Multicam sync in DaVinci Resolve: the complete workflow

Two different things called “multicam sync”

Before touching a menu, separate the two operations Resolve offers, because they solve different problems:

• The multicam clip — a single Media Pool object that contains every angle aligned on a shared internal timeline. You cut it live in the multicam viewer, switching or cutting between angles as the timeline plays. This is what you want for a talk, a panel, a concert, an interview covered from three cameras.
• Timeline auto-alignment — you drop clips onto separate video tracks, select them, and let Resolve slide them into sync with each other in place. This is what you want when you are not cutting a live switch but simply need a two- or three-camera scene laid out and aligned before you edit conventionally.

Both use the same underlying sync engine. The multicam clip is the right default for anything genuinely multi-angle; auto-align on a timeline is the escape hatch for odd cases and for repairing a group that synced wrong.

Building a multicam clip

Select every clip that belongs to the same event in the Media Pool — all camera angles and, if you have it, the sound roll — then right-click and choose Create New Multicam Clip Using Selected Clips. The dialog is where the real decisions happen. The field that matters most is Angle Sync, and it offers:

• Timecode — align angles by their embedded start timecode.
• Sound — align by analysing each angle’s audio waveform.
• In / Out — align to a manually placed in or out point on every clip (the modern equivalent of a clap you mark by hand).
• Manual — no automatic sync; you nudge angles into place yourself.

Set the multicam clip’s frame rate and starting timecode to match your project, name it after the scene, and create it. Everything after this is about picking the right sync method and then verifying it.

Sync by timecode

If your cameras and recorder were jammed to a common clock, timecode is the fastest and most robust route: Resolve reads each clip’s start timecode and places it on the shared timeline at that moment. Overlapping time ranges become aligned angles with no waveform analysis at all.

Three things decide whether it works cleanly:

• Every device must actually carry timecode. Professional cameras jammed from a sync box, or a recorder generating master TC, will. A mirrorless body with no timecode input will not — unless the clock was fed into one of its audio channels as LTC, which is a different route entirely (see below).
• Frame rate and drop-frame must agree. The same wall-clock instant maps to different frame numbers at 25 and 29.97 fps, and drop-frame counting shifts it again. A mixed-rate shoot where one camera was left in the wrong mode will sync with a constant offset that looks like sloppy jamming but is really a counting mismatch. If you are fuzzy on how the count works, our explainer on how LTC and SMPTE timecode actually work covers the frame math.
• Re-jam gaps show up as steps. A camera that drifted and was re-jammed after a battery swap carries two different offsets across the day. Timecode sync is only as good as the last jam.

When timecode is present and clean, it beats every other method for speed and is accurate to the frame. The residual sub-frame error is expected and is what waveform refinement, if you use it, cleans up.

Sync by sound (waveform)

No timecode? If every camera recorded some audio of the scene — even a tinny on-board scratch mic — Resolve can align by sound. It cross-correlates the angles’ audio: conceptually, it slides each recording against a reference and finds the offset where the waveforms agree best, which shows up as a sharp peak at the true sync point.

What makes the difference between a clean lock and a wrong one:

• The scratch audio must hear the scene. A camera thirty metres back records mostly its own surroundings and reverb; the correlation broadens and can lock onto an echo. Speech, claps, and transients give sharp peaks; continuous traffic or wind noise correlates weakly everywhere.
• Repetitive audio is a trap. Music playback, a metronome, looped announcements — anything periodic produces several near-equal peaks, and the matcher can pick the wrong one with full confidence. These groups need a human eye on the result.
• Silent angles cannot be sound-synced. A camera with audio disabled, or a drone, has nothing to correlate. It needs timecode, a manual in-point, or hand placement against the action.

Done well, waveform sync is sample-accurate — tighter than frame-accurate timecode — which is exactly why a strong workflow uses timecode for the coarse pairing and sound to refine it. The full hierarchy of sync methods and their failure modes is laid out in how to sync external audio to video without a clapper.

Folding in dual-system sound

Most multicam edits also want the good audio — the boom and lavs from a dedicated recorder, not the cameras’ scratch tracks. In Resolve you bring the sound roll into the multicam clip as one more angle: include it in the selection when you create the clip, and it syncs alongside the cameras by whichever method you chose. Then, in the timeline, you switch audio to the sound-roll angle while cutting video freely between cameras — the multicam viewer lets you switch picture and sound independently for exactly this reason.

The recorder’s files carry more than audio: the BWF bext time reference and iXML scene/take names are what make them line up and label themselves. If you have not met those chunks, the editor package guide walks through what a sound roll actually contains and why its metadata is worth protecting through ingest.

Angle names and order

A multicam clip is only pleasant to cut if the angles are labelled. Resolve reads the Angle metadata field, falling back to reel or clip name, to lay out and title the viewer tiles. Set the Angle field per camera before you build the clip — A-cam, B-cam, Wide, Drone — and the multicam viewer shows a named, ordered grid instead of a wall of anonymous thumbnails. Consistent reel names across a shoot also keep angle assignment stable if you ever rebuild the group. This is metadata discipline that belongs upstream of Resolve, at ingest, not something to fix one clip at a time in the edit.

The Sync Bin on the Cut page

Resolve’s Cut page adds the Sync Bin, which leans entirely on good timecode. It shows every angle that overlaps the current playhead position, already aligned on a shared clock, so you can audition and cut angles at a given moment without pre-building a multicam clip. It is fast for run-and-gun and news-style turnarounds — but it is only as trustworthy as the timecode underneath it. Jammed cameras make the Sync Bin feel magic; unjammed cameras make it show angles that are not really where it thinks they are.

Refining sync and catching drift

No automatic pass is above verification. Two failure modes deserve a direct look:

• A constant offset — one angle a few frames early or late for its whole duration. Open the multicam clip, select the offending angle, and nudge it frame by frame until a plosive lines up (lips closed on the consonant, audio on the release; the eye catches a one-frame error faster than the ear). A wrong-mode frame rate or a bad jam usually looks like this.
• Drift — in sync at the head, out at the tail, by an amount proportional to the take’s length. This is the audio sample clock, not the edit. A consumer camera off by 50–100 ppm drifts 3–6 ms per minute, so a one-hour single take can end a third of a second out even after a perfect head sync. The classic 48,000-vs-48,048 Hz pull-down mismatch drifts a blatant 60 ms per minute. The fix is a small speed/resample adjustment on the drifting track, not a re-sync — and it is far cheaper to catch before the clip is buried in a multicam group than after.

Verify every group by playing the first and last ten seconds of its longest angle with both the camera scratch and the sound roll audible: phasing means you are within milliseconds, a distinct echo means you are not. A group that reports zero problems across a whole shoot day deserves suspicion, not relief — real days contain a camera that was never jammed.

Where Launchr Post fits

Resolve syncs beautifully when the material arriving in the Media Pool is already honest: timecode converted through the right frame rates, LTC decoded off audio tracks, angles named, sound rolls paired to picture, and drift measured on the long takes. That preparation is upstream work. Launchr Post does it on your Mac before Resolve opens — pairing by file timecode first, decoding LTC where a camera carries it, refining to sample accuracy against the waveforms, grouping multicam angles, writing consistent Angle and reel names, and flagging the clips that drifted or could not be matched with confidence. You import a scene that is already grouped and labelled, and build the multicam clip on material you can trust.