Multiview is a production technique where multiple camera feeds or source streams are combined into a single tiled video output. Instead of broadcasting one camera at a time, a multiview output shows all feeds simultaneously, arranged in a grid or picture-in-picture layout.
The Live API handles multiview composition through two resource types: staticMultiviewEvents and staticMultiviewChannels.
Which resource type to use
Both resource types produce the same multiview output. The only difference is how they are billed and how long they run.
| Resource | Billing | Use when |
|---|---|---|
staticMultiviewEvents |
Per-minute | Sports matches, live shows, any broadcast with a defined start and end |
staticMultiviewChannels |
Monthly flat fee, charged immediately on start | 24/7 production infrastructure that runs continuously |
If for example you are running a 90-minute football match on a Saturday, use staticMultiviewEvents. If you are building a permanent control room feed that is always on, use staticMultiviewChannels.
How composition works
When the Live API receives multiple input streams for a multiview resource, it routes each one into a processing template that composes them into a layout and then encodes the result. The template is where the composition happens. The Live API resource is the control plane: it tells the platform which streams to accept, which template to use, and where to send the output.
What a pin is
Processing configs expose named input slots called pins when the actual config declares them. A pin is a labelled placeholder that says "a video feed goes here". A multiview config might expose four pins. When you create a multiview resource, each input includes a transformInput value that names its pin.
This is how the platform knows which camera feed to put in which position. The template controls what happens at each pin (size, position, any overlays). The transformInput value on your resource input connects a specific source stream to that pin.
Pin names are defined by the template and retrieved by reading the template's status.inputPins[].name. They are not standardized across templates, so always read them from your specific template before creating a resource.
Configure composition and encoding
Use an encodingLive template with the multiview layout and encoding settings.
List the encodingLive presets available to your project and choose one configured for multiview. Create a customer config from the selected preset, then reference that config in spec.transform.encodingLive. See Manage templates.
The setup sequence
Before you create a multiview resource, you need to set up these dependencies in order:
- A source for each input (Media API): each camera feed or ingest point needs its own source resource. See Content, sources, and destinations.
- An encodingLive template (Templates API): required for all multiview resources. List the current presets, create a customer config from one, then retrieve that config's pins. See Manage templates.
- An MK.IO asset (Media API): live resources write encoded output into an asset. See Assets.
For a complete walkthrough that covers all of these steps in order, see Multiview setup walkthrough.
transformInput also applies to single-input Live resources. For multiview, each input normally binds to a different pin. Resolved inputs, source Active state, and event-level health do not prove that each tile receives media. Check every encoder's contribution and verify that each expected tile updates in the composed picture.
Guides
End-to-end walkthrough
Follow a complete setup from sources and templates through to a running multiview event.
Static Multiview Event
Reference guide for creating and operating a per-minute multiview resource.
Static Multiview Channel
Reference guide for creating and operating a monthly-billed multiview resource.