Workflows
Generate a rough cut from a folder of clips and your voiceover
This article describes how to create a rough cut from a folder of assets and a voiceover (VO) track. An asset is a video clip or a still image. In this example, the folder contains 70 numbered assets. Each asset must appear while its matching line of narration is spoken.
MatchCut FCP creates the rough cut in three app steps:
- It transcribes the VO into timed caption blocks.
- It pairs your assets with the caption blocks, in order.
- It writes a Final Cut Pro timeline. Each asset is shown for the duration of its line.
Then, import the timeline into Final Cut Pro. Add transitions and other finishing work in Final Cut Pro.
What you need
- Your VO or narration audio file
- Your folder of numbered assets. The folder can contain any number of assets. This example uses 70.
- Final Cut Pro, to import the exported sequence
1. Number your assets
Number your assets before you start. The numbers determine whether the rough cut is correct.
- Number each asset in the order that it must appear. Start at
001. - Use leading zeros. Use
001to070, not1to70. Leading zeros make sure that the files sort correctly. - Do not skip numbers. The app pairs asset 1 with the first caption block, asset 2 with the second caption block, and so on.
- Use one number sequence for clips and still images. You can place a still image between two clips. Number it in the correct position.
The app sorts the folder by file name, in natural order. For example, clip_10.mp4 comes after clip_9.mp4. Use this naming pattern:
001_intro.mp4 002_still.png 003_broll.mp4 … 070_outro.mp4
The app supports these file formats:
| Type | Formats |
|---|---|
| Video | mp4, mov, m4v, mkv, avi, webm, mpg, mpeg |
| Still images | png, jpg, jpeg, gif, heic, webp |
2. Generate subtitles from your VO (app step 1)
- Open MatchCut FCP. The app opens at Step 1: Drop Your Audio. This is the only active step.
- Drop your VO file on the drop zone. Or, select Choose File….
- Select Generate Subtitles.
The app transcribes the VO on your Mac, with whisper.cpp. The app does not upload your audio.
The app divides the VO into sentences. Each sentence becomes one caption block. Each caption block has a start time and an end time. The app writes an .srt file to ~/Library/Application Support/MatchCut FCP/SRTs/.
If you already have an .srt file, select Already have subtitles? Use an existing .srt…. Then go to section 3.
3. Load your assets (app step 2)
- App step 2, Load Your Clips, is now active. The app already uses the SRT file from app step 1.
- Drag the folder of assets onto the drop zone. Or, select Open. Do not drag the files one at a time. Drag the folder itself.
- The app lists the assets in file name order. It shows the number of assets. It pairs each asset with one caption block.
4. Generate the sequence (app step 3)
- App step 3, Generate the Sequence, is now active. Select Export FCP XML.
- Select a location for the
.fcpxmlfile. Then, export the file.
The app places each asset on the timeline at the start time of its caption block. It shows each asset for this duration:
Duration = start time of the next caption block − start time of this caption block
The timeline has no gaps between shots. The app writes a valid Final Cut Pro XML version 1.9 timeline.
5. Import the sequence into Final Cut Pro
- In Final Cut Pro, select File → Import → XML…
- Select the
.fcpxmlfile that you exported. - Final Cut Pro creates an event and a timeline. Each asset is aligned with your VO.
Move the playhead through the timeline to check the alignment. Then, add transitions, music, and other finishing work.
When the counts do not match
The pairing is strict. The number of caption blocks must equal the number of assets. This is the most important check for a job with 70 assets.
- If your VO produces exactly 70 caption sentences, each asset is placed on its own line. The process is automatic.
- If the numbers are different, MatchCut FCP blocks the export. The app shows a warning before it writes any file.
- You can confirm a manual override. The override exports the smaller of the two counts:
min(asset count, caption count). Matched items are exported. The status footer shows a warning. A blocked mismatch does not use a free export.
To get a one-to-one cut, do one of these steps:
- Split your VO so that it produces the number of caption blocks that you want.
- Add or remove assets so that the number of assets equals the number of caption blocks.
Before you export, open the .srt file that you generated in app step 1. Compare the number of caption blocks with the number of assets.
Tips
- Keep the numbers in sequence, with no gaps. One-to-one matching depends on this.
- Check the SRT file before you align the assets. The number of caption blocks controls the timeline.
- Correct transcription errors before you export. If a sentence is split in the wrong place, edit the
.srtfile. Then, load it again in app step 1 with Already have subtitles? Use an existing .srt…. Export again. - You can repeat any step. Select the row of a completed step to open it. Then, change its input.
To see a visual walkthrough first, open the product tour.