BOOK REVIEWS (AA, Two, 1989)

by Gary Galo and Rod Rees

BOOKS Reviewed by Rod Rees

Perception of Reproduced Sound, edited by Bech and Pedersen; The Symposium, Technical University of Denmark, 1987; hardbound, 166 pp., plus softbound Annex. Available from Old Colony Sound Lab. PO Box 243, Peterborough, NH 03458, (603) 924-6371; $35, plus $1.75 shipping.

THE HISTORY OF reproduced sound can be viewed as a series of three overlapping periods. For the audiophile, these periods show up in the philosophy apparent in the creation of reproduced audio experiences, and in the kinds of articles that appear in audio publications such as TAA.

First, the Intuitive Period, during which the intuitive skills of audio inventors dominated, might have begun when some early hominids cupped their hands to shout across the valley, thus inventing the horn-loaded loudspeaker. The fabled musical instruments of a few hundred years ago came out of the Intuitive tradition. Then a host of twentieth century Intuitive audio inventors gave us a fascinating array of sometimes spectacular, but always flawed, audio equipment.

In the past 20 years we witnessed the Technological Period develop, during which audio engineers with analytical skills arrived at detailed technical knowledge. I think of A.N. Thiele working through the parameters of vented-box loudspeaker design, for example. The in credible jumps in the quality of reproduced sound during this time span can be credited in large part to these technological wizards.

The third period is well-represented by this collection of research articles, Perception of Reproduced Sound edited by Soren Bech and O. Juhl Pedersen. Let ’s call it the Psychophysical Period, borrowing an appropriate term from the field of sensory psychology with due respect to Gustav Theodor Fechner (1801-1887).

The Psychophysical Period began with the publication of Fechner ’s Elemente der Psychophysik in 1860, and research on the sensation and perception of sound has been a mainstay of experimental psychology throughout the past 100 years.

Carl Stumpf published his two volumes on the Psychology of Tone in 1883 and 1890. More recently, Carl E. Seashore (1866-1949), S. S. Stevens (1906-1973), and Richard M. Warren became leaders in the psychophysical consideration of sound.

In general, the psychophysics of sound has been the province of academic psychology and corporate research facilities such as Bell Laboratories, and few facts have leaked out to the general audiophile. But I see the glimmer of a ‘new age’ in the audiophile ’s world as psycho physical research becomes more accessible to a wider audience. Called ‘subjective evaluation ’ by the editors of Perception of Reproduced Sound, psychophysics is nothing more than the attempt to re late the perceived qualities of stimuli to the physical attributes of those stimuli.

In their Preface, Bech and Pedersen state: Today, it is thus possible to make subjective measurements of perceptual qualities as accurate as the well known objective measurements of physical qualities. But to ensure this accuracy a number of precautions has [sic] to be taken in planning and accomplishment of the measurements and in interpretation of the resulting data if the results shall not be seriously misleading.

Indeed, Fechner might well have said this 130 years ago. The time is right, I say, for audiophiles and the audio industry to embrace the Psychophysical Period.

So, what specifically does Perception of Reproduced Sound give us? Structurally, this 166-page volume, plus a small softbound ‘Annex,’ consists of 18 research articles covering a wide variety of topics of considerable interest to audiophiles and audio designers. The articles are brief with numerous figures, tables, and references; the look and feel is generally pleasing (depending on the quality of computer printers used for the different articles); and the technical knowledge you require should not be a problem for most audiophiles.

I ’ve grouped the 18 articles into six categories, although they are not organized as such within the book. Within each category the quality and usefulness of the articles vary widely. Most are directed toward loudspeaker testing but can be extended to other audio components as well.

Two articles deal with general psycho-physical research methodology. T. Poulsen presents a decent overview of signal detection theory, but is very sketchy on scaling methods. H. Staffeldt provides a useful overview of the widely used statistical technique called analysis of variance (ANOVA). Anyone who has not had a statistics course covering this method or who needs a refresher can benefit. Both articles include numerous references from which specific details can be filled in.

The largest group of articles, seven, deals with general procedures for con ducting listening tests and factors that can confuse the results. Floyd Toole and A. Gabrielsson each present good over views of the factors that affect the generality of loudspeaker listening tests, including consideration of individual differences between listeners and of instructions. Four other articles are of limited scope or require an effort to absorb a relatively small dose of information.

Two excellent articles deal with the specific issue of the perception of timbre, although Plomp ’s would benefit from less technical jargon and more common language interpretation of the results.

Staffeldt’s article ( ‘Evaluation and Scaling of Timbre in Listening Tests on Loudspeakers’) should be required read ing for all audio professionals, and I recommend it for every audiophile.

Two more excellent articles cover the perception of auditory space. J. Blauert discusses many interesting facts and hypotheses, and Gunther Theile presents the theory that stereo playback does not ‘‘reproduce ’ the original sonic sound stage but ‘simulates ’ it in somewhat the same way that a 2-D picture simulates 3-D visual space. On the negative side, both articles are rife with typographical and grammatical errors.

Three articles deal with the general problems of setting up a listening test.

V. Hansen describes how Bang & Olufsen chose and extensively trained a long term listening panel, and Fincham describes the listening room and test procedures used at KEF Industries. H. Moller describes several ‘official’ specifications for standardized listening rooms, including ‘Listening Tests on Loudspeakers’ (IEC Publication 268-13). All three articles are worthy, Hansen ’s particularly so because it documents the extensive and detailed training you need to become a skilled listener.

Two articles deal with the special problems of reviewing loudspeakers for the press. Martin Colloms discusses the ‘subjectivity of objective measurement ’ and points out how different varieties of that old stand-by, frequency-response measurement, can tell different stories about loudspeaker quality.

Overall, I rate 12 of the 18 articles to be of sufficient interest to TAA readers to warrant consideration. The article on ANOVA has rather less general interest, and the remaining five tend to be either overly specific and obtuse or overly general and simplistic. In many articles the number of typographical and grammatical errors is excessive, perhaps because English is not the primary language of most of the authors.

This research article collection ’s main contribution is its focus on the techniques of factor analysis and multidimensional scaling of the perceptual dimensions of reproduced sound. This research further de-mystifies the poetic vocabulary of audio description, and is an excellent starting point for audiophiles to enter the Psychophysical Period of sound reproduction.

Reviewed by Gary A. Galo Contributing Editor The Audio Dictionary by Glenn D. White; University of Washington Press, 1987; 291 pp., softbound. $14.95.

The Illustrated Dictionary of Electronics by Rufus P. Turner and Stan Gibilisco, fourth ed.; TAB Professional and Reference Books, 1988; 648 pp., softbound. $24.95.

Both available from Old Colony Sound Lab, PO Box 243, Peterborough, NH 03458, (603) 924-6371, FAX: (603) 924 9467.

GLENN WHITE ’S AUDIO DICTIONARY is one of the most impressive books to cross my desk in some time. White is the North west district manager of B&K Instruments in Seattle and previously lectured in music and engineering at the University of Washington. His introduction states that he diligently attempted to avoid mathematics in his discussions, making the text understandable to the non-specialist; he has succeeded admirably. Throughout, the approach is direct and concise. You need not be a physicist to understand his writing, and yet you won ’t put the book down feeling as if something has been omitted. A teacher who can synthesize complex material in to an understandable form without re sorting to oversimplifications or sliding into inaccuracies is rare, but White is such a person.

White rightly points out that inadequate audio education in our schools and universities results in the spread of myths and old wives ’ tales. Dispelling those myths is a principal goal in my Audio Fundamentals course at the Crane School of Music.

The Audio Dictionary covers much of the audio field, including audio reproduction, sound recording and reinforcement and acoustics. I picked definitions to cite that in other books are often thorny and usually highly mathematical, vague or downright inaccurate.

Let ’s start with root mean square (RMS). How many readers really under stand this term? White gets right to the point. Square the signal mathematically, average it over time, and take the square root of the average. How about impedance? White explains the differences between DC resistance and the loads presented to alternating current. His definition is lengthy, but explains how inductive and capacitive reactances are frequency dependent.

Damping factor is defined as the ratio of load impedance to amplifier source impedance, which is exactly right. He points out loudspeaker cable effects and how damping factor affects the sound.

I’m sure many believe it has no effect on the sound at all.

In his definitions, White capitalizes any term defined elsewhere in his dictionary. Thus, in a sort of zig-zag fashion, you can look up analog-to-digital converter, and work your way through digital, signal, analog, sampling, binary, amplitude, waveform, compact discs, amplifier, voltage and linear, to arrive at a concise understanding of digital audio basics.

I like his explanation of horn loudspeakers, since I do not like them. He states, ‘High-frequency horns also do not have quite as smooth a frequency response as direct radiators, and they pro duce significantly greater distortion as well, especially at high sound levels.’ I wish more writers would state this in print.

I found a few definitions that dissatisfied me, such as bias as related to tape recorders. He states that the high-frequency bias signal reduces distortion by making the recording process more linear, but he doesn ’t explain how. Bias is another greatly misunderstood term in audio and I wish this definition had done more to clarify it. His definition of micro groove states that Edison ’s LP record played for 40 minutes on one side. The playing time was actually 20 minutes per side, or 40 minutes total. This was still a remarkable feat for an 80-rpm record in the late 1920s.

White concludes the book with seven excellent appendixes, some of which pro vide more detailed explanation of topics such as impedance, audio measurements, balanced lines and digital audio.

He also presents an appendix entitled How to Subdue a Hi-Fi Salesman. Here he gives useful tips on shopping in a stereo store, including how to show the salesperson you know what you ’re doing.

This really isn ’t too difficult; I honestly believe audio salespersons are the largest group of incompetents in the country.

Most would be better off selling shoes.

Even in ‘respected’ high-end shops, how often do salespersons without any real technical background get defensive about technical matters? High-end audio equipment must be designed by mystics, not engineers.

I suspect that because of The Audio Dictionary ’s universal logic, engineers and audio newcomers equally will greet it with enthusiasm. White obviously possesses a broad background and he offers occasional historical insights. This book belongs on the bookshelf of any audiophile, whether professional or amateur.

The Illustrated Dictionary of Electronics is another outstanding offering.

Many electronics dictionaries are in print, but I doubt that you ’ll find one more comprehensive than this volume by Turner and Gibilisco. With roughly 35 definitions per page and more than 600 pages, this book contains about 21,000 definitions. The publisher claims more than 27,000 entries, including all the abbreviations. Although this book is technical, the definitions are clear, enabling anyone with a background in basic electronics to understand them.

Most definitions are concise, without a lot of supportive information. When the definitions contain terms covered separately, these are italicized. The authors cover all areas of electronics, including computer terminology, and include more than 450 illustrations.

To ‘review’ every topic in this dictionary would require that the review be longer than the book. I’ll select a couple of definitions TAA readers are likely to investigate. Root mean square is defined as ‘the square root of the mean of the squares of a set of values. ’ Exactly right; concise and to the point, but the authors do not offer any explanation, as found in The Audio Dictionary. Impedance is defined as ‘the total opposition offered by a circuit to the flow of alternating cur rent. It is the vector sum of reactance and resistance. ’ The formula given for impedance, Z = (R? + X?) ’2, applies only when resistance and reactance are in series, a fact the authors fail to point out.

The formula for parallel circuits, Z = RX/(R? + X?) ’2, is not mentioned.

If you look up classes of amplification you ’ll find several definitions related to Class A, B and AB operation. The definitions are accurate and to the point.

Aperiodic is defined as ‘not characterized by predictable periods or steps,’ and the explanation states that non-resonance at multiples of any frequency is a criterion.

The definition for aperiodic damping is: ‘Damping of such a high degree that the damped system, after disturbance, comes to rest without oscillation. ’’’ These are excellent explanations. I have been frustrated in my studies of aperiodic loud speakers by unacceptable definitions.

I found a few problems with crossover network. As defined by the authors:

‘Following final amplification in a sound reproduction system, an active outboard filter circuit or a passive one within a speaker enclosure that usually blocks or passes high frequencies, the purpose being to deliver each band of signals to the correct loudspeaker.’

First, an active filter is not used following the final amplification. An active crossover will be found before final amplification. Second, the bandpass filter is not mentioned in relation to three-way Crossovers.

No dictionary encompassing such a wide area will be absolutely correct and complete. Those of us involved in the specialized field of audio will invariably find quibbles with general books that at tempt to cover the entire field of electronics.

The appendix features eight pages of common schematic symbols and a variety of tables, including resistor color codes and an element periodic chart. Tables of mathematical operations, logarithms, electronic abbreviations and units are also included. Every TAA reader should have at least one electronics dictionary on hand and you won’t do better than The Illustrated Dictionary of Electronics. This book and the Audio Dictionary are indispensable references.

Also see:

A SIMPLE HIGH-QUALITY CD OUTPUT AMP, By Jan Didden

AC SENSING CONTROL, By LB. Dalzell

THE DC-300BC--TUBE MOSFET AMP, PART II, by Hans J. Klarskov Mortensen

Prev. | Next

Top of Page   All Related Articles    Home