This guide is a primer to setting up Audio Tracking with Seervision using Shure MXA920 panels. The end result should be that based on talker position data from the MXA920, Seervision automatically steers PTZ cameras to the speaker location and frames them.

There are a few individual parts to get connected, here is how the process breaks down:

1. Set up and install the Shure MXA920
2. Set up and install the Seervision server + PTZ camera
3. Connect the two via Node-RED (the main topic of this guide)

There’s some requirements up front that need to be met before you can proceed:

• The Shure MXA920, Seervision server and PTZ camera all need to be able to communicate with each other over IP. This means that they must all be located in the same IP subnet, or that the respective subnets are able to communicate with each other.

• During installation and commissioning only: the Seervision server(s) must be online and remotely accessible, so we can guarantee they are up to date before commissioning kicks off. They can be kept offline during normal operation

• Steering of the PTZ camera only works if it is a PTZ camera supported by Seervision, as listed on our integrations page.
• The Shure MXA920 must be correctly configured by accredited Shure representatives, in order for Seervision’s steering to be accurate.

Before proceeding with the rest of this guide, make sure you have already properly configured the Seervision server to correctly control the PTZs you intend to steer. If you’re not sure where to start, our Getting Started guide will walk you through it.

Only proceed with the next section once you have successfully completed the above, as this is a prerequisite to the next steps.

In this guide, we won’t be covering how to set up the Shure MXA920 for optimal audio, that’s best left to your Shure representative and the Shure MXA920 manual. The outcome here should be that you have an MXA920 that is ready to go and correctly configured. You will also need the IP of the Shure panel, as we’ll be using that to connect to the panel’s API, needed for talker position data.

Shure MXA920 Configuration Notes:

• The area of interest within which you wish to pick up audio must be defined as a covered area in the Shure designer.
• To simplify the calibration later on, make sure that the mounting orientation of the microphone panels and the PTZs are either parallel or perpendicular:
  • For square panels, this means that the edges of the panels are parallel or perpendicular to the orientation of the PTZs
  • For round panels, this means that the LED is at a 45 degree angle to the mounting orientation of the PTZs
• For optimal performance Automatic Coverage must be enabled for these areas. While our system can work Automatic Coverage disabled, it negatively impacts speaker tracking performance.
• If there are loudspeakers in the room, their signal must be fed into the AEC Reference of the Shure MXA920

To correctly steer PTZs based on talker position input, we need to measure all the distances between the microphones and the PTZs. To measure these, you will need the following:

• The Seervision servers should be on a network that is connected to the internet, so they are remotely accessible by Seervision (necessary only for the duration of the setup). Without remote access, calibration is not possible.
• A top-view schematic of the room that shows the microphone location and PTZ locations
• Our printed markers, one per microphone (sent to you as PDF)
• A laser meter to measure additional distances where necessary

In many cases, we can automatically establish the distances by attaching markers to the microphones, and steer the PTZs so they can see the markers. To prepare for automatic calibration:

• Print one marker per microphone that needs to be calibrated (we will send you these as PDF via email)
• Measure the length of the line on the front of each marker: it should be 25 cm. If it is not, let us know what the actual width of the line is, so we can scale accordingly
• Read the instructions on the back of the marker, it is very important they are placed on the correct corners of the microphone panel, as shown in the instructions!
• Once all markers are in place, let us know and we can start the calibration

Sometimes, the layout of the room or positioning of the markers does not allow for automatic measuring based on the markers. In that case, for each PTZ, you will have to measure using a laser distance meter (to approximately 5 cm precision):

• The X and Y distances from the center of the microphone panel to the lens of the PTZ (see schematic below)
• The Z distance between the PTZ lens and the ceiling to which it is attached

With the final results of the calibration, it should be possible to draw a schematic that contains the distance information. The below schematic is just an example, but it illustrates all the distance information that should be available at this point.

To validate the measurements are correct, the next step is to check that the PTZs are moving to the correct location in the room by setting up a conference call, and having somebody on-site speak from different positions. Depending on the observed accuracy, the size of the room, and the amount of cameras, this step can take multiple hours, also depending on the parameters that we need to tune and the use-case of the room.

In nearly all situations, you will want to connect your venue control panel, such as Q-SYS, AMX or Crestron, to our audio setup, to turn it on, off, or know when to switch video feeds. To be able to configure this, you will need to know a minimal amount of Node-RED and JavaScript. In this case we will provide you with access to the Node-RED instance running the Seervision room control.

The following paragraphs are purely intended as an overview of the programming that will be required to control our audio flows. It is therefore intentionally a high-level description. If you need more detail on getting started with Node-RED and what our integration looks like, check out our Developer Guide.

Thanks to the flexibility of Node-RED, it is entirely up to you to decide how to communicate with our room control. Node-RED provides multiple options, such as UDP sockets, TCP sockets, HTTP calls or WebSockets. You are free to decide which of these you prefer – all you will need to do is implement it correspondingly in Node-RED (which is usually straightforward, as many protocols are supported out of the box).

To enable or disable our audio tracking flow, we will provide you with hooks that you can activate to start or stop tracking. Additionally, you will need to send a “heartbeat” every 2 seconds with the current state (on or off) so that we can correctly synchronize in the case of a network interruption or reboot.

Our Node-RED flow will provide hooks that inform you when a camera should go live. All you will have to do in Node-RED is write a node to consume this information and send it to your control software.

Especially with the pandemic, a recurring question we get is “Can I operate the SV Suite remotely?”. The answer’s a resounding yes, but it does come with a few caveats and, depending on the implementation you choose, will require some LAN networking adjustments. In the sections below, we’ll attempt to walk you through the various “flavours” of remote operation, and you can decide which one would suit your workflow best.

At its most basic, the UI of the SV Suite is just an interface hosted locally on the Seervision server, and you access this interface through a browser. Any attempt to remotely control it will concretely mean that you’re looking for a way to remotely access the server and the interface it serves. Both of the approaches below will walk you through how to get that done without needing to install anything on the Seervision server itself – all of the routing is done using your own available infrastructure.

The easiest way is using any remote desktop software. There’s lots of software that does this – at Seervision, we’re pretty partial to Parsec, as it plays nice with discrete GPUs and is one of the smoothest implementations we’ve seen so far. Of course, there’s plenty of alternatives that will work just fine.

The setup

Connect the Seervision server and the PTU as you usually would, and use an on-premise device (laptop or desktop) to confirm everything works as expected. Once it’s up and running, install the Remote Desktop software of your choice, and leave the machine running (you can turn off the display, but don’t let the machine go to sleep).

Once remote, log in to the machine via your Remote Desktop software, and you should be able to operate the system from wherever you are!

With the VPN approach, the goal is to be able to access the Seervision server’s IP by routing your remote traffic via a VPN that is on the same network as the server.

The setup

The first step is to set everything up normally, and to make sure that the machine that will be the VPN endpoint is already able to access the Seervision server’s IP and UI as normal.

The second step is to configure this machine as the VPN endpoint. Most VPNs will work, though we like the open-source Wireguard the most as it’s probably the lowest-latency one we’ve seen, and has seen rapid adoption across the digital world. Alternatively, our go-to is ZeroTier which is also a wonderfully performant alternative. Note that ZeroTier would require additional configuration to act as the bridge network (as that’s not its default intended use-case).

We don’t have an in-depth guide to offer here, as the configuration of the VPN heavily depends on the available networking infrastructure and what exactly it is that you want to do. Of course, we’re happy to help out and offer some ideas, just shoot us an email!