Intensity-based stereophonic systems

The earliest recorded spatial audio experiments including, recording, transmission and reproduction were carried out at the Electrical Exhibition (‘La Lumiere Electrique’) held at the Paris Grand Opera in 1881 (Moncel, 1881). The system was designed by Clement Ader, and utilised four microphone pairs connected to binaural telephone receivers in four remote locations in the opera house allowing visitors to the exhibition to listen stereoscopically to the live performance.

a) b)

Fig. 2.5 a) The first stereophonic transmission at ‘La Lumiere Electrique’ by Clement Ader. b) Advert showing the subsequent commercialisation of the Theatrophone for stereophonic transmission and

However it was not until 1931 when the theoretical background for stereophonic systems was developed, following the experiments by Blumlein and his subsequent patenting of the first intensity stereo system (Blumlein, 1931), this contribution was complemented by stereo recording methods using multiple microphones.

Around the same time the Bell labs were also looking into spatial audio reproduction techniques and the work by Steinberg and Snow (1934) proposed an early three- channel stereo system, replaying a recording made with 3 spatially disparate microphones (shown in Fig. 2.7). Remarkably this early work into stereophonic reproduction shows many similarities to the now state-of-the-art wave field synthesis systems as described later this thesis. They suggested the idea of the Acoustic Curtain as shown in Fig. 2.6.

Fig. 2.6 Steinberg and Snow’s acoustic curtain concept, recording and recreating the wave fronts from

the stage (after Rumsey (2001))

The idea was to have a row of microphones across the stage that exactly matched the spatial configuration of loudspeakers at the reproduction end. Quite correctly, they pointed out that this would recreate the original wave fronts and therefore the localisation cues that would be present if the listener were at the location. However this was put forward as the ideal situation which due to technical constraints could not be practically realised. They therefore came up with a compromise solution which utilised just three microphones and three loudspeakers as shown in Fig. 2.7.

Fig. 2.7 Steinberg and Snow’s compromise version of the acoustic curtain concept, recording and

Whilst the system in Fig. 2.7 was original devised as a compromised version of that shown in Fig. 2.6, Steinberg and Snow did point out that these two situations are in fact different in the localisation processes, as the latter compromised version actually relies more on the precedence effect rather than the information from the wave fronts reaching the ears, however they note that the performance of the compromised system was actually very good, which provides another example of the surprisingly good performance of stereophonic reproduction which has seen it so widely adopted. Despite the works from Steinberg and Snow and Blumlein, it was not until the 1950’s that commercial stereophonic systems became realised; the market was fuelled in the 1960’s when two-channel vinyl records became popular. Stereophonic audio grew greatly in popularity and became ubiquitous for many years, to a great extent usurping monophonic reproduction.

Despite the success of two-channel stereophony however, there was still a desire in the cinematic world for a greater spatial impression and rendering to fit with the moving image; consequently quadraphony was introduced as a concept. Quadraphony provided an extra two channels at the rear of the listener, increasing spatial impression and the creative options available to the producer/sound engineer. Quadraphony was however rather unpopular for consumer audio systems primarily due to compatibility issues and consumers not seeing the need for extra loudspeakers in their living rooms thus it did not supersede two-channel stereophony.

With the invention of the CD in 1982 and the subsequent upsurge in digital audio technology, more audio channels have become easily available, driving the consumer market. Multi-channel sound reproduction is subsequently becoming more popular and is being utilised increasingly in an ever growing home cinema market. Since the consumer adoption of stereophony, people have increased the number of loudspeakers used within a given system such that it is now common to see 5.1, 7.1 and even 22.2 (Hamasaki et al., 2005) surround sound systems.

All of these systems use both level and phase panning in order to provide the necessary information for the listener to be able to localise the sound in either a two or three dimensional space, the greater the number of loudspeakers the greater the definition of the localisation. A problem with the aforementioned systems however is the so-called ‘sweet spot’. The systems are generally optimised for just one position meaning that the listener will only receive correct spatial imaging if he or she is sat in the correct place in the room. This is because sources are positioned in the reproduced sound scene based on level and phase differences between each loudspeaker signal (generating the required ILDs and ITDs at the listener); if the listener moves even a small amount, these cues will be altered and the perceived location of the source will change thus it is said that the phantom source is unstable. These problems can be reduced if the sweet spot is set to be further away from the speakers; however this method still only allows true surround sound for one or two individuals.