With all the speculation over the Internet of Things (IoT), we wanted to have a look at some networked devices which are already pushing the boundaries of transfer speed, automatic discovery and global data transfer – network audio devices.
Since the audio speaker was invented in 1874 (by Ernst W. Siemens, although Alexander G Bell was the first to patent audio transmission in 1876) the electronic transmission of sound has worked in the same basic way – low voltage electrical fluctuation carried along a wire. That changed in the last few decades, when in the 80’s analogue-to-digital convertors took over audio recording. They work by recording analogue sound as digital data and transmitting the data to a computer. Any data transfer method with sufficient bandwidth can connect converters of a sound card to a computer – USB, firewire, even (very recently) wifi.
But the breakthrough which ties audio technology to the IoT explosion is network audio – transferring many channels of audio over a network connection, typically a gigabit connection. Audio recording, especially at high quality, creates relatively large files – around 100mb per-minute per-channel. Audio is often “multi tracked” to record several sources at once. Stereo is 2 channels multi tracked together, but an Orchestral recording or broadcast transmission can use 64 or 128 channels simultaneously, sometimes more. This is where network audio comes into its own, with upto 2000 channels of audio able to be transmitted simultaneously over a gigabit network.
(image – Yamaha pro audio diagram of a network audio system)
Another thing which sets audio transmission apart from many other computing tasks is the need for real-time transmission. Even these huge amounts of data must be transmitted with low latency (delay between input and output), low enough to provide lip-sync capability between audio and video, or for broadcast to appear live on screen. Its generally agreed that latency over 80ms is unusable for sync with audio, and latency of less than 5ms is common. In fact, the latest network audio devices can be clocked at pico-second latency.
Lets have a look at some of the companies who are creating networked audio devices :
Audinate. Creators of the Dante protocol, which is rapidly becoming the most common network audio system used by manufacturers. Audinate supply Dante expansion cards to many leading recording and broadcast equipment manufacturers, including Yamaha, Soundcraft and Focusrite. They also created a rather cool “virtual soundcard” which is capable of transmitting dozens of audio channels between systems with no hardware necessary.
Dante Via creates a flexible software audio bridge for your computer to connect with local USB, FireWire, Thunderbolt and analog audio interfaces, transforming them into networked devices. When Via is launched we are expecting some very forward thinking tech which has implications for device discovery and file transfer / streaming.
Focusrite. Focusrite have long been a name associated with audio-excellence. They have created a Dante-specific sound card range called Rednet. Rednet has been adopted by a diverse range of clients such as Fifa world cup 2014 and rock band band The Killers. These lightweight, dual redundant systems are starting to look more akin to server technology than music industry. Except they look cooler!