Now, let's seat them both in a living room and play them a perfect reproduction of the same performance. If the reproducer delivers to their ears exactly the same pattern of air vibrations that reached them in the concert hall, each listener will perceive an exact replica of the original sound. Each can choose to "tune in" to whatever aspects of the sound he normally listens to, and as long as they are delivered to his ears in exactly the same proportions as they were originally, the sound, to each listener, will be a perfect reproduction.
Now, suppose the reproducer is not perfect. (We're still looking for one that is.) Suppose it reproduces all of the information needed to hear the main themes and the rhythm, like a typical cheap phonograph, but tends to change harmonic structures and textures. Listener B, if he has had no practice whatsoever in analytical listening to reproduced sound, may notice that it is not a perfect reproduction, but he will find it pleasing because it gives him what he wants. Listener A, on the other hand, won't like the sound one little bit, even though he may not (through the same lack of listener training) be able to explain why he doesn't like it.
The fact that every loudspeaker is an imperfect reproducer of at least some aspects of a total sound is what accounts for the tremendous diversity of opinions as to what constitutes "the best" loudspeaker. Different listeners judge reproduced sound according to different criteria, and assign different "weights" to the various aspects of the sound.
This is why The Stereophile attempts, as much as possible, to describe the sound of loudspeakers that we test, as well as simply passing opinion judgments on them. This way, a reader who has found that he does not generally agree with our opinions may still be able to tell with some degree of certainty whether he will like a particular loudspeaker.
He may, for example, have learned that he prefers some forwardness to a neutral or distant perspective, that he values full, fat bass more than bass detail, and that his hearing limitations at the high end may make it immaterial whether the system extends to 18kHz or rolls off above 12kHz.
Component Considerations
In order for a test report to have any real validity, it must attempt to assess a component's inherent capabilities, rather than how it will sound with an amplifier or an acoustical environment that booms up the bass or rolls off the highs. Thus, all speakers, from the cheapest to the costliest, should be evaluated on the basis of their performance under ideal conditions—with the best possible associated equipment and in an essentially neutral acoustical environment. Then, when an amplifier or pickup is subsequently tested, and is described as having certain colorations (as heard through a top-grade speaker), the reader can get some idea as to whether the components in question will complement one another or aggravate one another's colorations.
Of course, components that are designed specifically for one another, like the Marantz turntable and pickup, must be reported as a unit. But as long as most components are sold as "universal" items, intended for use with a wide variety of associated components, they should be evaluated as such.
It is easy enough to sweep an oscillator through a loudspeaker's range and listen for peaks, dips, rattles and rolloffs, or to listen to some recordings and say this sounds distant or that sounds bass-heavy, but how does one divorce these attributes from the acoustical characteristics of the listening room, the normally wide variations in response and perspective from one recording to another, and the effects of individual preamps and power amplifiers on the sound of a loudspeaker?
Preamp influences are easily coped with. Since we have yet to find a preamplifier that does not have at least some audible effect on the signal, we do not use any preamp at all for most of our loudspeaker testing. Only power amps are used, driven directly from the outputs of a professional-type Ampex 2-track stereo recorder. The only controls in the circuit are the playback volume controls on the tape recorder.
The choice of a standard power amplifier was a bit more difficult. We had found, through many years of testing amplifier/speaker combinations of all varieties, that certain breeds of amplifiers—now represented by the Marantz 8B and Dynaco Stereo 70, to mention two—tended to elicit cleaner, more musical sound from most loudspeakers than any others, and yielded the best sound from those speakers which were theoretically closest to perfection. We adopted the Stereo 70 because, despite its much lower price, its sound is virtually identical to that of the Marantz (footnote 2).
The Listening Room
The main listening room—always an unpredictable factor in loudspeaker performance, was "chosen" to the extent that the last time Ye Editor was house hunting, he located a residence whose living room was close to the ideal proportions of 1 to 1.25 to 1.6, and whose longest dimension (18 feet) would support lows down to at least 30Hz. From that point on, the problem was to find out what the room's acoustical properties, after furnishing, did to the sound, to make necessary corrections wherever possible, and to make notes of the uncorrectable things so they wouldn't unduly influence future listening tests. Then there was the period of adjustment, during which we "lived with" different speaker systems over a period of time, and carted them around to the homes of other audio hobbyists in the area to get an idea of how the bass performance in our standard listening room compared with that in other, typical rooms. Midrange and high-end evaluation of the standard room was effected by listening to the speakers outdoors, where they were unaffected by room acoustics, and comparing their outdoor colorations with the ones heard indoors. We found that the room tended to be a bit heavy in the 40Hz bass range, slightly depressed in the "presence" range, and almost perfectly neutral at the high end. These factors are now taken into consideration when testing loudspeakers. When in doubt, speakers are still checked outdoors and in other rooms, but we have found it possible to obtain a fairly consistently accurate impression of the performance of most systems within the confines of the standard room.
Choice of Music
Program material for the tests is derived mainly from two-track stereo tapes. Most of these are tapes we recorded ourselves, for these (despite some flaws in mike technique, hall acoustics and so on) are known quantities. We know the mikes that were used, the mike setup, and the hall acoustics, and we know the capabilities (and the state of adjustment) of the equipment used to record and play back the tapes. We also occasionally find commercial discs that prove to be dependable sources of wide-range, naturally balanced and natural-timbre material, and excerpts from some of these have been taped for testing purposes. (Again, occasional checks are made to ascertain that what we're getting from the tapes of these sounds like the originals.) Summing Up
This whole setup may seem to some observers like the "tower of science" that was depicted some years ago in a Life magazine article—an inverted-pyramid-like, top-heavy structure erected upon a single toothpick, which was labeled "Basic Premise." We can't argue with this. All we can say in support of our structure is that it yields test reports that most of our readers have found trustworthy in terms of their own observations of the units we test.
Footnote 2: More recently, we have observed that the bass performance of a given loudspeaker is often markedly improved when it is driven by a good transistor amplifier, so until such time as the last top-notch tube amplifier becomes a thing of the past, we will be reporting on loudspeaker performance using two basic criteria: tube operation and transistor operation.
In order for a test report to have any real validity, it must attempt to assess a component's inherent capabilities, rather than how it will sound with an amplifier or an acoustical environment that booms up the bass or rolls off the highs. Thus, all speakers, from the cheapest to the costliest, should be evaluated on the basis of their performance under ideal conditions—with the best possible associated equipment and in an essentially neutral acoustical environment. Then, when an amplifier or pickup is subsequently tested, and is described as having certain colorations (as heard through a top-grade speaker), the reader can get some idea as to whether the components in question will complement one another or aggravate one another's colorations.
The main listening room—always an unpredictable factor in loudspeaker performance, was "chosen" to the extent that the last time Ye Editor was house hunting, he located a residence whose living room was close to the ideal proportions of 1 to 1.25 to 1.6, and whose longest dimension (18 feet) would support lows down to at least 30Hz. From that point on, the problem was to find out what the room's acoustical properties, after furnishing, did to the sound, to make necessary corrections wherever possible, and to make notes of the uncorrectable things so they wouldn't unduly influence future listening tests. Then there was the period of adjustment, during which we "lived with" different speaker systems over a period of time, and carted them around to the homes of other audio hobbyists in the area to get an idea of how the bass performance in our standard listening room compared with that in other, typical rooms. Midrange and high-end evaluation of the standard room was effected by listening to the speakers outdoors, where they were unaffected by room acoustics, and comparing their outdoor colorations with the ones heard indoors. We found that the room tended to be a bit heavy in the 40Hz bass range, slightly depressed in the "presence" range, and almost perfectly neutral at the high end. These factors are now taken into consideration when testing loudspeakers. When in doubt, speakers are still checked outdoors and in other rooms, but we have found it possible to obtain a fairly consistently accurate impression of the performance of most systems within the confines of the standard room.
Program material for the tests is derived mainly from two-track stereo tapes. Most of these are tapes we recorded ourselves, for these (despite some flaws in mike technique, hall acoustics and so on) are known quantities. We know the mikes that were used, the mike setup, and the hall acoustics, and we know the capabilities (and the state of adjustment) of the equipment used to record and play back the tapes. We also occasionally find commercial discs that prove to be dependable sources of wide-range, naturally balanced and natural-timbre material, and excerpts from some of these have been taped for testing purposes. (Again, occasional checks are made to ascertain that what we're getting from the tapes of these sounds like the originals.) Summing Up
This whole setup may seem to some observers like the "tower of science" that was depicted some years ago in a Life magazine article—an inverted-pyramid-like, top-heavy structure erected upon a single toothpick, which was labeled "Basic Premise." We can't argue with this. All we can say in support of our structure is that it yields test reports that most of our readers have found trustworthy in terms of their own observations of the units we test.
Footnote 2: More recently, we have observed that the bass performance of a given loudspeaker is often markedly improved when it is driven by a good transistor amplifier, so until such time as the last top-notch tube amplifier becomes a thing of the past, we will be reporting on loudspeaker performance using two basic criteria: tube operation and transistor operation.
