• British engineer and inventor Alan Dower Blumlein pioneered the idea of stereo recording and showed in 1930 that accurate and controlled spatial effects could be obtained by relying on only one of the two locating clues - intensity or time of arrival. He invented the binaural recording (now known as stereo sound) in 1931.
  • The concept of stereo (from the Greek 'stereos', meaning solid) came from stereoscopic photography, which can produce an incredible 3-D illusion. Many have experienced this with the Viewmaster stereo slide system, popular in the 1950s and 60s. The earliest use of stereo sound was through experimentation with binaural sound in the late 19th century. Binaural audio used two microphones to capture sound and headphones for listening. This method remains in use to this day with, for example, the Bruel and Kjaer Type 4101 binaural microphone, which is worn by a person in much the same way as portable headphones. An early two-channel playback system, developed and sold in the early 20th century, used a two-channel phonograph cylinder and two mechanical pickups and horns. Our modern concept of stereo sound comes from two sources: Alan Blumlein of EMI in England and Arthur Keller of Bell Laboratories in the US. These two groups followed quite different paths. Bell Labs in the US began experimenting with stereo recordings in the 1920s. Their earliest recordings were made on double-groove records that required a special playback system equipped with two tonearms and cartridges. Shortly thereafter, they produced single-groove recordings in which the groove walls were cut at 45 degree angles. This method was adopted in the 1950s when the stereo LP was introduced. These early experiments led to the development of a three channel (left, right, and centre) recording system by the early 1930s. This was a compromise, as the Bell engineers believed the ideal system consisted of an infinite number of microphones reproduced through an infinite number of speakers. This approach puts the emphasis on the position of the instruments. It was believed that with a hundred channels, one could perfectly position the instruments in a large orchestra comprised of a hundred musicians. These could be placed in a two-channel field during reproduction or, better, a three-channel field. This was impractical because a record only had two groove walls and could only reproduce two-channel audio. Their solution was to add centre channel audio to both the left and right channels at a reduced volume. In the proper position, the listener could hear a phantom centre channel. Alan Blumlein, with EMI in England, followed a different approach to stereo, what he purposely called binaural audio. He filed a patent in 1931 for: “Improvements in and relating to sound transmission, sound recording and sound reproducing systems.” Blumlein, who died in an air crash in 1942, is also credited with helping to develop television and radar. His concept was to record sound using two microphones, placed at 90 degrees to one another and 45 degrees from the source. These were not placed close to the source, but at the listener’s position. Although most people think of microphones as either omnidirectional (i.e., equally sensitive to sound from all directions) or cardioid (i.e., most sensitive to sounds in front, becoming less sensitive as the source moves towards the rear of the microphone; the pickup pattern is somewhat heart-shaped), the best microphones at the time were bidirectional (i.e., equally sensitive at the front and the back, with little sensitivity to the side). When two bidirectional microphones are oriented 90 degrees from one another, four areas of maximum sensitivity are produced. This arrangement is known as a coincident pair or a Blumlein pair. The coincident pair is not perfect, however, because there is a physical space between the microphones. There have been many variations of this pattern developed over the ensuing years: some using an angle of separation greater or lesser than 90 degrees; two side-by-side omnidirectional microphones aimed at the source and separated by a fixed distance, sometimes with audio baffles between the pair; the mid-side or MS system, which uses one microphone aimed at the source and a bidirectional microphone aligned at 90 degrees; and the Soundfield microphone, which uses a four-microphone capsule to create a virtual coincident microphone that can recreate a sense of space in anywhere from two to six channels. The coincident pair preserves three audio cues that we use to locate sounds: - Precedence: if a sound reaches your left ear before the right, it will appear as if it is originating from the left and located fairly closely to the listener. - Intensity: if a sound is louder to your left ear than to the right, it will appear as if it is originating from the left. This sensation varies with the signal frequency, increasing as the frequency increases and becoming less at low frequencies. - Phase: sounds can arrive at your ears in phase or out of phase. This also helps the listener to locate the source of the sound, most noticeably in the mid frequencies. The Blumlein method works best when reproduced through headphones. It can be quite effectively reproduced with two speakers, depending on their separation and distance from the listener, as well as the way in which the performance was recorded and mixed. The Blumlein recording method attempts to recreate the performance environment, rather than creating a new one. The Bell Labs approach retains only one audio cue: intensity. One microphone is used to record each source, producing monophonic multi-channel recordings. Each signal is then placed in a specific location during mixing by adjusting the signal intensity (e.g., volume) on each playback channel. For example, with two payback channels the signal could be placed hard left, hard right, or panned somewhere in between. The editor does this to create a musical performance or simulate the position of the instruments on a stage in from of the listener, rather than recreating the actual performance. These two methods create recordings with a very different sense of space. We are aware of the location of each instrument in multiple-monophonic recordings, but we have little sense of spaciousness, except that which is provided during the mixing process (e.g., phase reversal, delays, reverberation). With the Blumlein method, cues are retained that provide a very different sense of space, particularly when heard through microphones and in the nearfield of the two speakers. It should be obvious which method is used most frequently for modern recordings: the multi- monophonic method developed at Bell Labs. I must admit, however, that a well-executed two-channel recording, made using the methods pioneered by Blumlein decades ago, sounds far better than a multi-monophonic recording, even those that are extremely well engineered. Great sound can be accomplished with two speakers; five- or six-channel systems are not automatically better. [With historical material from UHF Magazine, copyright 2004 by Broadcast Canada]

Copyright 2020, Wired Ivy, LLC

Answerbag | Terms of Service | Privacy Policy