‘Binaural’ simply means ‘relating to two ears’, and it’s only an accident of history that we don’t use the word instead of ‘stereo’ (which actually means ‘solid’).
As it is, binaural recordings are basically recordings made specifically for listening to on headphones rather than loudspeakers.
There’s one very important difference between headphones and speakers: the left earpiece of a pair of headphones is only audible to the left ear, whereas the left loudspeaker of a conventional stereo pair is audible to both ears, with a slight time delay and loudness difference between them. Good, convincing stereo sound is possible in both cases but for optimal results the recording must have been made, or mastered, specifically for one of the other.
The basic concept of binaural is simplicity itself: to capture the sounds that should be heard by the left ear, put a microphone where the left ear would be in the performing/recording space, then do the same for the right ear. With a few nips and tucks (baffles, or a dummy head, between the microphones), that’s how binaural recordings have always been made. It’s an obvious idea and it was first used as far back as 1881, just about the time when the concept of sound recording was dreamed up and only five years after the first electrical transmission of a sound signal (the first telephone was demonstrated experimentally in 1876).
In fact the inspiration for this first stereo sound experiment appears to have been the stereoscope, the then-popular trick of printing side-by-side photographs taken with cameras separated like two eyes. This binaural transmission was demonstrated by inventor Clement Ader, who used pairs of microphones in front of the stage of the Paris Opera, sending signals to left and right earpieces of listeners elsewhere in the city. Note ‘pairs’, plural: this was in the days of single-point-to-single-point telephone, before multicasting. In fact the electronic amplifier hadn’t yet been invented, so a telephone circuit consisted of a battery, a carbon microphone, an electromagnetic earpiece, and a lot of wire. Having a telephone was a very exclusive luxury at that time: having two (left and right) made the whole thing effectively a millionaire’s plaything, but it proved the point. There’s no evidence that anyone tried, or even thought of trying, to record these signals for posterity.
Not surprisingly, binaural didn’t catch on at once. The telephone remains to this day resolutely mono, and recordings were invariably mono for many decades. There were, however, some early experiments with binaural radio transmission. In the 1920s, a few American radio stations set up crude but effective systems with each microphone feeding a separate AM transmitter, each tuned to different frequencies. The listener at home needed two receivers, one tuned to each radio frequency, with one earpiece fed from each. Loudspeakers were in their infancy at this time and most people still used earpieces, so the concept of headphone stereo was an easy one to grasp.
The most famous name in stereo history is certainly that of Alan Dower Blumlein, the man who is credited with ‘inventing stereo’. In fact his original patent on the subject, rather blandly titled ‘Improvements in and relating to Sound-transmission, Sound-recording and Sound-reproducing Systems’, acknowledges as a well-known fact the possibility of conveying a stereo soundstage using what we now call binaural techniques. Rather confusingly, he uses ‘binaural’ in the way we now use ‘stereo’, that is, meaning any two-channel system that gives an impression of lateral sound placement. But the actual invention he describes is loudspeaker stereo, the system most often used ever since and which replaced true binaural as the default two-channel system. Ironically, his original method of obtaining loudspeaker stereo was to process the two channels of a binaural microphone system, and indeed one can convert between binaural (headphone) and loudspeaker stereo with a suitable signal processing chain.
Although Blumlein showed conclusively that the correct way to record stereo for loudspeaker replay was using either ‘coincident’ (realistically, very close) directional microphones or head-spaced microphones with suitable signal processing, plenty of engineers over the years have used other techniques that are more or less closely related to binaural.
Recordings made deliberately for binaural usually use omnidirection microphones, which pick up sound equally from any direction, though obviously if these are mounted in a dummy head their spatial response is modified a bit.
The frustrating thing for the lover of binaural – which, once heard in full flow, is a very addictive thing – is that it’s usually impossible to know in advance which recordings were made with suitable microphone techniques. Practically no LPs or CDs give information on microphone placement, unless they happen to include a session photograph of sufficient clarity, and even knowing the name of the recording engineer or producer doesn’t help because most recording folks have plenty of different tricks to choose from depending on recording venue, how many musicians are in the recording, and what mood they’re in as they set up for the session.
Just occasionally, recordings are actively trumpeted as binaural – recordings and, perhaps more often, radio broadcasts. The most consistent advocated of binaural on the airwaves has been American broadcaster John Sunier, who from 1985 until 1998 hosted a programme called ‘Audiophile Audition’ on national public radio. A web-based magazine grew out of the programme and continues to fly the flag for binaural, though its remit is very much wider than that. Another website, The Binaural Source, features more information and even some demos, though as they are all MP3 their quality is somewhat limited.
The BBC has also broadcast binaural material, rather inconsistently. Broadcasters are generally reluctant to use any stereo technique other than straight ‘intensity stereo’ (i.e. Blumlein coincident-pair, or panpot stereo) because so many people listen in mono and straight downmixing of binaural can give some rather odd ‘comb filter’ effects, but Radio 4 has an honourable history of binaural drama, going back certainly to the 1970s when some Sherlock Holmes-based dramas were put out in binaural. A noted recent success was the drama ‘The Dark House’, which was broadcast in 2007, and the technique continues to be used for special projects.
Most of the above assumes simple microphone techniques, basically a pair of microphones, one per channel. But what about the recordings – the vast majority from the last 40 years – which have been made with numerous microphones either recorded to multitrack and mixed subsequently or mixed down ‘live’? It would be perfectly possible to mix these for binaural listening but it’s practically never done. Conventional ‘panpot’ stereo mixing gives perfect Blumlein-style stereo which is not particularly well suited to headphone listening. Yes another reason to hate multimic recordings?
When all’s said and done, it can be perfectly pleasant, and often surprisingly convincing, to listen to Blumlein stereo on headphones, and indeed binaural recordings on loudspeakers. The real thrill comes when one hears a well-made binaural recording over good headphones: now that’s imaging!