How To Set Up A Spatial Audio Session? - Abbey Road's Mirek Stiles & Red’s Alumnus LifeScore

28th January 2021

In this, the second of a three part series, we give the floor to our Head of Audio Products, founder of the Abbey Road Spatial Audio Forum and board member at Abbey Road Red, Mirek Stiles, to give you a detailed overview of how he set up the microphones for the spatial audio aspects of the session.

I had the absolute pleasure of rigging up some spatial audio microphone experiments in Studio 3 for the brilliant composer Phillip Shepperd and his inspiring Red alumnus start-up Lifescore.

Philip had a group of first-rate strings players in the room to capture elements for his AI driven platform and very kindly gave me free run to set up some additional fruity microphone arrays in amongst the more traditional set-ups we see here at Abbey Road.

The room was laid out with 2 x 1st violins, 2 x 2nd violins, 2 violas, 2 cellos and 1 x double bass in the traditional left to right semi-circle, with the double bass in the middle at the back. Studio 3 has a lovely bright acoustic for smaller string ensembles and works beautify with this type of band.

The layout in Studio 3

In the middle of the strings I placed a First Order Ambisonic microphone and an Equal Segment Microphone array (ESMA for short). The First Order was the Soundfield, which is essentially a 4 capsule mic that picks up sound from Front to Back, Left to Right and Top and Bottom.

The 1st Order Ambisonic Sound Field microphone

The ESMA consists of 12 x cardiod pattern microphones spaced in a 50 x 50cm square and placed at head height. 4 x microphones pick up the TOP Layer (pointing up the ceiling) Left (L), Right (R), Left Surround (Ls) and Right Surround (Rs). The next layer of 4 microphones picks up the MID Layer (pointed at the musicians) L, R, Ls and Rs. Then the BOTTOM layer pointing at the floor, lower down to pick up the floor reflections, L, R, Ls and Rs. This cluster picks up a nice direct sound from the musicians but has a nice sense of space around it. This concept was originally designed by Dr. Hyunkook Lee form Huddersfield University.
Above the Conductor (or Phillip in this case) I placed a 2nd Order Ambience microphone, from Core Sound, at head hight to capture the conductor point of view. The 2nd order microphone has 8 x capsules instead of 4 and hence more information and spatial clarity is recorded. It’s a lovely sounding microphone.

The 2nd Order Ambisonic Core Sound microphone

And lastly, around the perimeter of the musicians I placed high up in the celling a Hamasaki Cube to capture the spatial elements of the satisfying Studio 3 room acoustics. The Hamasaki Cube consists of 4 x 2 layers. The first player points to the ceiling - Left, Right, Left Surround and Right Surround, these are set to a cardiod pattern. Then directly under the top layer is the bottom layer that consist of 4 x fig 8 pattern picking up the side reflections of the room L, R, Ls, Rs. The intention of the Hamasaki Cube is it picks up the indirect sound from the musicians bouncing off the walls and ceiling – it’s rather marvellous and a personal favourite of mine.
This set up is well placed to give the content creator enough elements to create realistic sound zones to import into a game engine and make a really nice VR experience, where the user can walk through the performance in real time. All the above microphone arrays also encode beautifully into Dolby Atmos and even 5.1, so there is plenty of flexibility on these types of set ups.
You can see some the exciting action from this session in this video, where me and my colleagues explain a little more about what’s going on behind the scenes.