Those who hold measurements in higher regard than listening experiences have a valid argument with regard to the variability of the human listening experience. Those who prefer to take advice from a skilled human-ear/brain combination have a solid position regarding the validity of pass/fail marks from a machine that can't tell Bach from Beatles.

To paraphrase an old saying: "neutral is as neutral does." Although conventional testing procedures were developed with the intent to insure the neutrality of audio components, arguments and verbal confrontations abound because many listeners feel that they have experienced this conundrum: better specifications often equal better sound. Their problem is with the word "often." Most audiophiles feel justified in their feelings of confusion (and maybe even "betrayal!") because the word should be "always." This has created a distrust of specifications, a too-often-confused consumer, and an entire industry: tweaks. (The resulting defacto "working definition" for the word "neutral" is closer to an audio-specific definition of the word "transparent": a lack of gross colorations or other sonic aberrations which would hinder the ability to hear small differences in the sound of recorded media and other components in that system.)

Because of this confusion, this essay introduces several new terms which have been coined in an attempt to resolve the problem. These new terms are utilized in a conceptual framework that attempts to unite the "perceived" with the "measured."

The first new term is actually a new label for describing the goal of conventional measurement techniques which have as their focus, measuring individual components (see Appendix B). This goal's new label: Static Neutrality. How well a component measures is the benchmark of its static neutrality. However, high levels of static neutrality have NOT historically correlated with high marks for perceived musicality.

The second new term is Dynamic Neutrality and is today only applicable to audio SYSTEMS. This term has to do with the distribution of mechanical and acoustical resonances in an audio system and how these resonances affect its perceived harmonic balance and neutrality. Unfortunately, test instruments other than the human ear do not as yet exist. (Or if they do, they are very very expensive and are rarely found outside of a high-tech research lab.)

Can't Get There From Here, Today...Maybe TomorrowAll acoustic and electric instruments are physically constructed from materials. These materials have mechanical resonances that will differ from instrument to instrument. These resonances will insure that each instrument, whether acoustic or electric, has its own individual sound. (Axiom #1 applied to instruments.)

As all stages of recording and playback equipment use wire or other conductors in their manufacture, the differing mechanical resonances of each stage (recording, mastering, and playback) will cause every stage to have its own individual sound, regardless of that stage's static (measured) neutrality. (Axiom #1 applied to all stages of recording and play back equipment.)

In order to preserve the individual resonance signature of the instruments that are being recorded or played-back, the mechanical resonance signature of the recording, mastering, and playback systems would each have to be dynamically neutral (presuming that the measured (static) neutrality level is high.) To accomplish this, each stage would have a wide band resonance response consisting of evenly distributed, overlapping, low-Q resonances. This design philosophy is rarely understood and even more rarely implemented.

(High-Q resonances accentuate fewer notes and overtones than low-Q resonances, but the amount of accentuation will be much greater. This is especially problematic if these few notes/overtones had previously played only a minor role in that instrument's harmonic structure. Low-Q resonances minimize the amplitude of these contributions and also cause less discrimination between one note and its neighbors, thus providing a more level support for all of music's notes and overtones.)

Audio equipment is supposed to be a "reproducer" of sound, not a "producer". Unfortunately, this is not true. Even if today's levels of static neutrality had achieved perfection, the lack of a scientific measuring stick for dynamic neutrality precludes widespread advancement in the ability of designers and manufacturers to build truly neutral audio systems...systems capable of the sonic recreation of a live musical event.

GO YOUR OWN WAY

A manufacturer of camera equipment makes tools for a photographer to use in making photographs. A manufacturer of violins makes tools for a musician to use in making sound. In the same way, a manufacturer of audio equipment makes tools for the serious listener to use in making the-sound-of-a-system-playing-music. (A sonic-portrait?)

In order to make tools, it is necessary to use tools. The tools used to manufacture other tools include machinery, design philosophies, and measurement tools and techniques. The design philosophies determine the "what," and the measurement apparatus determine and establish verification of "how well." These production tools shape, and more importantly, define the resulting manufactured tools.

The production tools of the manufacturer and the serious listener differ in that design philosophies are only of passing interest to the serious listener. He or she will eagerly subscribe to any and all design philosophies (and even change them in mid-stream) if more musical enjoyment is the result. Many (most?) designers/manufacturers have their design philosophies nailed firmly in place.

The measurement apparatus of the two also differ. The serious listener's apparatus is personal in nature: his or her ears and musical tastes. A manufacturer uses many measurements in an attempt to be universal. All tools used in the manufacturing of a product (including test methods), act as fulcrums that determine the direction taken by that product. However, there are inevitable problems that must occur when using a set of measurements that does not include tests for a component's resonance response. Because manufacturers are constrained with a limited set of "directions," they may not be able to "take you where you want to go." Achieving your sonic goals usually requires trying out products from many manufacturers...which can be defined as a guessing game.

Some manufacturers include subjective listening tests as a counterbalance to measurements. This is certainly a step in the right direction. However, it would be impossible for a manufacturer to conduct listening tests with all the possible combinations of equipment that could get used with its products.

The public, however, is not limited to a small number of combinations of components, and their goals are personal, not universal. That which will EXCITE one, will put another...to sleep.

The serious listener's goal of enjoying his music is quite a different goal from that of a designer or manufacturer: A manufacturer hopes to build a better tool; a serious listener hopes this tool can be used to smooth his path to musical enjoyment. Unfortunately, because of the limited scope of measuring tools available to manufacturers, listeners will have the ultimate responsibility for achieving maximum pleasure from their systems. Fortunately, tuning products abound, and this article will help listeners understand the necessity of a balanced resonance response, and apply the basics of tuning for the purpose of achieving balance in their personal (and unique) listening systems.

THE MISSING LINK

The concepts of accuracy and musicality can be seen as the basis for the two categories of neutrality. The history of audio componentry is peppered with instances of components that measure well but don't make the grade sonically, as well as components that measure poorly but are pleasant to listen to.

H. H. Scott was a manufacturer of tubed audio components back in the 1950s and 60s. The chief engineer was Mr. Donald von Recklinghausen. His most famous quotation goes something like this: "If a component measures good and sounds good, it is good. If a component sounds good but measures bad, you're measuring the wrong thing." To say that the wrong measurements have been made for decades would be to take the easy way out. It might be far more appropriate to say that in the absence of the "right" measurement, too much emphasis has been placed on the "wrong" measurement. While this "wrong" measurement is still an appropriate and valuable measurement to make, it is just not the most important measurement anymore. (The "wrong" measurement could be any of those used for measuring static neutrality. See Appendix B.)

The author hereby proposes a conceptual system that illustrates common ground between the "perceived" and the "measured." A Venn diagram is provided to make the mental leap less perilous. (See figure 1.)

Note: Areas 1-6 are subsets of the super set: All Audio Equipment.

Area 1:

A.P.T.: This stands for Asymptotic Pure Tonality and is a goal for both systems and components. (Please note: This is not necessarily THE goal, but it is definitely A goal.) Once a product/system has achieved exceptional levels of measured performance (a high degree of static neutrality) as well as an evenly balanced resonance response (a high degree of dynamic neutrality), its tonality will approach that of real life. However, nothing man-made is perfect, and we will always be approaching, but never reaching, the "Pure Tonality" of real life. (The definition of an "asymptote" is a straight line that continually gets closer and closer to a line with an infinitely extended curvature, but never intersects it.)

Area 2:

High Degree of Static Neutrality: There are numerous tests and measurements based upon sine and square waves: distortion, phase shift, frequency response, slew rate, dispersion, signal-to-noise-ratio, and other measurements such as rise time. These tests are representative of conventional thinking on the possible distortion mechanisms of audio components. For a system/component to occupy this subset of the larger set "Accurate," most conventional tests must be passed at the highest levels possible.

Area 3:

High Degree of Dynamic Neutrality: An even distribution of low-Q resonances, with few irregularities, is required for an audio system or component to reside in this subset of the set "Musical." At the present time, the test methodology relies on the human ear/brain combination.

Area 4:

"Accurate" Systems and components that measure well in at least a few of the many conventional measurements reside in this group.

Area 5:

"Musical": Systems and components that have a pleasant sound, rather than an "analytical" sound, reside here. Area 6:

Not Musical & Not Accurate: Plastic clock radios, many boom-boxes, and some rack systems.