RADVision MCU-323, User Manual

Page: 17 / 185

MCU-323 Functionality
1-3
Voice–activated Video Switching
The MCU-323 performs automatic switching of the video image viewed at each terminal
according to the voice level of each participant. Whoever is speaking the loudest is the
participant that is viewed. An adjustable minimum time interval for voice activation defines
the minimum time interval before video images are switched. Normally, all the participants
see the active speaker on their screen. The active speaker sees the previous active speaker
on his screen, or himself (configurable). If the conference chairman locks the image to one
participant, video switching is suspended and all conference participants see the participant
that has been locked.
Continuous Presence
Continuous Presence allows you to view a number of participants in one screen
simultaneously. The MCU-323 provides a Continuous Presence screen that is composed of
four quadrants of equal size, each displaying the video image of a different participant. The
upper left-or lower right hand quadrant (configurable) contains a voice–activated image that
updates as attention shifts from one speaker to another. The remaining three quadrants
contain the images of three other participants in the order that they join the conference on a
"first come first serve" basis. Once all four quadrants are filled, only the loudest participant
in the conference can replace the voice–activated quadrant. If the loudest speaker is already
present in one of the remaining three quadrants, no switching occurs and the voice–activated
quadrant remains as before. If one of the other three quadrants is vacated by disconnecting
from the conference, that quadrant is filled by the next participant in line in the order of
joining the conference.
The Continuous Presence image is formed by unifying (up to) four video images into a
single picture. This is achieved by the MCU-323 unifying four QCIF signals and sending
the resulting CIF signal to each participant of the conference. Because of this signal mixing,
the outgoing bandwidth from the MCU-323 is four times that of the incoming bandwidth.
This factor should be considered when setting up and allocating network resources for a
conference using Continuous Presence. A Continuous Presence conference is set up by
dialing a dedicated service that is defined for Continuous Presence. When a conference is
set up across a Gateway or VIU–323 connection, participants at the remote end can only
receive a single picture image, the voice–activated image of the active speaker. The
incoming image from a Gateway or VIU–323 connected participant can be mixed in the
Continuous Presence screen in the same manner as a participant on the LAN.
Chair Control Locking of the video image (via the MCU-323 web interface) can be used to
suspend the voice–activated video switching. When locked, the voice activated picture
quadrant remains the same, and video switching is suspended. Additional participants can
enter the CP screen only if one of the quadrants is vacated by disconnection from the
conference. In cascaded conferences, the Locking feature behaves differently for Master and
Slave units, see
Cascaded Conferences in this chapter.