We spend a lot of time talking about 3D sound for cinema, but when it comes to creating effective soundscapes for 360 degree video and virtual reality environments, many of the principles at work are slightly different, and you’re going to need correspondingly different kit. We asked our team to weigh in on the best mic options, how to handle different speaker arrays, and their top choice for headphone monitoring. Oh, and their new favourite app…
Sennheiser’s AMBEO VR mic is widely regarded as one of the best available for 3D capture. A first-order ambisonic microphone with four matched cardioid KE 14 capsules in a tetrahedral arrangement, the AMBERO VR captures four channels in “A-format” which can comprise a single, spherical soundscape. This can then be processed to “B-format” as required in post-production with the free Sennheiser AMBEO A-B converter plug-in, to adjust for the “virtual” microphone position and direction. The Ambisonics standard has been widely adopted across production, post and reproduction, so you have plenty of choice when it comes to compatible studio hardware and DAWs, and the spherical nature of the soundscape makes it ideal for VR (more on that later). AMBEO mics are also fairly easy to work with, as you don’t need a hardware encoder – just an audio interface with four or more mic inputs and the free software encoder.
Another excellent option is the Røde NT-SF1, which pairs with the SoundField by Røde app to offer an alternative approach to 3D capture. The mic records a full spherical sound field, which means that in post you can alter the mic directivity, position and rotation to suit your mix. Using the SoundField by Røde app, you can harness the latest beamforming technology to create cardioid, figure 8 and even shotgun-type patterns in post.
Both the above technologies are primarily interested in capturing audio for ambisonic manipulation and replay. However, some people can find that coincident mic arrays, whilst providing a rich sense of 360 degree localisation, can lack a sense of depth relative to other techniques.
Equal Segment Microphone Arrays (ESMA) have been recommended as a good solution for this problem, with significant academic work having been done on this subject by Dr. Hyunkook Lee. ESMAs rely on the trade-off between time and level differences and, when appropriately designed, give the best of both spaced and coincident techniques in terms of ability to localise and sense of acoustic space, similarly to the performance of Decca Trees relative to spaced or coincident pairs in the stereo domain.
There are no off the shelf options for ESMA solutions; it’s a case of carefully designing the system you require. If you’re thinking about putting together an immersive ESMA rig, please don’t hesitate contact our audio team for microphone and rigging recommendations!
At the other end of the market, there are binaural capture technologies like the Sennheiser AMBEO Smart Headset. Traditionally, capturing audio binaurally required a Neumann KU-100 dummy head or similar, which represented a significant investment. However, recent developments mean that there are now microphone pairs which use the recorder’s own head as a binaural dummy. This is great for the YouTube videographer wanting to capture immersive audio as simply as possible, although because of the nature of binaural audio it doesn’t always hold up well to much mangling in post.
The Applied Psychoacoustics Lab at the University of Huddersfield have made some fantastic resources for anyone working in 3D sound– from room impulse response libraries to research tools to renderers, but we’re big fans of their MARRS app. Available for free on iOS, Android and the web, MARRS provides “an interactive, object-based workflow and graphical user interface for localisation prediction and microphone array configuration,” according to APL. It allows you to predict the perceived positions of multiple sound sources for a given microphone configuration, and configure suitable microphone arrays for your desired spatial scene in reproduction – it’ll even take into account microphone height and vertical orientation, as well as the target loudspeaker base angle.
Dolby, long-term dons of cinema sound, have cornered the theatrical and traditional home entertainment markets for 3D sound with their Atmos system. We can help you spec up, install and configure an Atmos-compatible room (details here), but there are other options.
Ambisonics has emerged as the preferred 3D sound format for VR gaming and 360 degree video, and we can help you develop solutions for capturing, mixing and playing audio using it. There are several reasons why it has seen such widespread adoption in gaming. Firstly, Ambisonics renders sound as a speherical soundfield relative to the mic (when recording) or the listener (during playback), leading to very smooth ‘movement’ of the sound as its position changes.
Secondly, the mathematics required to convert to binaural for headphone listening are well established and widely available (although we love the Audeze Mobius virtual HRTFs; see later) and more likely to be supported on any particular platform (particularly mobile) than any of the major proprietary 3D formats.
Thirdly, an Ambisonics mix can be decoded to any speaker array. Unlike Dolby Atmos, which requires a precisely calibrated 7.1.4 room, an Ambisonics mix can be fed to any pre-existing speaker array (although if you were to play it back on, say, a standard stereo array, the sound would be folded down).
Furthermore, a first order B-format ambisonics mix requires only four mono audio channels to deliver, reducing bandwidth requirements compared to Auro, DTS or Atmos.
However, when it comes to gaming, most users are going to be listening through headphones. Which means if you want to QC an Ambisonics mix the way your users will hear it, you need some fairly advanced headphones. Luckily, we’ve found them.
These headphones have been getting rave reviews for their clarity, accuracy and overall sound quality. Using a combination of Audeze’s own planar magnetic technology, Waves Nx head tracking (your position is checked roughly 1000 times a second), sound isolation and realistic room emulation, Mobius headphones are designed to make you feel like you’re right at the centre of what you’re listening to, whether that’s an album, a film, a traditional game or an immersive VR experience.
This is great for end users who want high-quality playback, but it also means they can be used to test and QC a range of 3D sound projects and formats. For example if you’re building an Ambisonics, 7.1 or 5.1 mix and want to QC on headphones, this is the tool for you!
Most importantly, the binaural encoding is done in the headphones, freeing up your GPU and CPU for more plugins, virtual instruments and video playback.