   
|
||
 September 15, 2005 Playing Audio Mind GamesIn the past, I’ve discussed how two-channel stereo more or less accidentally became the standard form of audio because of the limitations of the vinyl record, and that the future seems to belong to multichannel systems that can simulate the three-dimensional soundfield we actually live in (see "Captives of Two-Channel Sound," December 2003). But the fact is that we do sense sounds coming from all directions using only two ears, and one of the most fascinating corners of audio over the years has concerned the various schemes devised to produce a full ambient soundfield using only two channels of transmission. The human hearing mechanism is astonishingly sensitive to timing cues: the minute differences in time of arrival of various sounds at our ears allows us to locate them in space, without being aware that we are actually hearing time -- or phase -- variations. With most two-channel stereo systems, both our ears hear sound from both speakers, which limits our ability to mentally reconstruct an aural image. Headphones don’t have that drawback, and have been used since the beginning of audio to create a very realistic soundfield, under the proper circumstances. The most impressive use of phones in this area is what is known as binaural recording, in which the pickup microphones are placed in such a way as to mimic the positions of a pair of human ears. The purest example is a technique called kunstkopf (dummy head) recording, in which the microphones are placed in the ear canals of a model of a human head, so that the timing relationships of the sounds reaching them are basically the same as they would be if one were present at the recording. Listening to a kunstkopf recording through good stereo headphones can produce an amazingly realistic sense of space, but the technique has some drawbacks. One is that the phones -- and thus the entire soundfield -- move when the head is turned, which is highly unrealistic. Another is that center sounds tend to sound as if they are inside our heads, and sounds that should be up front are often perceived as being behind us. This seems to happen because, in a live situation, our brains assume that if the producer of a sound is in front of us, we will be able to see it; if we can’t, we locate it behind us. Perhaps the biggest problem with kunstkopf recording is that it doesn’t relate to how most recordings are made. Using the dummy head is fine only if it records something that takes place on a single occasion in a single place, such as a symphony orchestra concert. But most recordings are built up in layers from many tracks recorded over a period of time, sometimes in several locations, and these techniques don’t lend themselves to dummy-head microphones. But as the base of knowledge grows about just how various sounds relate to each other, it has become possible to synthesize the effects of the kunstkopf system -- to predict how a sound emanating from a particular direction will affect the ear and to then electronically tailor the sound to have that effect. One of the most interesting applications of such technology came from a Florida company called Virtual Listening Systems. Their Toltec Processing modified the output of a Dolby Pro Logic decoder so that, when heard through headphones, the sound seemed to come from where the various surround-sound loudspeakers would be. Because the system was designed for home-theater use, the lack of a visual reference was not a problem -- sounds did seem to come from the screen. Watching movies is usually a social activity, however, in which the use of headphones is often inappropriate. But as it turns out, much of the same effect can be produced from speakers as well. The trick is to find a way to modify the stereo signal so that each speaker "addresses" one ear only, just as each earcup of a pair of headphones affects only one ear. With varying degrees of success, that kind of modification has been around for a while under such names as Image Restoration and Sonic Holography or, in its simplest form, the Stereo Wide circuits found in some boom boxes. A processor analyzes each channel and predicts what its form will be when it reaches the opposite ear; the inverse of that signal is fed to the opposite speaker to cancel the right information in the left ear and the left information in the right ear. Once such crosstalk has been eliminated, a further processing stage applies the proper phase to each of the surround channels to make it seem to come from its appropriate location. Theoretically, any number of phantom speakers can be placed anywhere in the soundfield just by applying the right equations. It may sound farfetched, but the reality can be disconcertingly realistic. Several years ago, I heard the Virtual Home Theater system developed by Harman International, which used just such technology. Because they felt that this sort of system would be most attractive for such things as computer workstations, the demo used a small pair of multimedia speakers placed on either side of a 14" TV, with a small powered subwoofer off to the side. The results were remarkable. Even though the speakers were only about a foot apart, the front signals appeared to come from the conventional 60-degree spread, and the surround material was definitely in the rear. When test tones from the Pro Logic decoder were cycled from position to position, they sounded correct, even though all sounds were actually being reproduced from the two tiny speakers up front. One big benefit of such a system, of course, is that there is no tonal difference from channel to channel to channel: all sounds are produced by the same two, identical speakers. Dolby Labs took such efforts seriously enough to develop a virtual surround standard, and a number of products have appeared that have used fewer than the standard 5.1 channels for surround sound. Are such systems only curiosities? Perhaps, but this sort of technological wizardry does have a way of cropping up in different areas of audio altogether -- areas that might not have been possible without such tinkering. ...Ian G. Masters
|
||
|
