Multicam Workflows - CBA Glossary
📖 Glossary

Multicam Workflows

Multicam workflows—coordinating multiple camera feeds through Vision Mixing, switching, graphics, recording, and simultaneous delivery in real time—are the backbone of professional live production. They're also one of the most complex elements to execute reliably at scale.

A multicam event requires that every camera has synchronization (genlock for SDI cameras, or timecode for IP-based feeds so frame alignment is exact). If camera 1 and camera 2 aren't genlocked, switching between them creates an imperceptible but jarring micro-stutter as frame timing realigns. For distributed productions like EWC with five separate arenas, maintaining genlock across venues requires careful planning—usually a central genlock signal distributed via fiber or GPS-derived timing.

The vision mixer (technically, a video switcher like BlackMagic ATEM or vMix) ingests all camera feeds, selects one as primary at any moment, and outputs the selected feed to encoding and delivery systems. This might sound simple, but the mixer is also typically handling graphics layers, color correction, slow-motion replays, and downstream keying. A single mixer failure takes down the entire broadcast.

For EWC, we ran redundant vision mixers in each arena—the primary ATEM handled all mixing, a backup ATEM stood ready, and an automatic failover system would switch to backup if the primary failed. During the three-month tournament, we had one primary mixer fail mid-event, but viewers never saw disruption because the failover was transparent.

Recording multicam streams adds another layer. You're simultaneously recording every camera feed (archival quality), recording the mix output (delivery quality), and sometimes recording separate audio stems so post-production can remix audio later. This means simultaneously writing 10-15 gigabytes per minute to storage during a live event. Storage, networking, and backup strategies become critical operational issues.

Multicam coordination also requires a communication system between the vision mixer, camera operators, and production control. The mixer needs to know if a camera is about to move so they can time the cut appropriately. Cameras need feedback about what's on air so they know if they're "hot" (currently broadcasting) or "standby" (not currently visible). This is managed through tally lights (physical lights on camera heads) and Comms Systems where the mixer coordinates with operators.

FAQ
How do we keep multiple cameras synchronized?
+
Genlock for SDI cameras (a timing signal distributed from a central reference) or timecode sync for IP cameras. During Site Surveys, we run a genlock reference from the main control room to all camera locations. For distributed events like EWC, genlock comes from a central timing reference (often GPS-based) ensuring all arenas stay synchronized.
What happens if we lose one camera feed during a live event?
+
The mixer switches away from that camera and continues with remaining feeds. If the lost camera was critical (only head shot of the speaker), we have backup—either a second camera covering the same position or a prepared graphics slide to hold on air. We design multicam setups with redundancy so no single camera failure creates visible disruption.
How many cameras can one vision mixer handle?
+
A BlackMagic ATEM can ingest up to 28 SDI or HDMI feeds simultaneously (varies by model). vMix can handle 30+ feeds depending on computer CPU. The limit isn't usually feed count but complexity of graphics and processing. A mixer handling 8 cameras plus graphics and color correction is easier to operate than one handling 20 cameras. We design mixer load based on operational complexity, not just feed count.
Do all cameras need to be the same model?
+
No, they don't need to be identical, but they need compatible outputs (all SDI, all HDMI, or standardized IP). Mixing camera types with different color science can create mismatched appearance. For EWC, we used different camera models for different purposes (stadium wide shots, player close-ups, crowd cameras) but chose models with compatible color profiles.

Need help with multicam workflows?

Book a Discovery Call