READ ALL ABOUT IT

During these times of economic recession, the audio industry, like so many others, has seen sales revenues drop significantly, re search and development projects cur tailed, and venture capital practically dry up. All of this has led to dreadful "restructuring" by many companies, with an attendant fallout of reduced staffs, production cutbacks, drastically restricted outlays for advertising and promotion, and myriad other unpleasantries.

Well, friends, the audio industry--be it the professional world of recording, broadcasting, or communications or the consumer world of the audiophile and music lover-has weathered other recessions. Audio is a technology-driven business, and every advance in the reproduction of sound has resulted in the introduction of desirable new products that create entirely new markets. Surely the best examples of this are digital audio and the Compact Disc.

Thus, in spite of financial constraints and the need for companies to run a tight ship, audio engineers and scientists in audio-related disciplines continue to toil in the research laboratories. They are bringing new understanding to audio and acoustic phenomena, developing new technologies and significant new products of ever-increasing sophistication.

Since the first Audio Engineering Society Convention in 1949, the AES has showcased the developments and products of its members and has pro vided a forum for the presentation of scientific papers covering their investigations. The AES alternates conventions each year between New York City and Los Angeles or San Francisco. Every spring, a European convention is usually held in London, Hamburg, Par is, or Montreux. For the first time, this year's (92nd) AES Convention was held in Vienna in March, and it enjoyed great success, producing a record number of papers covering a broad range of audio technology.

The AES papers, or preprints, as they are known, provide a fascinating glimpse into the cutting edge of audio research and developments. Some of these preprints cover very arcane subjects indeed, but there are quite a few that I feel would be of interest to the audiophiles and music lovers who read Audio. There is information in these preprints that is simply unavailable elsewhere.

On a very practical note, AES Pre print No. 3237, "Study of Corrosion Stability on DAT Metal Tape," written by F. Hayama et al. of Hitachi Maxell, Ltd., addresses a story that has been making the rounds since the introduction of R-DAT. This story purports that DAT metal tape is quite unstable and that in as little as three or four months, and most certainly within a year, some loss of signal can be expected due to corrosion of the tape coating. If tape is stored at high levels of temperature and humidity (60° C and 90% relative humidity) for one week, it is equivalent to four years of normal room storage at 25° C and 60% relative humidity! High humidity apparently can cause more destabilization than high temperature.

The conclusion of the Maxell engineers was:

As long as DAT cassettes were stored under a nominal environmental condition of 25° C and 60% RH for four years, the DAT tape was very stable and did not show any change in electrical performance. Even though under the conditions of high temperature and high humidity of 60° C/80% RH for one month, the degradation of RF output level and C/N were less than 0.5 dB and had few changes in audio performance.

However, in general, the lifetime magnetic tape also depends on the de compositions of chemical components, such as polymer binder and base film on the sticky phenomena on the magnetic tape caused by water between the tape layers. Therefore, it is better for users to store the magnetic tapes in the temperature range of 15° to 25° C (59° to 77° F) and humidity range of 40% to 60% RH.

If you use planar-type loudspeakers from such manufacturers as Quad, Magnepan, Martin-Logan, or Apogee, you know that the placement of these dipole speakers in a room is of critical importance for optimum sound quality and is a matter of considerable controversy. AES Preprint No. 3327, "Dipole Source Placement in a Room," by Jorma Salmi of Gradient, Ltd., provides invaluable information on the subject from both practical and experimental viewpoints. (Incidentally, Gradient, Ltd. is a Finnish company that manufactures an approved subwoofer for the Quad ESL-63US speaker.) Salmi concludes by saying that his . . . recommendation to follow the 1/3 rule . . is based mainly on minimizing the back wall effects at the midrange, assuming that this wall is highly reflective. The speaker can be placed closer to the wall if the wall is absorbent at middle frequencies. In practice, the minimum distance to the back wall is relative to the speaker's physical size.

However, we must keep in mind what will happen to the bass output.

If a dipole is weak in bass, positioning it close to the side wall will give some help but at the expense of the midrange quality.

A dipole source having a very light diaphragm (e.g., most electrostatic de signs) is generally more sensitive to the back wall effects and less sensitive to side wall effects than those having heavier membranes.

The position of the listener in the room relative to the speakers and boundaries is as important as the position of the speakers.

The influence of absorption is not discussed in this paper. The measurements were done in an empty room with highly reflective boundaries, so the results are presenting the worst case situation.

For recording engineers and dedicated audiophile recording enthusiasts alike, AES Preprint No. 3255, "Microphone Arrays Optimized for Music Re cording," by R. W. Woszczyk of McGill University, is a veritable goldmine of information on various mike arrays for stereo music recording. It has large, clear layout charts of patterns, polar plots, frequency response plots, and more on such setups as four coincident KM84s in line, four coincident KM84s in block, the famous Decca tree stereo array, various spaced arrays, X-Y, and ORTF. Three other papers on mikes and stereo recording are also worthwhile: Preprint No. 3252, "Frequency Dependent Hybrid Micro phone Arrays for Stereophonic Sound Recording" by Michael Williams of Paris; Preprint No. 3254, "A Matrixed Pressure Triplet for Full Surround Stereophonic Pickup" by Andre L.G. Defossez of Brussels, and Preprint No. 3313, "Standard Stereo Recording Techniques in Non-Standard Situations" by Albert G. Swanson of Location Recording in Seattle.

By now most readers are aware of the concept of virtual reality, an exciting prospect for the not-too-distant fu ture. Through manipulation of high-power, ultra-sophisticated computer graphics, a person wearing special electro-optical equipment would enter a three-dimensional space that he perceives as being a real environment.

The environment could be whatever suits one's fancy, providing the computer graphics processor is programmed to simulate it. No doubt, a Walter Mitty type could become an ace Grand Prix driver, zipping his racer through the streets of Monte Carlo! This is the ultimate step beyond video games! How ever, a problem must be solved before virtual reality can more closely simulate the sensory inputs of a real event or environment. As one would expect, to accompany the visual stimuli we must have sound of "virtual reality" quality.

Binaural sound might seem the obvious solution, but it has always been plagued by the "sound inside the head" phenomenon and, worse yet, the inability to perceive frontal localization. There have been attempts to rectify this by using special dummy heads coupled with equalization processing.

I have heard quite a few of these special recordings, but none of them really provided me with a convincing sense of frontal localization.

Much work is in progress to solve the problem of frontal localization and sound externalization, most of which involves digital signal processing, and five fascinating AES preprints cover various approaches to binaural recording problems. An offshoot of this re search-and a very important thing for the many people who listen to stereo phonic music through headphones-is digital processing that can simulate loudspeaker listening while listening through headphones. The preprints on binaural sound are: No. 3291, "Im proved Externalization and Frontal Perception of Headphone Signals," by Soren Gert Weinrich of Oticon NS Re search Unit in Snekkersten, Denmark; No. 3323, "BAP Binaural Audio Processor," by F. Richter of AKG Akus tische in Vienna; No. 3289, "Head-Related Transfer Functions: Measurements on 24 Human Subjects," by Dorte Hammershei et al. of the Institute for Electronic Systems at Aalborg University in Denmark; No. 3290, "Transfer Characteristics of Headphones," by Henrik Moller et al., also of the Institute for Electronic Systems, and No. 3332, "Improved Possibilities of Binaural Recording and Playback Techniques," by K. Genuit et al. of HEAD Acoustics in Herzogenrath, Germany.

Preprint No. 3343, "The Sound of the Orchestra," by Jorgen Meyer of Physikalisch-Technische Bundesanstalt in Braurschweig, Germany, explores the physics and acoustics of the instruments in a symphony orchestra and the qualities that determine the characteristic orchestral sound. The various instruments' transit and decay times, resonances, sound power output, and dynamic expression are analyzed.

Meyer used the orchestra of the Vienna Technical University for his research, and his sound examples are from their performances of the Overture to Die Freischutz by C. M. von Weber and Joseph Haydn's Symphony No. 89.

The examples illustrate: ... the chorus effects produced by the groups of the strings, tone mixture in unison and in octaves, location-related tonal balance between upper and lower voices, seat arrangement of the strings

-motif separation and blending [and] loudness level balance between instrument groups-masking in fortissimo, dynamics and space impression [and] time structure of the onset of tutti sounds. This is, indeed, a fascinating and instructive paper.

In the March issue, I reported on Ambisonic sound. Michael Gerzon and Geoffrey Barton of the U.K. are major proponents of this technology. Their AES preprint, No. 3345, "Ambisonic Decoders for HDTV," offers very persuasive arguments for Ambisonics as the multi-channel sound that will complement HDTV. However, it will have to compete against the new Dolby SR-D digital stereo film format. As I noted in the February issue, Dolby SR-D follows the SMPTE layout for six discrete channels of sound. Time will tell which for mat will predominate, but at least owners of typical home surround theaters will be able to play either one.

There are many other preprints from the Vienna AES Convention that would be useful to audiophiles, but I've tried to pick the plums. Each preprint costs $5 and can be ordered from the Audio Engineering Society, 60 East 42nd St., New York, N.Y. 10165.

(adapted from Audio magazine, Jul. 1992; Bert Whyte)

= = = =