Abbey Road Head of Audio Products, Mirek Stiles, writes about what spatial audio means for actual recording.

4th December 2018
Pushing aside all the exotic post-production manipulation software, Binaural rendering options and the awesome VR/AR headsets, I thought it might be good to focus on what Spatial Audio means for actual recording.

The good news is it can mean that not much changes in terms of your usual workflow. You will still need great spot and room microphone techniques, the ones that you have been developing to marvellous effect. In fact, you could just record everything as normal and use various post-production mixing software to create your 3D world. Having said that, there is room to expand the recording process – and come on, who doesn’t just love the idea of further complicating the recording process.

There are two areas for consideration when recording for SA. The first of which is ambisonic microphones, which have been around since the mid-‘70s. Abbey Road still has some early examples of the Calrec Soundfield microphone that former Abbey Road engineerJohn Kurlander requested and told me he loved using. Unfortunately, today these Carlecs are in bad shape and I wouldn’t trust using them as a talkback microphone, let alone trying to capture some crazy spherical ambience to impress my colleagues with. Fortunately, thanks to the growing SA trend, a new generation of fancy microphones are emerging onto the market. Ambisonic microphones come in different flavours (you didn’t think this was going to be simple did you?), the most basic are FOA (First Order Ambisonic) microphones. These consist of four capsules in a tetrahedral arrangement that encodes into four channels = omni, left and right, front and back, up and down.

There are HOA microphones (Higher Order Ambisonics), which in a nutshell mean more channels; more channels means more spatial resolution and more resolution probably means better localisation/recordings (in theory – it actually opens up a whole host of other challenges and question marks, but again, for another article). It is worth noting that Ambisonics have two different channel order standards, Furse-Malham (FuMa) and Ambix. Again, not going into detail here, but my advice is to stick to Ambix whenever you can, and always keep track of conversion workflow – you don’t want to get your FuMa and Ambix mixed up. Ambisonic microphones can be an awesome tool in your microphone arsenal. I tend to find when placing them too far or high in a room, they don’t really work to great effect, but when in the mid and near field, once decoded, can sound stunning. For example, I recorded a solo acoustic guitar/vocal performance with the usual close mics and a nice stereo pair to pick up some room. In between the close and distant mics, I placed an FOA microphone. Once decoded and blended into the final mix it created a beautiful sense of depth and space. This would have been very challenging (if not impossible) to create using standard mics and processing. The same can be said for close micing. For example, an FOA microphone on a harp or piano sounds stunning – the ambisonic microphone has the ability to pick a large spread of detail from all around in every direction. A midfield ambisonic microphone can also act as a sort of directional glue between close mics and ambient microphones.

The other side of the coin includes spatial multi-channel microphone arrays (MMAs), which I believe pick up much more useful information for reflections and room than ambisonic microphones do. Most spatial arrays, beyond the Decca Tree, were alien to me before taking part in a joint experimental recording session with Gavin Kearney, from the University of York and Hyunkook Lee, from the University of Huddersfield a couple of years back. The session in question was conducted in Studio Three as part of then student, now Abbey Road Red employee, Hashim Riaz’s final Masters’ degree. The concept was capturing a live band for an immersive virtual reality experience. On this session, we had all sorts of spatial (MMAs) arrays out on the floor (around 80 channels worth to be precise). As explained, ambisonic microphones can provide spatial glue. Obviously close micing gives you definition but what do you need to create a sense of magical externalisation? As far as I am concerned it’s a Hamasaki Cube or an Equal Segment Microphone Array.
 

Hamasaki Cube

The Hamasaki Square was designed to capture the ambient and diffused sounds of a concert hall with surround sound in mind. As we are now in the world of 3D sound, we add another layer to capture some height – hence the Hamasaki Cube. The cube consists of eight fig.8 microphones placed 1m apart from each other. For the bottom layer, the null of the fig.8 pattern is pointing towards the sound source, so we can reduce the direct sound and increase the diffuse sound captured by the array. The top layer has the null pointing towards the sides, so we capture the diffused sound from above and below. The overall array should be placed as far and high from sound source as to capture some exciting room reflections.
 

Equal Segment Microphone Array:

The Equal Segment Microphone Array (or ESMA as I like to call it) is another 8-channel configuration, this time consisting of the humble cardioid pattern. The idea is to capture sound in a 360-degree azimuth plane positioned in a square arrangement (angle of 90 degrees between each mic) with 50cm between each capsule to minimise inter-channel crosstalk. This is then augmented by another four upward facing cardioids to capture, some height.
Once the eight channels of each array are processed via an object or ambisonic panner, the result can be extremely lush Once blended in with close and ambisonic microphones they really help create an overall immersive audio scene.
This is just an intro to the sort of recording workflow that’s clicked for me so far. It will most likely change, but it’s a good start. I genuinely think that fully immersive audio is key and crucial to playing a significant role in the future of entertainment, not only for film and VR but music in general. It vastly expands the sonic landscape for artists to present their work. It’s in the best interest of Abbey Road Studios and all audio professionals to explore this technology and create the best sounding content we can.
I hope this inspires you to explore – have fun.

This is an edited version of an article that originally feature in Resolution Magazine V17.2 | March 2018