By Andrew Starks.
Plug your monitor into your laptop, and you will expect to see video on the screen and hear audio through your speakers. Your laptop and your monitor will negotiate the most optimal settings and even allow you to override its selection with an alternate mode supported by your display, should you choose. Thanks in part to EDID, and its replacement DisplayID, this works perfectly… almost always.
However, throw in a distribution amplifier or multicast with AV over IP to drive multiple monitors, and you’ll begin to understand the pain and frustration of your average Pro AV installer. This is because EDID is designed for simple one-to-one connections, and everything beyond this is done with hacks that are not based on standards or best practices. As a result, there is a small cottage industry of donkey-knuckles (aka dongles and dohickeys) designed to fake out source devices, replacing the EDID information that the source device would normally receive with whatever the installer wants. Believe it or not, these kinds of workarounds represent the state of the art in AV over IP installations as of 2021.
For IPMX, AIMS recognized the opportunity to improve this situation dramatically by thoroughly supporting content negotiation between sources and displays, especially in multicast scenarios. With that goal in mind, AMWA is developing a new specification called NMOS IS-11 Sink Metadata Processing. Along with NMOS Receiver Capabilities (BCP-004-01), a prior AMWA group’s work, and NMOS EDID to Receiver Capabilities Mapping (BCP-005-01), IS-11 defines how NMOS Senders negotiate with one or more Receiver devices connected to displays using EDID, and eventually DisplayID.
To illustrate how it works, consider the simplest scenario: one Sender and one Receiver device. The Receiver device detects when a monitor is attached to its output using HDMI’s signaling facilities and reads the EDID information. The Receiver then maps the EDID’s timing, video, and audio profile data to receiver capabilities endpoints (a URL) on one or more NMOS Receivers. These Receivers now contain a list of all the video or audio modes supported by the attached monitor. When the user connects a Sender to one of these Receivers, the NMOS controller gets this information from the Receiver and gives it to the Sender. If the Sender can support one or more of the modes it receives, it’s configured and ready to make a connection. If a controller attempts to change the flow to something that is not supported, the Sender shuts down the flow and returns an error.
Now imagine there are multiple monitors. In the case of video, the NMOS controller retrieves the receiver capabilities of each connected Receiver. Typically, the controller would combine them into a constrained set that all monitors support. However, it could alternately use custom logic to determine how mismatched capabilities are handled or further constrain the choices, as desired by the user. With this constrained set, the Sender can pick a mode that every monitor can support or fail if it can’t.
Hidden in these examples are the gory details and exceptions that the IS-11 group must address, thanks to the rocky world of real-world EDID implementations and the ever-increasing variety of displays. Currently, they are in Phase 3, which includes validation and testing of the design with implemented contributions from Pebble, Matrox, and Macnica. Guided by the feedback received from integrators, customers, and manufacturers, the group is confident that they can deliver a robust and deterministic way of managing one of Pro AV’s thorniest issues.