Camera Setup

The Camera Setup page lets you select an intelligent camera mode and configure how the collaboration bar frames meeting participants. Choose from five modes depending on room size, participant count, and desired meeting experience.

Mode Description
Manual Pan, tilt, and zoom are controlled directly from the in-room controller. No automatic framing.
Auto Framing Uses vision detection to identify faces or bodies and automatically crops the image to frame everyone in the room.
Active Speaker Framing Combines audio localization from the microphone array with vision to steer and frame the active speaker.
Tile View Detects faces from the full field of view and places each participant into an individual tile cell. Supports up to eight participants.
Active Speaker Tiling Shows up to four of the most recent active talkers in equal vertical slices, combining speaker detection with tiled presentation.

Auto Framing continuously adjusts the camera crop so all detected participants remain in frame.

Parameter Description
Transition Speed Controls how quickly framing adapts when a new person enters or exits the frame.
Frame Padding Controls the tightness of the crop. Tight focuses closely on faces; Wide shows more of the surrounding space.

Active Speaker Framing uses the microphone array to localize the talker and steers the camera to frame the person speaking.

Parameter Description
Transition Speed Controls how quickly framing adapts when a new person enters or exits the frame.
Frame Padding Controls the tightness of the crop. Tight focuses closely on faces; Wide shows more of the surrounding space.

Tile View identifies each participant's face from the full field of view and presents them in individual cells.

  • The maximum number of displayed participants can be limited (up to 8).

  • Once positioned within a tile, the system continues tracking the participant's face within that cell, even if the person stands up or shifts position.

Active Speaker Tiling

Active Speaker Tiling presents up to four of the most recent active speakers in equal vertical slices.

Parameter Description
Room View When enabled, shows a full-field view of the room alongside the tiled speakers, providing context for participants who may not be among the recent active talkers.

Video Fence defines a visual region of interest for the camera. Intelligent camera modes, like Auto Framing, Active Speaker Framing, Tile View, and Active Speaker Tiling, will only detect and frame participants within the defined region, ignoring activity outside it.

Setting Room Dimensions

  1. Enter the Room Width and Room Length to match the physical dimensions of your meeting space.

  2. Select your preferred unit of measurement (feet or meters).

Defining the Region of Interest

Draw or adjust the region of interest in the Top View of the room diagram. The camera will only use this region when performing face detection and framing calculations.

Camera Preview

When no Teams call is active, you can enable a live camera preview. The 2D region of interest is projected into the camera field of view so you can visually confirm alignment with the intended meeting area.

Note: The live preview captures approximately one frame per second and is intended for commissioning verification rather than full-motion monitoring. If a Teams call is active, camera preview is unavailable.

Example Use Case

A conference room with glass walls facing a hallway can use Video Fence to mask out the hallway area. People walking past the room will not trigger auto framing or appear in tile view, keeping the camera focused exclusively on meeting participants.

Related Topics