In reading this article, this phrase jumped out at me:
Wouldn’t it be great if the camera automatically detected the speaker? What if the camera could detect the number of faces in the room, and adjusted its frame of view to accommodate accordingly?
I respectfully, yet completely, disagree. If we're discussing any sort of user experience, I would wager that a constant fixed camera is ideal. That being said, in all transparency, I have developed several fixed camera products over the years so as usual take my opinion with a grain of salt. I look at it like this; video conferencing is a meeting. That's it. We're having a meeting. Period. It may be over video and we may be in different locations, but breaking it down to its simplistic essence, we are just having a meeting. Entering dynamics like a moving camera breaks the entire flow of the meeting. I've been in enough meetings where someone (or an AI) keeps fiddling with the camera. You end up concentrating on where the camera is going to go next and forgetting what you're talking about and sometimes who you're talking to as that person can end up out of frame.
A moving camera absolutely calls attention to itself and more often than not, it's negative attention.
As usual, I welcome all dissenting opinions.
PS: I will say this - Eagle Eye Producer does an infinitely better job than Director with the above issues.