The XML file /usr/local/freeswitch/conf/autoload_configs/conference_layouts.conf.xml contains the additional configuration where are specified the different ways to organize and populate the screen real estate during video conferences.

The moderator can change a video conference layout dynamically, during the conference, to adapt it to changing situations: during presentations you may want to have a little video stream of the presenter's face superimposed ("overlapped") to the screen she is sharing, so the effect is a little talking head as a picture-in-picture over the PowerPoint slides.

Layouts are defined by a name, the coordinates that divide the screen into areas, and the proportion the video stream will be scaled for each area. All numbers, both of coordinates and of scaling factors, are in the range 0-360 and proportional to the canvas. So, a layout that divides the screen real estate into 3 rows of 3 equally sized areas (eg, a 3x3 disposition) will work the same both with a canvas of 1920x1080 and with a canvas of 800x600.

<layout name="3x3"> 
<image x="0" y="0" scale="120"/> 
<image x="120" y="0" scale="120"/> 
<image x="240" y="0" scale="120"/> 
<image x="0" y="120" scale="120"/> 
<image x="120" y="120" scale="120"/> 
<image x="240" y="120" scale="120"/> 
<image x="0" y="240" scale="120"/> 
<image x="120" y="240" scale="120"/> 
<image x="240" y="240" scale="120"/> 
</layout>