An auto-tracking conference camera automatically keeps the right people in frame: it uses AI to find faces and bodies, a microphone array to locate the active talker, or both, then pans, tilts and zooms (mechanically or digitally) to follow them. There are three main modes: group framing (fit everyone in), speaker tracking (cut to whoever is talking), and presenter tracking (follow one person who walks around).
The three tracking modes
| Mode | What it does | Best for |
|---|---|---|
| Group framing | Zooms to fit everyone in the room, reframes when people arrive or leave | Most meeting rooms |
| Speaker tracking | Switches the shot to the active talker | Medium and large rooms, boardrooms |
| Presenter tracking | Locks onto one person and follows them | Classrooms, training rooms, stages |
| Multi-stream / gallery | Sends separate tiles of each person | Teams IntelliFrame, Zoom multi-stream |
How the camera knows where to look
Vision-based tracking. An AI model detects heads and bodies in the image. It handles silent participants well and is how group framing works. Its weakness: it can't tell who is talking on its own.
Voice tracking. A microphone array measures the tiny timing differences of a voice arriving at each mic to estimate its direction. The camera turns toward the sound. It knows who is talking, but reflections off glass walls can mislead it.
Combined audio and vision. The best speaker-tracking systems use both: audio to find the talker, vision to frame their face properly. Dual-lens cameras keep a wide lens watching the room while the second lens moves to the talker, so the cut to the new speaker is clean rather than a visible pan.
Mechanical vs digital tracking
- Digital (ePTZ) crops into a high-resolution wide image. Fast and silent, but zooming too far loses detail. Common in small-room bars.
- Mechanical PTZ physically moves the lens and uses optical zoom. Better for deep rooms, and needed for presenter tracking at distance.
- Dual-lens combines a wide camera with a zoom camera for smooth speaker switching.
Model picks
| Need | Camera |
|---|---|
| Medium/large room, voice tracking | AVer CAM570: dual-lens 4K PTZ with built-in audio tracking |
| Presenter tracking in a classroom or training room | Poly Studio E60: 4K PTZ with presenter tracking |
| Large Teams/Zoom room, speaker framing | Poly Studio E70: dual-lens smart camera |
| All-in-one medium/large room | Logitech Rally Bar or Rally Bar 2 |
| Two-lens AI framing PTZ | AVer CAM550 |
| Large rooms and halls | Logitech Rally AI Camera Pro |
| Small room, group framing | Logitech MeetUp 2, Jabra PanaCast 50 |
For more picks, see best auto-tracking conference cameras.
Getting tracking to work well
- Mount the camera at eye level, centered on the table, so faces aren't foreshortened.
- Keep strong backlight (windows behind people) out of the shot; it hurts face detection.
- Treat glass walls if voice tracking jumps to reflections.
- Choose a less aggressive tracking speed if the camera hunts during fast back-and-forth discussions.
- Turn tracking off for presentations where a fixed shot is better, and save presets instead.
Brand-specific steps: AVer auto tracking and presets, Teams IntelliFrame, Zoom Rooms Intelligent Director.
Can a camera follow you around the room?
Yes, that's presenter tracking. It works best when one person moves in a defined zone like the front of a classroom. In a meeting with several people, use group framing or speaker tracking instead; presenter tracking will lock onto one person and ignore the rest.
Want a tracking camera matched to your room? Share the room size and use and we'll recommend a model.
Related guides
Outfitting several rooms, or reselling? Get a volume quote or apply for wholesale pricing →
