Dante networked audio

The way in which we interconnect audio equipment has hardly changed since the early days of the technology. Each audio path is represented by a single cable that runs from one piece of equipment to the next. In modern telecommunications parlance, the logical and physical paths are the same: one cable equals one channel. Signals are routed via individual copper cables, and implementing unplanned – or even planned – changes in a system, such as moving or adding channel capability, can result in extreme expense, time and effort.

The advent of digital audio networking, also called "audio-over-IP" means the removal of much of that time, effort and expense, by separating a system’s physical and logical connections. A single network cable or fibre-optic path carries multiple channels, thus making it possible to make infrastructure changes relatively cheaply, quickly and inexpensively, and without the requirement for major building changes and works. At one end of the network, a single connection is made to the music workstation(s) that will be used to work with the audio. At other points on the network, a single interconnect links each interface, which can provide multiple analogue and/or digital I/O wherever they are required. Multiple analogue cable runs and expensive, heavy multiway cables are things of the past as multiple channels are carried digitally between locations.

Dante, developed by Australian company Audinate, is a tried and tested ultra-low-latency digital audio networking system based on Ethernet and IP technology – implementations of internationally-agreed standards. Dante networks have been used in the live sound industry for years. Dante is what's known as a Layer 3 protocol - it uses standard networking hardware and infrastructure rather than requiring special switches and other components. Dante is also compatible with the AES 67 audio networking standard. Suddenly, audio can be made available anywhere it is required, capturing or relaying a live performance or distributing audio to previously hard-to-reach or prohibitively expensive locations.

Dante is the audio networking system of choice for Focusrite products including the RedNet range of modules and the Red range of multi-standard interfaces.


Breathe the Air

The Red Evolution mic pres in the Red Range include the ‘Air’ function. ‘Air’ is a purely analogue effect that emulates the sound of the transformer-based mic preamps of the ISA range, the foundations of Focusrite’s “Heritage Sound”.

'Air' recreates the sound-shaping effect of the classic ISA transformer, by using an authentic analogue model of the original transformer-based preamp design. Engaging 'Air' reconfigures the three-stage preamp signal chain to switch in the correct transformer-style input impedance, followed by a linear gain stage, and finally inserting a passive frequency-shaping stage to recreate the tight low end and elevated high-end response – thus creating the effect described by our customers as the classic 'Air' sound of the heritage Focusrite mic pre.

Parallel Path Conversion

At Focusrite, we’ve been building mastering-quality digital audio converters for decades, and the converters you’ll encounter in the Red range are the best we’ve ever made. Needless to say, we’ve selected high-performance, precision conversion chips for the Red range. In addition, the Red range features what we call 'parallel path summing' conversion. We use two conversion chips in parallel to significantly enhance the signal-to-noise ratio. Using two converters in parallel and summing the result gives an increase in signal level of 6dB (as the two signals are in phase) while random noise is only increased by 3dB (as phase cancellation happens on average 50% of the time). The result: a net improvement of signal/noise ratio of 3dB.

We also improve performance through premium component selection, design and board layout. We pay as much attention to the analogue side of the equation as to the digital, optimising noise levels by keeping digital and analogue parts of the circuitry well apart and using multi-layer boards with internal ground-planes.

The Red range's design strikes a precise balance between multiple factors – including dynamic range, distortion, noise floor and conversion latency – to achieve the optimal overall sound, with main inputs and outputs delivering ≤0.0009% THD+Noise and frequency response flat from 20Hz to 35kHz ±0.25 dB or better, with 118dB (A-D) and 121dB (D-A) dynamic range. The Red range operates at standard sample rates up to 192kHz, 24-bit – ideal for high-resolution audio. And these are real-world figures, not something lifted from a chip's data sheet. Our goal is to give you the best sound quality. We won’t compromise sound quality to get a more impressive number on the noise floor that you’ll never hear, for example - once you reach a certain level, THD is much more important.

High Resolution Audio (HRA)

By "High Resolution Audio" (HRA) we generally mean recordings made to 24-bit, 96kHz (or 88.2kHz) specifications or better. All Focusrite interfaces can operate at these specifications. HRA is becoming increasingly popular, with music streaming and download sites offering high-resolution content. In addition, high sample rates offer the best option for archiving. Future replay systems will no doubt offer better performance than they do today, and archiving music at the highest possible quality levels means that we will give future listeners the best chance for an even better musical experience.

For more on the use of higher sample rates, see the Focusrite Precision Conversion page.

Red 4Pre Latency Table

Round Trip Latency* (ms) tested at 96kHz on OS 10.11 and Mac Pro

Buffer Size Logic Pro X Cubase Pro 8 Pro Tools 12 HD Native Pro Tools 12 Pro Tools 11 Ableton Live 9 Reaper 5
32 1.67 1.67 n/a n/a n/a 2.09 1.68
64 2.34 2.34 1.88 3.03 3.00 2.42 2.34
128 3.67 3.68 3.21 5.03 5.00 3.75 3.68

Round Trip Latency (ms) tested at 48kHz on OS 10.11 and Mac Pro

Buffer Size Logic Pro X Cubase Pro 8 Pro Tools 12 HD Native Pro Tools 12 Pro Tools 11 Ableton Live 9 Reaper 5
32 2.58 2.59 2.38 3.25 3.27 3.42 2.61
64 3.91 3.93 3.71 5.26 5.26 4.08 3.94
128 6.58 6.60 6.38 7.92 7.93 6.75 6.60

*Logic Pro X @ 96kHz, 32-sample buffer