Quote:
Originally Posted by
Chu Gai /forum/post/12071025
Let's look at the data and compare this to what Dick Pierce has stated in the two links that CharlesJ provided. Indeed, let us look at what's stated more carefully.
On GR Research's website, there is a technical section which can be found at,
http://www.gr-research.com/burnin.shtm
1) Intially, DR states that he made 3 measurements for "consistency and accuracy", yet the public does not have the measurements obtained so that an assessment can be made for his quoted comments. Had he done so, then one could assign a +/- to the averages using appropriate statistical weighting for the small number of trials. The numbers given are devoid of an understanding of what the "consistency and accuracy" are. Further, to report data to the ten thousandth's decimal place implies an accuracy that is completely unsupportable. As an example, an NBS certified 25.0000 gram weight is not cheap and DR's reporting that he added such mass is a stretch.
2) DR performed measurements on a fresh woofer. Straight out of the box. Never been played. DR considers that a baseline from which meaningful comparisons can be made. Of course this ignores the fact that drivers can be stored in different positions resulting in differing extents of being out of center. It also ignores that the manufacturing process itself can result in differing amounts of stiffness due to the way the epoxy, varnish or whatever is used. So, what does Pierce say about this?
Some manufacturers run their drivers through a process that excercises the driver. Some don't.
3) The
second measurements involve physically pushing the driver back and forth. No electricity was applied. Presumably this was done once in each direction by hand. IOW, it's just ONE CYCLE. It is entirely reasonable to assume that if this is the case, the changes in temperature were irrelevent. The measurements now show a change. As Pierce noted above, this is due to some initial cracking of the epoxy or varnish used. Not a lot of cracks, but enough to indicate that just one cycle causes an effect. The change in F(s) is approximately 2.7. Again, note that it is impossible to assign any standard deviations (66%, 95%, 99% confidence levels) to the numbers. We'll have to use them as are.
4) The
third test ran the woofer
"hard for 10 seconds". What hard is, is undefined. 50% full excursion? 75%? Something else? No idea. We could assume it meant very near full excursion but the assumption is unwarranted as it could also mean it was run slightly past full excursion in which case one may've damaged the woofer. By way of example, take a spring and let it oscillate with a light weight. Now use a heavy weight and watch the spring deform permanently.
In any event, the F(s) dropped 3.7 from the
original, never used, never tested woofer. After a period of time which we don't know, temperatures cooled. From what value to what value we don't know. Nor do we know how the temperatures were measured. In any event, the F(s) is now 2.7 from the original value.
5) The next measurements occured with
20 hours!! of hard run in. Again note, that no one knows what hard is. Well maybe your spouse knows a Mr. Softie
F(s) has changed by 5.0 from the original measurement of the never used woofer returning to a net 3.7 change after things have cooled down. The same comments about the temperature apply.
Danny Richie states,
"Now I wonder if the woofer burned in or if it was my test equipment?"
Well let's look at it shall we? In the AES paper I mentioned earlier, Clark found that it takes about a minute worth of work using a signal at the resonant frequency to effect break-in. Another manufacturer stated about 5 minutes. This 5 fold variation is still less than the time it takes to smoke a cigarette. As Danny Richie notes in his Technical section, there were no tests done at 1 or 5 or even 10 minutes. Instead, he took a T=0 seconds and T=10 seconds and then took a leap to
20 hours!!! Not even Marion Jones vitamin water can do that!
So, Danny presented data without...
a) The requisite number of data points at different times which could be curve fitted and equation developed. 10 seconds to 20 hours is just ridiculous.
b) No individual data points so that the averages reported could have appropriate error bars associated with them.
c) No specification for what driving hard means.
d) No temperature measurements.
e) No additional measurements after say 24 hours which would've been interesting as elastomeric creep takes more time than temperatures coming to some set value.
So, yes Danny Richie has presented data. What now follows is my personal opinion regarding this data. The data is incomplete and the experimental design is sloppy. Whatever statistics learned while pursuing a business program were not applied unless one considers adding three numbers and dividing by 3 to be statistics. Danny ignored or didn't think to investigate what happens to the T/S parameters after the driver has sat for a day which would have allowed him to investigate the elastomeric aspects in addition to temperature. He has not addressed the issue of driver to driver variablity within a lot of drivers and among different lots. Hence, he is unable to put into perspective for the reader the realities of manufacturing variabilities. Various unsupported claims were made elsewhere in the GR-Research link that appeal more to folk-lore rather than any verifiable scientific claims for audibility. In a very real sense, this is like the Nobel Laureate in Physics, Brian Josephson, attempting to use quantum mechanical explanations to justify things like mind reading and other paranormal events.
It is further my opinion that one benefits enormously by considering prior art and endeavoring to see what's been done in the field. Hence, that he does not have the AES paper nor apparently anything else pertaining to this field, has not boned up on matters of auditory science, is largely ignorant and dismissive of bias controls as it pertains to human psychology, other measurement techniques, etc. is indicative of why this paper and conclusions reached exist. Further work is obviously needed and since the ability to contact some of the people who don't see things quite the same way exists (Pierce, Clark, Nousaine), I might just start there.