Â鶹ԼÅÄ

What is Next Gen Audio and the EAR?

The response to the first version of the EPS was really positive, but there’s always room for improvement, especially given the myriad possibilities that NGA allows. NGA builds on the idea of object-based rather than channel-based production, meaning the listener will get the best possible surround sound experience, regardless of how many speakers they have. Also, we have collaborated with other industry partners to create the Audio Definition Model as we believe that open standards like ADM are crucial to the future success of NGA. With its improved features and bug fixes, we hope the new EPS release will motivate DAWs to support ADM natively. In short, we want more audiences to have access to immersive and personalisable content, and free, open-source software is the way to encourage creatives to produce it.

Higher-order ambisonics

The 2020 release of the EPS (v0.6.0) included support for object-based and channel-based assets. We’ve extended this with support for Higher-Order Ambisonics (HOA) from v0.7.0 onwards. Unlike channel-based assets, HOA (sometimes referred to as scene-based audio) can provide a 3D representation of a soundfield without using discrete virtual sources. This is a significant advantage over channel-based assets for some use cases since HOA can deliver a perceptually continuous sound field rather than the perception of being comprised of multiple distinct point-sources. This difference can become particularly apparent when head-tracking is used. HOA can also be flexible with channel count depending upon the degree of spatial resolution required (which affects the localisability of sound sources within the sound field). The HOA order determines the spatial resolution. A full 3D representation of a soundfield can be contained within as little as four audio channels, referred to as first order. The EAR Production Suite supports first order (4 channels) through to sixth order (49 channels).

A typical usage scenario for HOA would be to provide background ambience for a scene, such as the singing of birds and the rustle of trees in a forest environment. In this case, a low order is appropriate since spatial resolution is not particularly important. This HOA asset could be created by either:

• Capturing real-world environments using an ambisonics microphone, such as the Sennheiser AMBEO,
• Producing HOA from individual pre-recorded assets and placing them in 3D space. There are now many spatial audio plug-ins available that can do this, such as the SPARTA plug-ins, for example.

In cases where we wish to build up our soundfield from many individual point-sources and maintain precision, a higher order would be more appropriate. We would typically use audio objects in this case, but HOA can offer advantages in some scenarios. HOA allows you to encode as many point-sources as you would like into the soundfield without requiring additional channels. Therefore, there will be a break-even point where HOA begins to offer channel-saving benefits. For a small number of sources, it is usually best to use audio objects. When the number of sources becomes greater than the number of channels required for an HOA asset of sufficient order, then it becomes beneficial to use HOA instead. However, it should be considered that this would not be suitable where interaction with individual sources is required since these sources become mixed and ‘baked-in’ to the same HOA asset.

The second major feature added to the EPS since v0.6.0 is a binaural monitoring plug-in. This allows users without access to a surround sound speaker system to experience the immersive, enveloping capability of NGA content with 3D sound on normal headphones. The plug-in is based around our BEAR library (Binaural EBU ADM Renderer), an innovative binaural rendering solution for ADM media which we recently released open-source on GitHub.

Binaural monitoring

The new binaural monitoring plug-in also supports a range of head-trackers. Head-tracking allows producers to explore their mix virtually in a novel way and has been shown to increase the realism and immersion that listeners experience as self-motion cues are consistent with binaural information. The soundscape will rotate according to head-orientation, offering three degrees-of-freedom. The plug-in responds to Open Sound Control (OSC) messages understood by many of the most popular spatial audio plug-in suites, making it compatible with a wide range of existing head-trackers, including SPARTA, IEM and Ambix. Any head-tracker compatible with those suites should also function with the EPS.

Finally, many REAPER digtial audio workstation (DAW) templates are now available as part of the standard installation. These range from simple templates for learning how to use EPS features, up to fully featured multi-channel and object production-ready examples. We hope these will help new users to get started more easily and mean experienced users spend less time on set-up and more time on creativity.

A full list of bug fixes and improvements to the EPS - several of which have been suggested in our MakerBox community - is available in .

We’re delighted with version 1.0 of the EPS. We hope our new features give current users even more ways to realise their artistic potential using NGA and bring in new users to the MakerBox community, making a stronger case for adopting the ADM standard in all DAWs. In particular, binaural monitoring opens up NGA production to producers without immediate access to a full surround sound mixing studio.