do I keep the structure version while making a fine cut?
Export a clearly named structure reference before making pacing changes. Retain the original recording, the source selection notes and the editable project. Save the fine-cut export under a different name so the two files remain easy to compare.
What if the transcript and the recording disagree?
Use the recording to verify the words and their context. Correct your notes and check the actual timeline boundary. Automatic text is a navigation aid; it should not decide which claim or sound belongs in the finished interview.