Analogue audio cable is bulky and only provides unidirectional point to point connections. Large jack-fields take up studio capacity, especially in outside broadcast vehicles where space is a premium, and twisted pair cable is susceptible to noise, interference and frequency loss.

MADI overcomes much of the interference and frequency loss, and improves distribution by multiplexing multiple channels over a single cable. However, the multiplexed signals travel over one cable and complex broadcast centric equipment is needed to extract channels and route them elsewhere. Although fiber can be used to reduce the use of twisted pair cables, they are still new to studio infrastructures and require specialist equipment to terminate the ends.

CAT5 and CAT6 structured cabling is quite common in most TV stations, and can easily be added to studios or installed during the build. Being an IT commodity product, its cheap and durable. Bi-directional bandwidths of 1Gbps can be achieved and the full duplex capacity is one of the most misunderstood aspects of IT structured cabling; we send data in both directions on the same cable, unlike twisted pair audio and coaxial video.

Non-real time audio is distributed over IP in a similar way to file distribution. The main difference between non-real time audio and streamed audio is non-real time audio has a defined beginning and end, and has no timing constraints on its arrival. This facilitates the use of File Transfer Protocol (FTP) and Transmission Control Protocol (TCP). 

Both these protocols trade data accuracy for time taken to deliver. We know when sending a file over FTP or TCP, it will either get there intact or an error message will be sent to the user advising it couldn’t be delivered. Either way, we are certain that it was delivered or not.

Time Dependent

Real-time audio distribution is time dependent, hence the reason we used fixed point to point cabling in analogue and MADI distributions. However, with AES67 and proprietary formats like Dante, we can deliver audio within strict time frames to meet the constraints of real-time streaming.

IP-Audio has an inherent routing system built into it because of the addressing scheme used in Internet Protocol (IP), we can send streams of data to specific devices. A microphone with an IP connection can send its output to a specific destination such as an IP enabled loudspeaker. Both the microphone and loudspeaker will have their own IP addresses and the microphone destination address is set to the source address of the loudspeaker.

Multicasting is a DA

This primitive distribution clearly becomes an issue if we want to send the microphone output to many different devices. We might want to send it to a sound desk, recording device and audio monitor. Multicasting provides a simple solution to achieve this.

Multicasting is the IP equivalent of an audio distribution amplifier (DA). Destination devices use Internet Group Management Protocol (IGMP) to signal to its attached router that it wants to receive the output from a device. The microphone must use one of the assigned multicast IP addresses to signal to the routers that it is available for multicasting, but once this is done, no further configuration of the microphone is required.

The major advantage of multicasting is the microphone will only send its IP datagrams once, regardless of the number of destinations. It’s the routers that do the clever bits of distributing and copying IP datagrams to attached devices. And no extra cable is required.

IGMP Opt-In

IP routing and ethernet provides broadcast engineers with a toolkit for routing audio and video signals, but the user control of the routing is still an emerging technology.

In broadcast and radio stations audio is routed using jack fields and solid state switches, elaborate control panels are available for the user to route signals or select channels for monitoring. Although we can achieve this with IP routing, its cumbersome and dangerous to manually set IP addresses and configure IGMP opt-ins using command line interfaces on a PC.

Vendors are developing ways of turning this toolkit into a viable solution, and organizations such as AIMS (Alliance for IP Media Solutions) are facilitating this through the interoperability programs.

Monitoring on iPad

Quality of Service (QoS) is the mechanism routers use to guarantee delivery of different types of service. Using QoS, audio over IP is prioritized over other services such as HTTP to take priority when a router sends the datagram over a link, thus guaranteeing the audio IP packet gets to its destination within a preset timeframe. This is needed so the decoding equipment can re-assemble the packets in the correct sequence and turn them back into pulse code modulated (PCM) words.

IP distribution offers other benefits including monitoring. Traditionally, engineers relied on finding an audio monitoring unit to listen to the audio, or a pair of headphones and plug them into a jack-field. Audio over IP allows the engineer to stream the audio to any IP enabled device, including their mobile phone or iPad. Extrapolating this idea, we can have apps running on mobile phones and iPads to fully monitor all aspects of an audio signal, including metadata.

Automate Automate Automate

IP distribution for broadcasting within studios and transmission facilities is a new and emerging technology. We are still working out the problems we must solve, and the key to successful IP installations is having the ability to monitor a system, work out where the bottlenecks are and automatically fix them, not have somebody tweak them, but automatically fix them. The system needs to be highly automated to reduce the risk of human intervention and error.