How Does Gallery View Work?

Lync Server 2013 Gallery View works great, and looks great on the client side, but how does it work? The addition of this functionality has introduced another level of intelligence to the Lync Server A/V Conferencing Server. The A/V Conferencing Server in Lync 2010 would actively manage all video streams from participants, but it still was just a video switch and would not encode or decode media. The easiest way to think of this was in the form of a simple on or off switch. Lync 2010 was limited to a single active speaker, and there were only two active video streams in a conference at one time. The current active speaker video stream would be sent to all participants in the video conference, and the previous active speaker video stream would be sent to the new active speaker. Although all participants were sharing their video, they would not be actively sending a video stream until they were designated the active speaker. This helped to reduce the bandwidth requirements and processing load on the A/V Conferencing Server. This is also why there was a slight delay in the video switching to the active speaker.

In Lync 2013 the intelligence is increased to include not only up to five different video streams, but also multiple resolutions and frame rates. The H.264 SVC codec handles most of this functionality. Each video endpoint is responsible for providing the A/V Conferencing Server what it is capable of sending and receiving for video streams. The A/V Conferencing Server is capable of dynamic optimization during a conference, including the following situations:

• New Participants Joining—When a new participant joins the conference, the A/V Conferencing Server will automatically adjust video streams that are sent and received based on participants who enter and leave the meeting.

• Bandwidth Changes—For each participant, if the network conditions change at all, the A/V Conferencing Server dynamically adjusts video streams to accommodate.

• Video Window Size Changes—Each participant has the capability to view video streams in different sizes. The A/V Conferencing Server dynamically adjusts the video streams that are sent and received by each participant however they want to see it. For example, if Randy has his video full-screen, but Alex has his video smaller as part of the presentation window, Tom’s video stream will be sent to the A/V Conferencing Server as a full-screen (HD) stream, but Alex will receive only the reduced video size. If Randy was to shrink his video window and HD video was no longer needed from Tom, then Tom would stop sending HD video to the Conferencing Server.

• Forward Error Correction—Forward Error Correction is built into the Lync Server and Lync Client media codecs. This is the capability to rebuild media streams in packet-loss scenarios. Essentially, if packet loss is detected, FEC will kick in and endpoints will start sending redundant packets. The receiving endpoint then has the intelligence to rebuild the media stream with the packets it receives. The A/V Conferencing Server is also able to use FEC for video conferencing scenarios.

Lync Server 2013 video conferencing utilizes much of the same intelligence described previously for non–Gallery View conferences. The only difference between regular video conferences and Gallery View conferences is that there are more than two streams to manage for a conference; other than that, the way the A/V Conferencing Server functions is the same.