There are more things in heaven
and earth, Horatio, Than are dreamt of in your philosophy. ... Hamlet by William Shakespeare: Act 1. Scene V
[B
]"If the facts don't fit the theory, change the facts." --Albert Einstein[/B]
I am troubled by the assertions - common to this forum, that double blind testing has proven the inaudibility of differences among amplifiers. I, personally, have, on many occaisions, swapped amplifiers and receivers in and out of a high quality system and have heard differences large and small almost every time. I'm not saying that the amps didn't sound essentially alike on a superficial level - but once settled into the music apparent differences in subtle but important qualities like dynamics, soundstage width and depth, image specificity, rhythmic drive and timbral accuracy were audible. I don't propose that I'm particularly a golden eared person - although I do care a lot about audio quality and play with audio gear a good deal. Since the audibility of amplifier equipment is such a regular and familiar attribute for me (and for many other admittedly audiophile folk I know) I am forced to consider the following possible conclusions:
1) Every instance of what I (and what many others) hear is pure placibo effect.
2) The methodolog(ies) used in double blind listening tests is flawed or incomplete.
Excuse me if I come to the second conclusion. I find it credulous that folks feel that something as etherial and subjective as audibility and perception can be considered scientifically known and proved - especially given the limited set of attributes usually tested (THD, IMD, power). The fact is that there are many many aspects of amplifier performance that are not routinely tested that have been shown to have significant audible effects. See Daniel Cheever's thesis on amplifier audiblity testing methodology (
http://w3.mit.edu/cheever/www/cheever_thesis.pdf)
And then, tested exactly how? The AES lists a specific methodological standard for testing loudspeakers (
http://www.aes.org/standards/b_pub/aes20-1996.pdf) but none for testing amplifiers - presumeably because they have already dismissively determined that there is no point. However, even the standard for speaker testing says nothing about such important factors as listening selections or length of sample listening period. The fact is there is no consistency among double blind listening tests in terms of such important psychological factors as listening selection and length of sample.
Given these problems, it's more than credulous, it's fatuous to invoke science to dismiss arguements for the audibility of differences among amplifiers as Rip Van Woofer did:.
Mr. Van Woofer misses the point. It isn't that, by declaring the audibility of amplifiers, we reject the scientific method. Rather we reject the conclusions reached by the current spate of non-rigorously performed tests. Science has historically embraced many incorrect conclusions over the years, from phlogiston to the lumiferous aether. The fact that a current theory has some science to back it up isn't worthless - but it's only worth the quality of the science that it's based on. I haven't seen anything convincing at all in the science that the current AES theory that amplifier differences are inaudible.
It all comes down to the double-blind tests. The problem with the current methods of double blind testing (other than the lack of standards or consistency) is that they take music listening out of context. In an environment where test subjects are listening for differences, rather than experiencing the music emotionally, the test becomes one of whether you can catch an extremely subtle effect as an instanteous audible fact - not whether you can perceive it in the way that we listen to music.
The fact is (I certainly agree) that modern amps, as has been duly noted, are able to produce the full frequency range - with a fairly similar degree of resolution, so the instantaneous sounds are extremely similar. Listening for instantaneous sounds - e.*. "are the drums there?", "can I hear what she's saying" chases away the ability to exerience the deeper and more subtle qualities of experiencing music (like appreciation for dynamics, imaging specificity/width/depth, timbre etc). These are rocks upon which current double blind methodology is wrecked. I've experienced it myself, trying to listen critically while a friend switches between components is like trying to touch smoke - or judge depth in a 3 dimensional field and then close one eye. The act of attempting to listen for differences banishes the depth of perception necessary to experience those differences.
Ultimately, exactly what we should be testing for in amplifier performance is open to question.
The aforementioned article by Mr. Cheever on feedback, used extensively to improve THD specs in many (especially lower priced mass market) amps "convincingly illustrates that even minimal amounts of negative feedback covert moderate amounts of low-order harmonic and intermodulation distortion into a multitude of high-order distortion products". These distortions are shown to top 10% at levels of feedback commonly used, but in Dynamic Intermodular Distortion products that are not conventionally listed in spec sheets. Factors like these need to drive future double blind tests.
Simply counting which double blind tests resulted in null results, and which ones resulted in subjects successfully being able to discern between amps (see
http://www.pcavtech.com/abx/abx_pwr.htm for some examples of both) doesn't help you if a particular test is comparing apples to apples (i.e. comparing between two essentially similar amps - like two amps of similar design and construction with comparable performance). To back up an assertion like "differences among amplifiers are inaudible" requires that amps with different design parameters be tested head to head. I haven't seen results for this kind of test.
Science hasn't proven the inaudibility of differences between amplifiers. In fact, it has hardly been brought to bear on the issue.
