Header image  
by Ralph Glasgal
line decor
Home Tutorials Tech
Kudos and
Demos Bio Free Ambio
Glossary The Home
Concert Hall
Rec Engineers
FAQ/Forum Links Contact us
line decor

Audio Engineering Society

Presented at the 99th Convention
October 6-9, 1995 New York, NY



The Synthesis of Concert-Hall Sound Fields in the Home

by Ralph Glasgal


The basic technical premise of Ambiophonics and most binaural sound research is that if you can present to the ear via recording a close replica of what the ear would or could have heard in the concert hall or other musical venue,then the goal of "You Are There" realism for the reproduction of recorded music must be achievable. To accomplish this we do not necessarily have to know everything there is to know about how the ear-brain system works. We just have to deliver the proper sonic goods to the entrance of the ear canal, including pinna and head modulation. Ambiophonics does not rely on any psychoacoustic shortcuts, but does make use of the psychoacoustic research that defines what it takes to make a musical performing space sound good. Ambiophonics does not extract recorded concert-hall ambience signals from the LP or CD. what it does do is synthesize the necessary early reflections and reverberant tail in a way that matches as closely as possible the space in which the recording was made or creates a space in which a dry recording will sound best, i.e., pipe organ music in a cathedral. Thus Ambiophonics may not always generate a hall that exists, such as Carnegie Hall, but it always produces an ambience that, coupled with the Ambiophonically transformed front channel provides a sound field that the ear will accept as real and as a well-designed performance space.


James Moir has reported that many, or even most, blindfolded listeners can locate a caller indoors at a distance of 55 feet, with an accuracy of about l°. with the sound source limited to the band between 3000 and 7000 Hz, the angular localization was twice as good. It is interesting that this is also roughly the frequency range of 2 to l0 kHz where the combined pinna and head diffraction frequency responses gyrate the most and where interaural intensity differences are said to govern binaural perception. When it comes to the detection of sharp transients such as clicks, music, or moving objects, the ear may be even more sensitive in the horizontal plane.

To put this astonishing sensitivity of the ear in perspective, a movement of one degree in the vicinity of the median plane corresponds to a differential change in the arrival time at the ears of only 8 µsec. Eight µsec is equivalent to a frequency of over 100,00 Hz and it is unlikely that the ear is responding to such a small change in delay directly. Eight µsec also represents a phase change at 10,000 Hz of about 30° between the ears, but again there is no evidence that the ears can make phase comparisons of this type at such high frequencies. Thus it is overwhelmingly likely that it is the very large frequency-response shifts involving a wide band of frequencies at each ear and between the ears that occur for small movements of a source or the head and it is these gross changes in microresponses that make the ears so sensitive to position. Researchers at the acoustics laboratory of Aalborg University have shown that reflections off the various parts of the pinnae produce comb-filter peaks and nulls in such profusion that they are almost impossible to plot. Just considering frequencies below 10 kHz at least one null of 30 db is possible for most people at points close to the medial plane on the side facing the sound. Above l0 kHz, the dips and peaks are even more extreme, but it is not clear that such high frequencies play a significant role in frontal localization. The response of the ear on the far side of the head is even more irregular and depends on individual head, pinna and nose shapes. Sound from the right side will reach the left ear by diffraction over the top of the head, the back of the head, under the chin, around the nose, and all points in between. Each of these sounds is then altered further by the far pinna and this process is extremely frequency- and angle-dependent, as the sounds arriving by these many paths hit the pinna and add or subtract at the entrance to the ear canal, producing a very irregular and serrated frequency response with dips of up to 40 db.

The intensity differences at each ear, and especially between the ears, are magnified by comb-filter effects in each ear, first caused just by the pinna and then by the interaction at the remote ear between its pinna and the signals going around and over the head. One can easily see that a minute shift in a sound source could cause a comb-filter null at the near ear to become a peak and a peak at the far ear to become a null. As D.'B. Keele's measurements show, swings of 20 db or more in interaural intensity are possible with a change in interaural delay of just 63 µsec. The frequency of such a peak /null pair can shift more than one fifth of an octave for the same delay change. To put this in perspective, a 63 µsec delay change can be introduced by moving one loudspeaker .7 in. closer to the listener.

The point of all this is that reproduction-system errors that cause differential delay in the 2-to-10 kHz region, or that have errors in the direction from which signals originate, will affect the ear's ability to determine position accurately from pinna, head, and interaural intensity cues. Furthermore, if the cues provided by phase, or time of arrival of lower-frequency sounds, conflict with the higher-frequency, comb-filtered directional cues, location vagueness results, and the brain says the music is canned or that the reproduced sound lacks depth, air, or presence.

Next Page >>