Frequently asked questions

Are SuperMAC and HyperMAC going to comply with the IEEE 802.1 AVB (Audio/Video Bridging) Ethernet standard?

No, we see a clear distinction between the frame-based approach employed by both SuperMAC and HyperMAC utilising the IEEE 802.3 Ethernet Physical Layer, and packet-based Ethernet IP audio systems that are already compliant with the AVB standard, or are working towards compliance. IP-based systems collect the audio samples to be transmitted into data packets and then recover and reassemble the packets back into continuous audio streams at the receiving end. These systems have an advantage in that conventional IT hardware such as Ethernet routers and switches can be utilised to build networks and that the Ethernet ports on computers can be used as audio interfaces without additional hardware. It should be noted though that to guarantee uninterrupted audio performance, good quality “managed” switches are required, together with the expertise to set them up correctly. If control data is to share the same network, then it is necessary to establish Quality of Service parameters (typically by separating the network into several separate Virtual Local Area Networks or VLANs). This protects the audio from interruption by limiting the control bandwidth, but again requires considerable knowledge to set up for each individual type of switch. Also, the switches only provide routing for whole Ethernet packets – it is not possible to separate a single audio feed from the other signals in the same packet. Some IP-based systems can provide a few channels of audio at low latency, but only by making very inefficient use of the available bandwidth. This is because the Ethernet preamble, header, error check, and idle space are a fixed size (38 bytes) – if you only send a few bytes of audio data you use up most of the bandwidth sending headers.

SuperMAC and HyperMAC use a very different concept. They use the physical layer only of the Ethernet technology – the cables and transceivers at each end. However the signals that are sent are not based on the IP protocol – instead a deterministic protocol is used that is essentially a time-division multiplex (TDM) at the hardware level. Put more simply, there is a fixed pattern of data sent in a repeating fashion – there is no need to send and decode complex header information because the link is only designed to send audio data from point to point, rather than generic computer data across the Internet. The disadvantage of this approach is that conventional IT routers and switches cannot be used, and dedicated routing hardware is required. Whilst this is to some extent a disadvantage, it does mean that they can be a much better fit to the actual requirements of an audio network. Typically such a router provides individual one-to-many routing for every discrete audio signal, and incorporates a conventional Ethernet switch for the control data, which is treated separately from the audio. The routers are also typically used as the clock sources or slaves in the system, allowing the synchronisation strategy to be extremely flexible, and avoiding the need for a separate word-clock distribution system. Because such routers are designed specifically for professional audio applications they can also use ruggedised connectors such as the Neutrik Ethercon, rather than the fragile RJ-45 connectors typically used on commercial Ethernet switches. A major advantage is that no specialist networking knowledge is required to set up such a router – it is merely a matter of connecting up the AES50 connections and selecting the clock source.

The key differentiators that SuperMAC and HyperMAC have – the simultaneously high channel counts and fixed, very low latencies per link, plus ease of configuration – cannot be offered by IP audio systems due to their packet-based approach. Both frame-based and packet-based approaches to transmitting audio over Ethernet have their advantages and disadvantages, and we see them as complementary.

< Back to frequently asked questions