There is little doubt that IP brings greater flexibility and efficiencies of scale to broadcasters and media organisations. An IP environment offers a more flexible and scalable way to support a greater number of channels. The rate at which broadcasters are beginning to adopt hybrid SDI-IP infrastructures as well as end-to-end IP is on the rise. And the integration of IP Audio networks is a key part of that.
On the video side, we have seen the development of standards such as SMPTE 2022-6 – which sets out specifications for transport of high bit rate media signals over IP networks – laying the groundwork along with the soon to be ratified SMPTE 2110. Added to this, manufacturer-led initiatives such as AIMS and specifications such as AMWA’s NMOS are giving broadcasters and media organisation the added confidence to invest in new IP infrastructures and workflows.
The audio industry has always been more fragmented than the video side with many different networking options and connection schemes in place, perhaps because audio product manufacturers have been implementing proprietary audio over IP (AoIP) systems for a decade or more. Added to this, there is a significant proportion of audio within a facility not associated with video (monitoring, tie-lines, audio playback, microphone circuits, etc.), so audio and video signals were often kept separate.
At the beginning
The audio industry as a whole began to address this with AES67, a networked audio interoperability specification developed by the Audio Engineering Society. It describes techniques for interoperability between Audio over IP network streaming (AoIP) with the use of RTP over UDP, which is also used for video.
The specification also allows different audio networking protocols (Dante, RAVENNA, etc.) to pass audio between each other.
Additionally, AES67 aims to lay out a set of constraints to facilitate interoperability between implementations; this gives broadcasters and media organisations the confidence that any equipment they invest in will be able to talk to each other. We have seen a raft of products hit the market with AES67 support, allowing them to connect with other compliant devices within the network.
AES67 defines clocking, network transport, encoding and streaming, session description and connection management. While this is a huge plus, there remains one potential area of weakness within AES67. The specification gives manufacturers a choice of packet size and order, control protocols and number of channels making them open to manufacturer interpretation. This can cause issues when it comes to interoperability.
What is also open for manufacturer interpretation is the media transport itself, allowing them to write their own protocol for routing, level monitoring and so on. In an AES67-enabled packet you would have all the normal Ethernet and audio network protocols, MDNS, ARP, PTP. But when it comes to data payloads, they are proprietary. This is not necessarily a bad thing, as it does enable manufacturers to come up with unique features.
Furthermore, while some issues are being addressed by AES70, such as the absence of a common discovery system, there is still work to be done. AES67 aims to ensure interoperability when it comes to audio transport streams, but this doesn’t cover the “high levels” of any protocol such as routing and, ultimately key for true interoperability, control.
In essence, therefore there are “levels” of interoperability. For example, you can connect and stream AES67-enabled Dante to a device that accepts AES 67 but you would not be able to set up routing without the use of the Dante API (Dante controller software). In other words, interoperability doesn’t extend to 100 percent.
AES 70 (OCA - Open Control Architecture) is addressing those issues but uptake by manufacturers is at a very low level. In terms of systems demand, this is already there, but manufacturers need to catch up with the implementation.
There must be in-depth understanding between multiple vendors of the existing protocols to ensure interoperability. Systems integrators like TSL Systems also have a role to play. As customers seek to implement IP infrastructures – both for video and audio – SIs have the experience to help them find the best solutions. In terms of ensuring interoperability, they are able to assess the available options and make recommendations about what systems will work best together. By making use of an SI, broadcasters and media organisations have a means of going the extra mile to ensure interfaces are compatible - or if needed, find one that will work better.
Interoperability between AES67 and Dante and RAVENNA has helped and we can see evidence of this in the fact that broadcasters like Canal+ are beginning to invest in and deploy AoIP infrastructures. Earlier this year, TSL Products provided 23 MPA1 Mix Dante units for the Canal+ Factory facility to enable custom audio mixing within an IP-based workflow. Dante units were also supplied to MBC in Dubai for its all new facility at Dubai Studio City – TSL Systems acted as the systems integrator for the project.
There are some encouraging signs that we are moving in the right direction. The AIMS consortium is developing an industry standard for networked video, which includes SMPTE 2022-6, and it has been agreed that AES67 will be the specification for the audio element.
This is the sort of work that is needed – and much more of it – to ensure that vendors can help broadcasters to make the transition to IP-based infrastructures. It is by embracing interoperability and open standards that we can drive the future development of the industry.