Australian sound engineer and inventor Joe Hayes has been at the forefront of cutting-edge sonic technology for more than 30 years. He continues to impress engineers and musicians around the world with his A3D acoustic technology and its application in the low-cost, high quality Emergence AS8 speaker system. It is a system which advances the diﬀusion technology used in many concert auditoriums around the world by producing that diﬀusion at source in the speakers, rather than through expensive panels and room treatments. Here, Hayes explains the development of that system and recording engineers and experts Dr Toby Gifford, Ted Orr and Shane Fahey reveal how eﬀective it is in the studio and in the ﬁeld.
These experts’ views inform from the perspectives of professionals at the production stages through insights based on their adoption of the AD3 system into the workflow. Supporting their findings, SoundStage! Australia’s Editor-in-Chief Edgar Kramer provides further observations with a subjective listening appraisal of this unique technology.
Joe Hayes, Inventor
About two years ago I was asked for a clear technical explanation of what A3D technology offered that was unique. To that point we knew how to ﬁne tune a Quadratic Residue Diffusor to a spherical wavefront and reach a point of audio nirvana where suddenly the audio cues of an acoustic space jumped out of the image. But that in itself did not explain why this happened.
After considering just about everything available in the academic literature covering wavelets, seismic science DSP, diffusion, contemporary cochlea psycho-acoustics and neural biology, it was a Melbourne based academic Dr Toby Gifford who facilitated a breakthrough for us. We hired Toby to do a simple Matlab model of A3D and of normal piston radiators. The result was published at the AES Conference of Automotive Audio in 2017.
The model showed A3D emulates a point source: the sound-ﬁeld radiates power uniformly in all directions. For anyone who understands the art they will know this is a good thing. But not necessarily unique to A3D. Then we noticed that an unusually ‘shimmering’ quality to the polar sound ﬁelds that act at a higher frequency multiple. The wild thing about all of this is with the A3D process you hear the room cues in the recording as though they are those of the listening space. It is like they are real acoustics.
Purple = 15,000Hz, Yellow = 9,000Hz, Red = 5,000Hz, Blue = 2,000Hz. Across the bottom is 8 cycles of 2,000Hz. Animation shows sound intensity patterns as the 2,000Hz sine wave moves through its cycles. It’s these patterns that change (at a speed of 4 x the frequency of the signal) that completely diffuse the listening pace eliminating specular reflections.
Then recently Toby broke through, on this shimmering quality. Using his considerable math skills and understanding of time and space manifolds he realised the shimmering in the A3D wavefront is an Eikonal of the QRD itself!
But before Dr Gifford explains this in his own words, it’s important to mention one of the great contributors to A3D, loudspeaker engineer Brad Serhan who provided guidance in terms of the fine-tuning/voicing. I knew we needed someone who was prepared to take on technical challenges. In the movie, Apollo 13 Jim Lovell's mum is quoted as saying “... if they could make a washing machine fly, Jim could fly it!” Similarly, I say “… if you can make it vibrate, Brad Serhan could tune it to its maximum musicality!”
Brad was fearless in tuning the first A3D systems and, through that, we first noticed “another presence” in the sound field. Through his dedication to his craft, we were able to determine the tolerance we could work with. As a direct result, we now have the maturity of this new technology we need for licensing it.
Dr Toby Gifford, Research Fellow, SensiLab Monash University
A3D technology emulates a point source: the sound-ﬁeld radiates power uniformly in all directions. This minimises sound-ﬁeld distortion from self-interference that is typical of traditional membrane drivers. In particular it streamlines the air particle velocities along the direction of radiation, minimising off-axis (tangential) energy that is associated with non-uniform sound radiation. As a consequence spatial information encoded in stereo recording is maintained in pristine form by A3D, whereas traditional loudspeakers distort this information by confounding it with directionally confusing ‘eddies’ formed by self-interference. Thus, A3D enhances signal to noise ratio by allowing for physically larger drivers (with greater capacity to excite a sound-ﬁeld) to behave like small point sources.