Funktion-One founders Tony Andrews and John Newsham have been experimenting with surround sound and immersive audio since the 1970s. In that time, they’ve made important discoveries when deploying these types of setups, particularly in large-scale environments. Here, in their own words, they share some of their experiences with us…
Although common wisdom holds sight as our primary sense, vision only perceives information from less than half a sphere through a relatively narrow electromagnetic window. It can also be ‘switched off’ by closing or covering the eyes, or by darkness. Sound, however, is sensed full sphere and with a much greater number of octaves.
Perceived light has one octave, whereas sound is perceived across 10 octaves. Hearing can never be completely avoided, even the strongest earplugs only muffle, plus sound is also received by the body, not just the ears. As such, it is the major contributor to our sense of where we are in space and the nature of that space, in a much more complete and accurate manner than light.
Sound conveys a constant stream of information about events located anywhere in our audio sphere to which the brain devotes considerable resources to achieve sound localisation. The brain would not use this much energy unless sound localisation was of major importance and is therefore intrinsic to human survival.
We propose that at an early stage in human evolution, a moving event would be prioritised because of its indication of immediate danger, such as the possibility
of being eaten, or alternatively the presence of possible lunch, and is consequently more stimulating to the emotions than stationary sound.
Stimulation of the emotions implies that moving sound has high entertainment value, which was recognised as far back as the late 1930s and implemented by Walt Disney and Leopold Stokowski in their collaboration to produce a surround sound system for Fantasia.
During the late-1960s, it was realised that there was commercial value in bringing surround sound to the living room. No doubt, the expansive mood of the times was part of the driving force. Stereo was being used
to its maximum effect on albums such as The Beatles’ Sgt. Pepper and Jimi Hendrix’s Electric Ladyland. The stereo panning on Electric Ladyland’s ‘Voodoo Chile’ is astounding and made a permanent impression on our young minds.
By 1971, a number of systems were on the market, the most prominent being CBS’s SQ, Sansui’s QS and JVC’s CD4. Most of them were matrixed quad with left and right rear channels derived from extra information within the standard stereo grooves of vinyl records. JVC’s CD4 system was different in the fact that it employed a high frequency component around 35kHz to carry the rear channel information, requiring a special stylus
as well as a decoder. Opinion has it that the Sansui
QS system with its Vario Matrix decoder gave the best results at the time.
All of these systems were still restricted to four loudspeaker positions with a front and rear orientation and derived right and left rear channels, although they did allow play on standard stereo setups. The ambisonic system – developed by Michael Gerzon, John Hayes, John Wright, David Brown and Professor Felgett of Reading University – was, to our knowledge, the first serious development towards a system capable of regenerating a full audio sphere with a multiplicity of powerful loudspeakers.
Our first foray into surround sound was with Jonny Rotten’s Public Image back in the 1970s, employing the eminently simple approach of using the difference between the two normal stereo channels to derive
a third rear channel as per Haffler. It was reasonably effective and drew the apparent sound away from the stage and into the audience. Signals panned hard to either side would be more present in the rear channel. We also built an eight-channel Haffler system for Steve Hillage’s Rainbow Dome Musick at the first Festival of Mind, Body and Spirit in 1979.
After discovering ambisonics in the late 1970s, we acquired a Minim decoder with which we attempted enhanced stereo. We knew that as ambisonics relied
on subtle phase information, it was important that the loudspeakers were as accurate as possible in terms of time coherence and fidelity. To avoid upsetting the bands with delayed percussive information we avoided aiming any of the loudspeakers directly at the stage. In the early ‘80s, we used the Minim decoder to carry out our first ever large-scale experiments with the enhanced stereo at the Glastonbury Pyramid Stage after we finished building it in 1981.
The first Experimental Soundfield was at Glastonbury in 1992 and used Pink Floyd’s quad mixer for panning between four stacks. The four-point setup featured a central FOH position in the middle, which doubled-up as a performance space. By putting the control position in the middle of the sound field, we not only halved the distance to the loudspeaker positions, it also put those involved – musicians (Underworld), DJs (Danny G and
Darren Emerson) and engineers (us) – in the same sound field as the audience and in close proximity to each other, removing the separation of the stage with its dedicated monitoring system.
Live input could be monitored on headphones to overcome the delay issues. It was an unforgettable experience for all concerned. The Experimental Soundfield attracted a lot of attention and opened the door to more adventures in sound. Soon afterwards, we sketched some ideas for building an automated panner using voltage controlled amplifiers (VCA), which led to the original analogue Azi Controller being developed by Chris Blythe and Bill Croston of Audio Dimensions, with input from Ben Duncan particularly on the VCAs.
In 1993, we used the newly built auto-panner with Underworld, Danny G and Roy Roach for an ambisonics stage at the Anti-Racism Alliance’s ARAfest in London. Despite some hardware and software teething problems, the audience reaction was astounding, and we were encouraged to continue the evolution. From this event we realised that we should leave the centre of the dancefloor for the people and move the FOH position to a halfway point between the perimeter and the centre. We also learned that four stacks of loudspeakers
placed equally on a 30m diameter circle was too large a spacing. Despite using ambisonics, a person located close to an individual loudspeaker would get the impression that the loudspeaker was turning on and off when the entire programme material was rotated. We decided that in future we should use six positions and, anyway, it was more cosmic that way.
We took our six-point ambisonics to Brixton Fridge (now Electric Brixton) on a couple of occasions for Escape From Samsara and then back to Glastonbury for the festival’s first ever dance tent in 1997. We deployed a 12-point ambisonic system at the Millennium Dome in 1999.
We went on to meet Dr Peter Lennox and Bruce Wiggins of Derby University in the mid-2000s through a shared interest in ambisonics. We’ll always remember Peter telling us that in audio you can trade level for precedence. Anyhow, with the advent of decent affordable DSP and Peter and Bruce’s help, a digital system for ambisonic panning was developed using a laptop computer, a multichannel soundcard and the VST ambisonic plugins designed by Wiggins. This was used from 2006 onwards, firstly with Sancho Panza and then the Origin Stage at the Glade Festival in Aldermaston before we returned to Glastonbury with the Experimental Soundfield in 2009 onwards.
In 2016 we merged with The Glade and deployed our latest loudspeaker innovations in a six-point system. It comprised main left and right stacks of Evo 7s and Evo 7T-215, four surround positions of Evo 6SHs and BR221s to create a 38m diameter hexagonal configuration, and our asymmetric bass setup featuring F221s and F132s.
By then we were fully digital – with Dante and amps with crossovers built in – meaning we could run a network cable and control everything on a laptop. It worked very well. In fact, by 2017 we felt we’d got it
as technically good as we could possibly make it, but any large-scale surround sound or immersive audio environment is challenging – mainly because the speed of sound (343m/s) is so slow compared to the speed of our perception.
Once a surround sound environment is larger than 10m in diameter, it presents challenges in terms of delays and arrival times. It’s possible to manage those challenges up to around 30m but beyond that the whole thing begins to unravel.
To avoid the annoyance of a sound at the perimeter that is doubling, we aim the speakers so that the focus of their directivity is around two thirds of the distance
– point to point – of the hexagon. Effects and non- percussive elements are less reliant on timing, so it’s possible to introduce movement, even in an environment that isn’t coherent and immersive in the purest sense. That said, it is the dimension of the enveloping sound that we’re most interested in.
Once you get beyond the workable size of surround sound setups, we’re finding that the loudspeaker technologies we’ve developed in recent years are capable of producing a truly immersive stereo image from just two speaker positions, without the constraints that we’ve encountered with surround sound during this rather long journey. This is achieved through design, engineering and geometry rather than digital manipulation. Large-scale shows we’ve done with Vero have been particularly rewarding in this respect.
Our experiments and adventures in surround sound and immersive audio are set to continue. This work is such an exciting and stimulating experience when it is done well.