A New Way of Reproducing Stereo Sound
The next revolution in loudspeaker design comes from the
the patent specification entitled "Ambient Expansion Loudspeaker". As the title implies,
this technology expands the perceived sound stage to at least right angles relative to the listener position. It could be described as two speaker "surround" sound. It employs the ubiquitous dual voice coil in a novel, seemingly self-evident way.
The Problem
The promise of stereophonic sound through loudspeakers depends on the ability of the listener to perceive a difference between the left and right channels. In sensing this "difference" component, the ear should be able to reconstruct the acoustic signature of the original hall or studio as well as localize the relative locations of instruments or voices. The mechanisms involved are the same ones used in listening to live music (or, for that matter, any acoustic event). Sounds coming from the left are perceived first in the left ear, then travel around inside the head (the cranial shield) to the right ear reduced in intensity and arriving later in time. Thus the ear-brain system computes all of this instantaneously and correspondingly localizes a sound on the left.
This simplification does not explain why early stereo through loudspeakers could simulate credible left-right "ping-pong" effects, but confined them laterally to the space between the speakers without reproducing the slightest sense of the room where the recording was made.
It seems that the requirements for dimensionalizing sound in the lateral plane involve reflections from the side walls of the recording locale normally situated at right angles to a listener. In live sound these dimensional clues arrive at the listener's ears from reflections at right angles and even behind a listener. The ear-brain employs its phase sensitive mechanism to perceive these late arriving signals and combines them with intensity difference information to construct the subliminal acoustic architecture of the live acoustic event.
But when reproduced through two loudspeakers located at a finite distance, everything falls apart! The longer wavelengths required to simulate the original event cannot maintain their differential separation because of human physiology. The left and right ears are spaced approximately 8 inches (separated by the cranial shield). At approximately one-half the wavelength of this inter-aural ear spacing the difference information becomes commingled in both ears at about 400 Hz and perceived monophonically.
This means that even the very highest performing loudspeakers are perceived monophonically at frequencies lower than about 800 Hz. An experimenter can verify this by inserting a variable low-pass filter with cutoffs from 400 through 800 Hz and then switching between "STEREO" and "MONO" on his or her preamp. The switching differences will become increasingly inaudible below 800 Hz and then totally monophonic below 400 Hz.
The Solution
Is it even possible to design a loudspeaker capable of supplying "difference" information down to at least 100 Hz? The answer lies in another attribute of the human hearing mechanism called the "precedence effect". This says that two sounds of equal intensity from different locations can be PREFERABLY LOCALIZED if one of them is DELAYED in time. The ear-brain will tend to reject the delayed sound in favor of the earlier arrival.
By employing dedicated BINAURAL (dual) voice coils in each left and right speaker system, it is possible to route left and right total signals to the "primary" windings and filtered differential L-R and R-L to the "ambient" windings. In this way, the anti-phase differential radiates from the same acoustic source as the primary left and right signals.
Note that there is NO TIME
DELAY, either electronic or
acoustic, employed! Precise
localization is maintained,
accompanied by an enormous
increase in soundstage width
and depth.
