Conference layouts and speaker names

A layout is the view that each participant has of all the other participants in the meeting. The layout for a meeting is configured in advance by the administrator, but can be changed during the meeting by Host participants using SIP/H.323 endpoints that support DTMF keypad controls or by using the Connect app.

A wide range of layouts are available, including Pexip's AI-driven Adaptive Composition layout and a variety of classic layouts that have different combinations of speaker-focused and equally-sized participant arrangements. You also have the ability to define your own custom layouts.

By default, the "Large main speaker and up to 7 other participants" (1+7) layout is assigned to each VMR.

You can enable showing the names of active speakers and participants as text overlays, and there are a range of dynamic in-conference options.

There are also some differences in layout behavior depending on whether the conference is being held in a Virtual Meeting Room (including scheduled conferences) or a Virtual Auditorium.

Adaptive Composition layout

Adaptive Composition is an intelligent meeting layout with real-time automatic face detection and framing:

• Powered by AI and machine learning to give users a more natural, engaging, video-first meeting experience.

• It continuously analyzes each video feed from all participants and uses automatic face detection and framing to create an optimized view of that participant, or group of participants where there are several people in that video feed.

• Layout placement of each participant is based on a combination of face count (number of faces detected within that participant's video feed) and their recent speaking frequency, so as to allocate more space for larger groups of people.

• Independent of where the participants are and which device they are using. No end-user action or configuration is required.

• When using Adaptive Composition you may want to consider disabling speaker tracking camera features in your room systems and endpoints to avoid conflict between the different framing technologies.

When using the Adaptive Composition layout:

• A maximum of 12 video participants are shown, spread across one row of 2 large images, a middle row of 3 slightly smaller images and a bottom row of 7 thumbnail images.
• Single-screen endpoints automatically receive any presentation content as part of the layout mix (replacing some of the other video participants), rather than as a separate stream.

• Any inactive video participants are automatically removed from the video layout. This is triggered if the participant:

  • Walks away from the camera, leaving an empty chair in frame i.e. no face is detected.
  • Points the camera at the ceiling or other place where no movement is detected.
  • Physically closes the camera lid or puts a cover over the camera.

  Note that these participants still continue to receive the full video layout.

  The participant is returned to the video layout if video is restored, or if any faces or movement are detected. Face detection is applied periodically on SIP/H.323 endpoints, if audio is detected (all device types), or on mouse activity (Connect apps).

• An extended Adaptive Composition view is also available* Technology preview only where up to 23 video participants may be shown, initially across three rows (2/3/7 extending to 2/5/7 and then 3/5/7) and then across four rows (3/5/7/8) when required. This is a technology preview feature and it can only be enabled via the transforms functions in the Pexip client APIs (transformLayout (PexRTC client API) and transform_layout (REST client API)).

This layout increases backplane bandwidth and CPU resource usage. See Hardware resource allocation rules for more information.

Adaptive Composition is fully supported in all Pexip Infinity conferences, and gateway calls to Microsoft Teams and Google Meet.

Note that Adaptive Composition applies face-detection technology, but does not apply any biometric or facial-recognition technology or store any such related data.

Classic layouts

In addition to Adaptive Composition, you can select from a range of other classic layouts for your Virtual Meeting Rooms and Virtual Auditoriums. These alternative layouts do not apply any face detection or framing technologies.

Speaker-focused layouts

The set of speaker-focused layout options are:

• 1+7 (1 large main speaker and up to 7 other participants) — this is the default layout
• 1+0 (full-screen main speaker only)
• 1+1 (1 large main speaker with one overlaying participant)
• 1+21 (1 main speaker and up to 21 other participants)
• 2+21 (2 main speakers and up to 21 other participants)
• 1+33 (1 main speaker and up to 33 other participants)

Equal layouts

The set of equal layout options are:

• 2x2 (4 speakers)
• 3x3 (9 speakers)
• 4x4 (16 speakers)
• 5x5 (25 speakers)

For full details on the differences between these layouts, see Virtual Meeting Room layouts and the examples.

Custom layouts

You can design your own layouts and use them in the same way as the standard adaptive composition and classic layouts that are included by default.

Custom layouts are specified through JSON configuration files that are uploaded via themes. The theme can then be assigned as the default theme or applied to specific VMRs and gateway rules as required, to control where your custom layouts may be used.

See the custom layouts section in Rules and requirements for customized themes for instructions on how to create and add custom layouts to your themes.

By using this feature, Pexip may reserve its right to waive at least some of the obligations stated in paragraph 9 of your Service Provider Agreement related to this feature.

Custom layouts support:

• All service types (VMRs, Virtual Auditoriums, gateway calls, Microsoft Teams and Google Meet calls etc).
• Name overlay and active speaker indication, and spotlighting/pinning.
• Receiving the presentation stream as part of the layout mix.

Note that resource usage is similar to that used by the existing classic layouts.

The following features are not currently supported:

• Face detection or framing technologies.
• Multiscreen participant display (the default 1+7 display is used instead on both screens).

Three example custom layouts are included by default in the base theme (see the examples for more details):

• 1+9 (One large main speaker and up to 9 other participants, with support for receiving the presentation stream as part of the layout mix) — the image, above-right, shows an example of this (the numbers indicate the order in which participants are displayed)
• 1+12 (One large main speaker and up to 12 other participants)
• 2+8 (Two main speakers and up to 8 other participants)

These example custom layouts are available for use in all service types in the same way as the existing standard layouts.

Conference indicators

All in-conference indicators, such as participant counts, audio participants, recording indicators and locked status are shown at the top-center of the layout.

	Total participant count. This is displayed when there are more participants than can be shown in the current layout.
	Number of participants connected as audio-only.
	Number of inactive video participants and video-muted Connect app participants, who are excluded from the video layout.
	The number of participants with a raised hand. The indicator bar also alternates every 5 seconds to show a message containing the name of the first participant in the list of participants with a raised hand.
	An audio or inactive video participant's display name is shown if they start speaking.
	Conference locked indicator. Conference locked/unlocked messages are also temporarily displayed when a conference is locked/unlocked.
	Live captions / transcription indicator. Live captions on/off messages are also temporarily displayed when captions are enabled/disabled.
	Recording/streaming indicator. Recording/streaming messages are also temporarily displayed when recording/streaming is enabled/disabled.

Showing the names of active speakers and participants

Each VMR can be configured to show the names of the participants in a text overlay along the bottom of their video image.

When text overlay is enabled, the name that is shown depends upon whether the participant was invited into the conference, the type of endpoint device, and whether the participant had to authenticate, as follows:

• Automatically dialed participants (regardless of the type of endpoint that was dialed): either the ADP's configured Participant display name or, if this field is blank, the Participant alias, is shown.
• Manually dialed participants (regardless of the type of endpoint that was dialed): either the participant's configured Participant display name or, if this field is blank, the Participant alias, is shown.
• Connect app participants: the name entered by the user (in the initial Type your name here or Display name field) is shown, unless participant authentication is enabled for the VMR, in which case the name shown may instead be that provided by the Identity Provider.
• SIP endpoints: the endpoint's display name is shown.
• H.323 endpoints: the endpoint's alias is shown.
• Skype for Business / Lync clients: the username portion of the user's sign-in address is used, for example if Alice signs in as alice@example.com the text alice is shown.

The size of the text overlay varies automatically according to the resolution being received by the endpoint and the type of layout. In low resolutions the text overlays are not shown.

The default font for the in-conference display of participant names is Roboto (which cannot be changed), or if that is not available for the character set, Noto Sans.

Using the Administrator interface

To turn on participant names by default for a service:

1. Go to Services > Virtual Meeting Rooms, Services > Virtual Auditoriums or Services > Scheduled Conferences and select the service you want to change.
2. Next to the Show names of participants field select Yes.

3. You can also choose to Show name of active speaker, as well as, or instead of, showing the names of all participants.

   • When participant names and active speaker display are both enabled, the name of the person speaking is shown in green instead of white.
   • When only active speaker display is enabled, the name of the person speaking fades in and then fades out again after a few seconds.

Using Connect apps

Host participants on Connect apps can dynamically change the layout during a conference and they can also show/hide participant names. See Controlling the layout during a conference for more information.

Using local policy

You can also use local policy to enable participant names for gateway calls into Microsoft Teams or Google Meet conferences. See Example local policy scripts for more information.

Dynamic layout controls, presenting content and single/multiscreen endpoints

Participants have a range of options for dynamically controlling the layout of a conference, how presentation content is displayed, and some additional layout options for multiscreen endpoints.

The options are summarized below, but for more information about the dynamic layout controls, see Controlling the layout during a conference and Using a DTMF keypad to control a conference.

• Changing the layout during a conference: host participants on video endpoints can change the layout currently being used in the conference by sending DTMF/keypad commands to the conference, and participants on Cisco endpoints can also use the Pexip Layout Controls macro.
• Receiving the presentation stream as part of the layout mix: when using Adaptive Composition, single-screen endpoints automatically receive the presentation stream as part of the layout mix (replacing some of the other video participants), but you can choose to switch to receiving it as a separate stream.
• Multiscreen participant display: if you have a dual screen endpoint you can display conference participants across both screens, effectively allowing you to show more participants than if you have a single screen. This display mode is currently only available for a subset of meeting and layout types.

Determining if an endpoint has single or multiple screens

Pexip Infinity follows a set of rules to determine whether a specific endpoint is a single-screen or dual-screen device and thus whether it can send presentation content in the layout mix, or make use of multiscreen participant display:

1. In the first instance, if provided by the endpoint, Pexip Infinity uses the display count signaled in the contact header.
2. Otherwise, if the endpoint's user agent string is listed in the theme's vendordata.json file, the screen count as defined in that file is used.
3. In either case, the display count can be subsequently set or overridden by participant policy.

If the display count is unknown, then:

• When content is being presented, a separate presentation stream is always sent to the endpoint.
• Multiscreen participant display is not available.

For more information about the vendordata.json file, see Theme configuration.

Allowing end-users to change the default layout

A VMR's settings for its layout and display of participant names are initially assigned when the VMR is created by the administrator.

The Pexip VMR self-service portal is a separately-installable component that allows end-users to manage their personal Virtual Meeting Room without having to send requests to their administrator to change the configuration or branding of their VMR. See About the VMR self-service portal for more information.

Layout features

The following layout features and characteristics apply to all layout types (Adaptive Composition and classic layouts):

• Any video-muted Connect app participant, or a participant who joins without a camera, is removed from the video mix and is included in the inactive/muted indicator count.
• When a participant is the only device that is sending video, that participant sees a holding screen until other video participants join the conference. The holding screen indicates either that the participant is the only participant in the conference, or that all of the other participants are audio-only. The holding images are fully customizable via the theme associated with the service (Virtual Meeting Room, Call Routing Rule etc).

• Watermarks and content classifications can also be enabled/disabled and customized via the theme associated with the service:

  

  • Video watermarking overlays a small transparent image/logo onto the main speaker video during a conference.

    The default logo watermark is a white "Pexip" logo (shown here against a blue background) and is enabled by default.

  • Content classification indicators can be displayed within a conference to, for example, display the current security classification level to meeting participants. By default, there are no content classification indicators configured (and, hence, none are displayed).
• A full list of participants is available using Connect app clients.

• If you send a banner message to all conference participants, it is displayed at the very top of the screen, above any conference indicators. If classification indicators are also in use, they are combined into the same banner and are displayed on the left side of the overlay:

  --- Classification --- and/or --- Banner message ---

  ------------- In-conference indicators -------------

Features common to classic layouts only

The classic layout types (1+7, 2x2 and so on) have some minor operational differences:

• Video participants who are on hold, or who are experiencing connectivity issues, are represented by a frozen image.
• When sending presentation content to a single-stream endpoint or broadcasting via an RTMP stream (when not dialed out with dual streaming), Pexip Infinity sends the video stream of the active speaker in a small window in the upper right corner of the presentation.

Virtual Meeting Room layouts

All video-enabled participants in a Virtual Meeting Room or scheduled conference see the same layout, regardless of whether they are Hosts or Guests. Administrators can select the layout to use for each VMR, however Hosts can change the layout during the course of a meeting.

To select the layout to use by default, from the Administrator interface go to Services > Virtual Meeting Rooms or Services > Scheduled Conferences, select the Virtual Meeting Room, and then select one of the options from the View drop-down menu.

The table below lists the layout options that are available. In all cases (except for Adaptive Composition), if one or more participants are spotlighted, they take priority over the current speaker.

Virtual Auditorium layouts

In Virtual Auditoriums, Guest participants typically only see the Host participants, and Hosts see Hosts and Guests (read more). Administrators choose from separate options for the layout shown to Guests and the layout shown to Hosts (unless using Adaptive Composition). Additionally, administrators can configure the Virtual Auditorium so that when a presentation is being shown, the presenter is kept in the main view. Host participants can change the layout seen by other Hosts during the course of a meeting.

To select the layouts to use by default, from the Administrator interface go to Services > Virtual Auditoriums, select the Virtual Auditorium, and then select one of the options from the Host view and Guest view drop-down menus.

Host view

The table below lists the layout options that are available. In all cases, if one or more participants are spotlighted, they take priority over the current speaker.

Guest view

Guests in a Virtual Auditorium typically only see the Host participant(s), and can hear but not see any of the other Guests, even if a Guest is speaking. However, there are some circumstances where Guests may see other Guests, such as when there are no Hosts left in the conference, or if none of the Hosts are sending video. You can control the precise behavior via the Guests can see other guests option when configuring the advanced options for a Virtual Auditorium.

The table below lists the layout options that are available. In all cases, if one or more participants are spotlighted, they take priority over the current speaker.

Lock presenter as main speaker

You can configure a Virtual Auditorium so that when a presentation is being shown, the main speaker position always shows the presenter instead of the current speaker.

Option	Description
Yes	When a presentation is being shown, the selected layout rules are overridden so that: The main speaker position always shows the video image from the endpoint that is showing the presentation, even if others are speaking. The image that would have been shown in the main view is instead shown in the first available slot/thumbnail.
No	When a presentation is being shown, the main speaker position is voice-switched as usual.

Portrait layouts

Any WebRTC device requesting a preferred aspect ratio of less than 1 (such as 9:16) receives a layout specifically designed for a portrait display.

The layout displays a maximum of 6 people, and builds as shown in the images below when using the Adaptive Composition layout:

And builds like this in classic layouts:

It is supported by default in Connect Webapp3, and can be enabled via a suitable plugin for Webapp2 users. It is not supported in custom layouts.

Layout examples

The numbers in the layout examples indicate the order in which participants are displayed as they join the meeting, or become voice switched into view.

Changing aspect ratios

Endpoints send and display video images and presentations in various aspect ratios. The most common ratios are 16:9 and 4:3 (i.e. landscape mode), but mobile devices in particular may send video in portrait mode.

If there is a difference between the aspect ratios of the sending and receiving endpoints, then the receiving endpoint and/or Pexip Infinity may crop the image or add black vertical or horizontal borders (also known as "letterboxing" and "pillarboxing"). Pexip Infinity chooses aspect ratios that maximize interoperability and quality. How and when aspect ratios are changed depends on a number of factors, but the general principles are described below.

Main video

• Pexip Infinity: If required, Pexip Infinity will change the aspect ratio of main video by cropping the image (to avoid black bars in the thumbnail images). The exception is when video is received in portrait mode; in this case, the image is not cropped and is displayed as is. Pexip Infinity sends a 9:16 portrait-oriented stream to any WebRTC participant that requests a preferred aspect ratio of less than 1.
• Endpoints: If the aspect ratio of the main video image being sent from Pexip Infinity does not match the aspect ratio of the receiving endpoint, the endpoint will typically add a border to make it fit. Generally endpoints will not crop the image.

Presentations

• Pexip Infinity: In general, Pexip Infinity chooses the resolution and aspect ratio that best matches the presentation content and that will not result in both vertical and horizontal borders (also known as "windowboxing") on a 16:9 endpoint display. It will never crop presentations, but may add either pillar or letterboxing to send a resolution that maximizes interoperability.
• Endpoints: If the aspect ratio of the presentation image being sent from Pexip Infinity does not match the aspect ratio of the receiving endpoint, the endpoint will typically add another border to make it fit. Generally endpoints will not crop the image.