I'm not going to argue whether or not the "long-term" speaker break-in is a result of physical changes in the drivers -- simply because I don't have meaningful data or reference to add. But I would like to present some random thoughts on what people have said in this thread so far.
|I'm not an entire break-in believer, but here is my understanding of why it is hard to be evaluated scientifically:
1. Since it is related to material property change, the effects vary largely from speaker to speaker, which makes a publication to be unsubstantial.
2. Since the changes, if there is any, are mostly related to sound characters and maybe a little distortion wise, they may not be measurable. Judgements by ears are normally not qualified for a technical paper.
I have to disagree with both of these points.
1. If the inter-specimen variability is large, all you need is a larger number of specimens to increase the statistical power of the study. A large variability surely makes the study more involved and costly, but it does not automatically render the study unsubstantial or irreproducible. Ill-controlled experiments and statistics are what make the study unreliable. On top of these general points, I would say (as Bigus would probably do), a large inter-unit variability usually suggests a poor engineering.
2. First, you shouldn't underestimate the power of today's loudspeaker measurement methodologies. The old simple era of steady-state frequency responses and distortion measurements were long gone. Second, the "subjective" assessments from well-controlled listening tests are perfectly acceptable "hard data" for the science of loudspeaker engineering. What separates between science and pseudo-science is not whether the data are derived subjectively by human listeners, but whether the listening session is properly designed and controlled.
|However, I'm just surprised that, as a hotly debated topic for a number of years, some magazine/publication/etc hasn't taken a shot at it.
I believe that there are obvious reasons for this one: for one, there is absolutely no economical/financial incentive for the layman's trade presses (including all "audiophile" magazines) to conduct such studies. You know why.
In my own view, essentially the only place where a systematic study on this subject can ever take place is in the academia. I would say, it would make a very nice Ph.D. thesis if correctly done, whether the results are positive or negative. Perhaps, there are numerous such theses already around; just that we have no way of finding them.
|I am a firm believer in what my ears tell me.
That is a perfectly legitimate and even enviable position as a consumer.
However, today's loudspeaker design engineers cannot and should not survive with a my-ear-is-everything attitude, simply because it has not and will not progress the art of loudspeaker design. If not for the marvelous breakthroughs in both the materials science and the audio measurement technologies in the last three decades, we would have been stuck with the sound quality of the '70s and earlier.
|You see, what I have discovered through this wonderful forum, is that those who believe in lab testing will disregard the double-blind tests and ask for lab data. Those who belief in "the human factor" will disregard any lab data and cite the double-blind tests. This will again prove that, when it comes to audio equipment, we like what we like and believe what supports it...much like everything in life.
My own observations in this and other audio forums are in fact totally opposite of what you said.
I observe that, in general, people who believe in lab testing and physical measurements are also firm proponents of well-controlled (blinding is one of the absolute prerequisites here) subjective listening tests. These are the people who tend to have a strong desire to know exactly why it sounds different. People who believe in "the human factor" (whatever that means) and claim that measurements do not reveal much about sound quality are usually the ones who often belittle the value of controlled (blinded) listening tests. They tend not to delve into questioning the exact reasons behind the sound difference. If it sounds different, that is it for them.
|It just bothers me when people rely so much on the cop-out of "science hasn't proven it so it doesn't exist" when people experience it in everyday life.
Science very rarely can prove that "Phenomenon A" does NOT exist. Usually, all science can tell is that we do not have to assume that Phenomenon A exists. Whether you still "believe" that Phenomenon A exists or not is the matter of faith, not science. When somebody says "science hasn't proven it so it doesn't exist," he is in fact proclaiming his own faith. No more, no less. Does it still bother you? :D