Surround Sound Recording Techniques
I have been interested in Surround Sound for a long time. While I was still an undergraduate I assisted Granville J. Cooper and and Sid O’Connell in experiments that were presented at the London Audio Fair in the Hotel Russell. The BBC were also at that time presenting to a very small section of the public late night/early morning experimental broadcasts in surround sound that utilized two stereo radio receivers tuned to differerent stations. These took a lot of setting up, but I recall hearing enough interesting and convincing effects on these broadcasts to stimulate further experimentation.
As a founder member of the English branch of the AES, I came into contact with Peter Felgett and Michael Gerzon, who were then developing Ambisonics and took part in several sessions evaluating recordings that they made to test the system — in all three planes.
As a recording engineer at Abbey Road Studios, I persuaded my bosses to let me do some experimental recordings in Surround Sound, using a four-channel Studer J-37 tape machine. The first of these was done at King’s College, Cambridge, where we were recording Christmas Carols. Using a figure of eight pair of KM56 microphones for the back channels, and the regular stereo pickup for the front channels, I played this back to the Director of the King’s College Chapel Choir, David Willcocks. Hearing a take immediately after conducting the choir in the chapel, he declared that it “felt like being there”. Of course, like many musicians who were among the first to hear the improvements that have been made in the craft of sound recording over the years, he was unknowingly exaggerating; but what he heard was definitely an improvement on stereo.
Two other recordings followed — one in Abbey Road’s No.1 Studio and the other at Guildford Cathedral — but they were not as successful. I concluded that the King’s recording had worked largely because of the large amount of reverberation in the building, which gave a good wash of surround sound, but little directional information, which caused me to place the rear-channel microphones relatively close to the singers. This meant that there was an acceptable delay between the front and rear channels, and the latter, being a Blumlein stereo pair, preserved a good image of the choir. The walls of Guildford Cathedral, on the other hand, had been treated by the original builders with a material that greatly reduced the reverberation. I was unable to monitor this recording in surround on the spot, and because of the larger size of the ensemble used a pair of U87 microphones in omnidirectional pattern for the rear channels. In order to get any effect of distance, I had to move the microphones so far back that the delay involved was too great. This was a useful lesson, but of course, I only discovered it back at the studios, when it was too late!
The recording done in No.1 Studio, Abbey Road was of a band of Gurkha soldiers playing bagpipes while marching back and forth. In order to protect the microphones and keep them out of the way of the heavy boots, I installed them in the catwalk above the studio. These microphones were four STC hypercardioids mounted on a wooden frame and as close to each other as I could get them in an attempt to simulate a Soundfield Microphone. Walk-about experiments indicated that, when played back over four loudspeakers, this array did indeed enable a listener to locate and follow the direction of a moving source of sound. The problem was that the bagpipes were so loud, and the sound source was so large because they were all playing the same tune. This caused the sound to bounce randomly between all the walls, with the result that it was impossible to tell where the players were, except in a general sense.
It was after this that EMI and other record companies launched Quadraphonic Sound, which utilized four loudspeakers, but which was doomed to failure by an over-optimistic system of matrixing four channel into the two channels of a vinyl disc. This was played back by a conventional stereo pickup and processed. In the best case, this gave the effect of sound at the rear, but there was little true directional information. The crosstalk of the system was just too great.
I did some recordings at the University of Surrey and Indiana University that employed the Calrec Soundfield Microphone. Early models of this microphone were difficult to use, as many examples needed careful adjustment of levels in order to avoid excessive noise or distortion. In those days, Nimbus Records invested heavily in ambisonic recording, but its commercial releases on vinyl were encoded in UHJ, a 2½ channel compromise, and did not issue their products on discrete-channel media until recently.
With the advent of digital video recordings and home theater systems, Surround Sound became a reality for the consumer, and the establishment of SACD and DVD-A allowed six discrete channels for audio-only recordings in the 5.1 format. Encouraged by the ability to cut my own DVD-A multichannel discs, and having heard some good commercial examples — notably from the German company, Tacet — I again embarked on my own experimental Surround recording. These experiments are detailed here.
advocating greater coverage of Surround Sound
Technical information about Ambisonic Recording may be found HERE.