The complete guide

How to auto-cut a multicam from raw clips.

From the camera cards and recorder files to a finished Final Cut Pro edit that cuts to whoever's talking. Speaker Cut syncs the sources, builds the multicam, does the switching, and hands you a timeline. About 10 minutes end to end — no prep in Final Cut first.

Speaker Cut → Final Cut Pro ~10 min macOS

Only have one recording, with no separate mics?

Use Highlight Cut instead. Speaker Cut tells the voices apart in a single video, names who's who, and turns the best moments into Final Cut timelines. Follow the Highlight Cut guide →

Part 1

Gather your sources

Two minutes of sorting before you open the app. Get every file in one place and decide which feed is your master.

Collect every camera and recorder file

Pull every camera angle and every external audio-recorder file (lav packs, a Zoom or field recorder, a StreamYard program track) into one folder. Media stays where you put it — Speaker Cut references it, never copies or uploads it.

ATEM shootsIf a camera or ISO record split each angle into several files, keep them together — they all go into the same angle in the next part, and Speaker Cut plays them back-to-back as one continuous feed.

Pick your master (monitoring) angle

The master is the one feed that hears everyone — your program mix, a room mic, or in an in-person shoot any camera that picks up the whole table. Every other angle is synced against it by audio, so it just needs a clean recording of the full session.

Which one?Remote recordings (StreamYard, Riverside): the program mix. In-person with a recorder: the room mic or the recorder's mixdown track. Cameras-only: the wide, since it hears every seat.

Think in angles: one per camera or mic

Each person's camera is one angle and their clean mic is another — even when both came off the same camera card. A wide or room cam is its own angle too. Naming them by person + role (April — Cam, April — Lav, Room — Wide) makes the lane assignment later a two-click job.

Part 2

Build the multicam in Speaker Cut

Drop the clips into angles, mark the master, and let Speaker Cut listen and line everything up.

Choose “Build a multicam from clips”

Open Speaker Cut. On the home screen pick the recommended path, Build a multicam from clips, and give the session a name (a date and the guest works well — it becomes the project name in Final Cut).

Add an angle for each camera and recorder

Click ＋ Add an angle, type its name, and drop that source's clip(s) onto it. Drop all of an ATEM angle's split files onto the one angle — Speaker Cut orders and joins them. External audio recorders drop in as audio-only angles; that's expected.

Mark exactly one angle as Master

Click Set master on the feed you chose in Part 1. It turns gold. You need at least two angles with clips and exactly one master before Build multicam lights up.

Build, then read the sync report

Hit Build multicam. Speaker Cut listens to every angle against the master and reports each one's offset, a confidence meter, and any drift across the session, flagged Locked, Low conf. or No lock. Locked all the way down means you're done here.

No lock?That angle didn't share enough audio with the master. Usually it's a camera with its mic off, or a master that can't actually hear that person. Give the angle scratch audio, or pick a master that hears everyone, and rebuild.

Continue to cut

Press Continue to cut →. The assembled multicam loads straight into the cut — the same screen you'd land on with a Final Cut export, angle names and all.

Already have a multicam in Final Cut Pro? Use it.

Speaker Cut still takes a Final Cut export unchanged. Build the multicam clip in Final Cut (right-click your clips → New Multicam Clip…, sync by audio or timecode), name each angle by person + role in the Angle Editor, then File → Export XML…. On the Speaker Cut home screen choose I already have a multicam and drop the .fcpxmld bundle, a zip of it, or the Info.fcpxml inside. Both paths get the same cut.

Part 3

Cut it

Tell Speaker Cut who's who, set the feel, and let it build the edit.

Name your people & assign lanes

Add each speaker, then for every person set which lane is their Video (the camera to cut to) and which is their Audio (the mic to detect them on). A webcam is both; an external recorder is audio-only; a wide or room cam is video-only, or set it as the "all speakers" shot.

Assigning each speaker's video and audio lane in Speaker Cut
Two-role lane assignment — video and audio, per lane
Two cameras on one person?Assign both as their video and star the one you want as primary. The rest become B-cams, used sparingly for variety — up to two per person. A two-shot that covers both hosts works as a B-cam for either of them.

Dial in the cutting feel

Set the pace and minimum shot length, decide how eager the switches are, choose how often B-cams get a turn, and when to fall back to the wide. This is where the edit gets its rhythm — faster for energetic banter, calmer for a formal interview.

Turn on Auto-mix (One & Pro)

With Auto-mix on, each person's mic lands in the Final Cut timeline on its own lane with volume keyframes that duck it whenever they're not talking — mic bleed and room tone gone, every reaction kept. Nothing is baked in: the keyframes are ordinary Final Cut keyframes, so you can drag any of them if a moment needs a different call.

Generate, then open it in Final Cut

Hit generate and Speaker Cut builds the auto-cut. On One or Pro it also transcribes the episode and writes AI show notes, highlights & hooks. Export the package and import the FCPXML into Final Cut Pro: the multicam, the synced angles and the cut arrive together, linked to your original media. Tweak from there — you keep full control.

Good to know

Common questions

How does Speaker Cut sync my clips? Do I need timecode?
By audio. It listens to each angle against the master at several points across the session and locks the offset, checking for drift along the way. Timecode isn't required — but if your cameras carry it, Speaker Cut keeps it, so the angles Final Cut sees reference the same source frames you'd get from a jam-synced multicam.
I recorded audio on an external recorder, separate from the cameras. Does that work?
Yes — that's the whole point of the master angle and the two-role lanes. Drop the recorder's file as its own audio-only angle, sync it against a master that hears everyone, then assign it as that person's audio lane and their camera as their video lane. Speaker Cut detects who's talking from the clean mic and cuts to the right camera.
What if an angle shows “No lock” in the sync report?
Speaker Cut couldn't find enough shared audio between that angle and the master. Check the angle actually recorded sound, and that the master can hear that person. Switch the master to a feed that hears everyone and rebuild — it only takes a few seconds.
Do I need to build the multicam in Final Cut first?
No. Drop the raw clips and Speaker Cut builds it. If you've already synced one in Final Cut, export the XML and use “I already have a multicam” instead — both paths give you the same cut.
Is my footage uploaded anywhere?
No. Syncing, detection, transcription and auto-mix all run on your Mac against your local media. Your video and audio never leave your machine.

Ready to cut your first one?

Drop your clips, mark the master, and watch the edit build itself. The app is free to try.

Download Speaker Cut → See editions