Lou, et.al.,
"Define what constitutes good sound"... isn't 'good sound' the sound of the real instrument and voice? That to me is the true reference. Science doesn't create a great musical instrument, it takes experience, intuition and good ears.
Exactly! - as far "good sound" being that of the sound of a real musical instrument having been reproduced faithfully. My comment regarding "what constitutes good sound" was in reference to science's ability to identify those physical parameters and effects of various audio devices (amplifiers, speakers, etc.), that must be in place and of sufficient fidelity necessary to reproduce the sound of a real, live instrument or voice - or any combination thereof.
If you try to say that science doesn't matter, it's the final sound that counts - you're only half correct. It's a matter of nothing more than the means to and end. I will prove my point.
With my "science" I will construct a speaker that has a flat response and in the end, let's say that because of whatever combination of variables it posseses, it just so happens that it is determined to deliver a very high level of performance by a majority of individuals - both experienced and inexperienced listeners. ( I should get rich$$$ - right?)
Now, I will take this wonderful speaker back in my lab and, using the same brands and quality of components, I will alter its frequency response such that my instruments tell me that I have produced a 30dB rise in its frequency response at 1kHz, that exhibits a "Q" of "1."
Now, how many of those original individuals that had determined this speaker provides excellent sound, will continue to hold that view? Wanna bet? I would venture to say virtually none. Through "science" I have managed to ruin a perfectly good speaker. I could accomplish similar effects by altering the enclosure construction such that it resonates terribly, or selecting different drivers that exhibit very high levels of distortion even at low drive levels. and be able to verify those effects via science and instrumentation.
So... isn't the opposite just as true? Can I not first identify parameters that seem to have correlation with perceived "good sound" and then optimize them to the best of my ability? Of course I can, this is the essence of all good product design and understood by any designer of quality products. Don't think so? Just look inside your CD/DVD player. There's enough "science" in there to make your head spin - and by God it better work or you won't get so much as a "peep" out of it.
The above dissertation is obviouse to all but the most willfully ignorant of the facts and should not even have to be pointed out. We start with the science and optimize it to the limit of our abilities - then we engage in subjective analysis in order to refine what our science cannot tell us. Nevertheless, the parameters that science cannot as yet identify are orders of magnitude below those that it can. If we optimize the science, then adjusting for the finer paramters that science can not help us with should only bring about further refinement of an otherwise already excellent product. Changing from copper to silver wire inside a "scientifically verifiable and subjectively perceived" lousy speaker, will not transform it in to an excellent one - period.
Whether or not the market perceives it to be a good thing or not, many designers "whistle past the graveyard" of science and then try dismiss its value publicly. Ironically, behind the scenes they still use it to the limit of their ability (and often very limited at that) to achieve a design that is marketable. To me, the whole notion smacks of hypocrisy.
I propose a challenge. In order for the scientific naysayers to validate their stand, I suggest that whatever designers that are out there that feel this way... try this:
Design a speaker (or amplifier for that matter) without using any formulas or instrumentation whatsoever. Design the thing on paper using only mechanical construction drawings and then build it. Then tweak and adjust it totally by ear until you have what you perceive to be an excellent product. If you ever get done within our life-times, then submit it to various members of AudioCircle for review. If the majority feel it is a superior product, I'll take it all back. I'll eat my words and admit I'm wrong.
Any takers? I didn't think so. Only a fool would consider this, let alone attempt it. Hey Frank, throw out all you RC time constant calculations and phase margin measurements and build us an amp, will ya? Oh yeah, just to make it a bit of a challange, I want it to be at least 500W/ch.

Never mind all the smoke and fire, that's just the price you have to pay in order to develop a truly supperior product.


Frank knows better - and so does any REAL engineer. Anything less is a hack - and I don't think Frank is looking for a "hack job" speaker. I rest my case.
At SP Technology, we optimize every scientifically identifiable parameter possible within the limits of the applied technology and cost limits of a given dsign. Frequency extension, frequency response linearity, frequency response matching between pairs of enclosures, THD, IMD, horizontal and vertical dispersion, phase error, diffraction, dynamic linearity, power handling, electrical impedance, enclosure resonance - you name it! All of these things are identified by science and are quantifiable with the proper instrumentation. Only after all of these areas are optimized, do we then subjectively evaluate internal wiring and crossover component quality effects - which are essentially un-measurable. Do these un-measurable variable have a sonic effect? You bet they do! And its funny, they seem to make a lot more difference once the scientifically measurable paramaters are optimized. Go figure.
-Bob