|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
64.13.14.96
In Reply to: RE: I now understand the emphasis on amplitude measurements over time measurements posted by unclestu52 on July 31, 2007 at 20:48:42
Great Stu,
I thought that's what you meant, but wanted to be sure. I completely agree with you. I use two-way speakers (Celestion 3's) in the study and early on flipped the tweeter wiring so it would be in phase with the woofer. Sure it caused an on-axis notch at crossover but the result was well worth it. And I don't listen on-axis anyway.
My theory is that we are used to tuning out FR variations because they also occur in nature from things like foliage absorption. Timing problems on the other hand are less common, and much of the information we rely upon to survive in the wild, or on the street is encoded temporally. Yes, a flat FR looks good. And it IS good all else being equal, but it's only one view.
Something of interest to me currently is learning the frequency range where we are the most sensitive to timing (and hence harmonic alignment errors). I'm pretty sure that it doesn't extend all that high but may go lower than one would think. I've been toying with how to measure it but haven't hit upon an appealing scheme yet. Thoughts welcome. Heck, SWAG's at the result are welcome. For instance, I'm guessing that the upper critical frequency is ~8KHz.
Regards, Rick
Follow Ups:
of a driver in a system designed with it inverted actually creates a hump, rather than a notch. The inverted driver, being reversed to the polarity of the adjacent driver will generate cancellation where the frequencies overlap, and thus give the system a perceived steeper roll off at the crossover point.
I find the humps generally acceptable as it is not always that there is music in that overlap range. In your case, I would guess that the crossover is in the 2K to 3KHz range. Being that tuning A is about 440 Hz, you're in the fourth harmonic range for the most part.
As for hearing sensitivity range, calculate the distance between your ears (seriously !). That would translate to a wavelength of about 2.2 K Hz or thereabouts. Hearing sensitivity should be centered on about that frequency. Also consider the range of human voice: about middle C (~260 Hz) and up three octaves to about 2.1k Hz. Piano scale runs about 30 Hz to about 10k Hz. Because of the logarithmic nature of hearing, human perception is remarkably skewed to the bottom end of the scale, numerically.
Stu
Wow Stu, that may be THE answer.
I've read that hearing peaks around 3KHz but it never occurred to me that it might be tuned to match head thickness. I've also read that we can resolve interaural timing differences of 20uS, but didn't put the two together! I just bet you're right, that we are the most sensitive to phase and timing in that area also. Cool. It really makes sense, higher frequencies would be more difficult to analyze because the next wave coming along would cause interference, lower frequencies would provide worse timing resolution due to the slower slopes. And as you point out, instruments have plenty of harmonics in that area, probably because they were invented by humans!
In fact, your answer is such a good one that I'm just going to believe that it's correct and press onward. What a simple, elegant answer: The hearing range that matters the most for music is the voice range. I owe you a few, I like the brew and view at Whalers brewpub but it's a ways out of town for me...
Inverting the tweeter did cause a mild peak as I recall, but it's fairly benign. Think it sounded just a tad "brighter" than stock but it's been a long time. The notch is of course much more noticeable being deep and narrow. Of course the notch can approach infinity while the peak's limited to around 3dB. Since I listen off axis I don't notice the notch and overall find the sound is more satisfying.
Regards, Rick
d
s
nt
"silence tells me secretly, everything..."
a
Nope. Never even heard of it... I just read the plot summary on IMDB and it sounds wretched.
I find it a solid comfort being out of the popular culture. The minor downside is that even though I live in Springfield, I don't get most of the references on the Simpsons. Which actually makes perfect sense...
Rick
...I think it's stupid.
And so's Seinfeld!
clark
Clark, I don't have a clue of which you speak.
If I'm a being messed with, "they" are doing a great job because I don't know who's doing it, what they are doing, or in what manner I'm being influenced. Tell me more, perhaps my brain has been so skillfully laundered that I don't even recall the trip through the wringer...
I'm actually inclined to feel that I just gained an excellent insight from Stu with no effort on my own part. How can it get better than that? By golly, this forum is both fun AND informative!
As far as crossover's go, their compromises and trade-offs are hardly a secret or new news. I'm not a 'speaker guy' so I wasn't trying to optimize the design, just choose a different compromise. I believe that "speaker guys" are typically better informed and upfront about the tradeoffs of their designs than are "electronics guys". (And I'm one of the latter!)
Regards, Rick
If not, then they're not being terribly "upfront" about it are they?
As for their being "better informed", while they may (or may not) be aware of the havoc they wreck polarity-wise, is their being informed about it helpful to anyone?
clark
No, I think Stu nailed it: Frequency response is easier to relate to and sell. Perhaps due to the efforts of the FTC. However if the phase thing can be boiled down to something meaningful and possible then it could become a real marketing factor. And a factor in improving design as well.
I suspect that the key to it is identifying a subset of the audible bandwidth that provides the most meaningful polarity cues and specifying acceptable temporal performance over that bandwidth (the voice range?). Trying to do it full-range is insane unless you keep your head clamped or wear headphones. Perhaps nut-case audiophiles could help promulgate a standard. You would make an excellent point man to sell the concept.
I think audiophiles underestimate the importance "real" people attach to our opinions. Not that they would ever wish to be one, or associate with one for that matter, but still they want to have gear that is "audiophile quality" or "audiophile approved".
Rick
And it was all thoroughly covered in The Wood Effect (1988). I blame the magazines for neglecting this aspect; the old Audio used to do speaker reviews in which the phase distortion was measured *and published* in easy form, not the rather more confusing (albeit revealing) Melissa plots of today.
clark
Sorry, couldn't resist Clark...
Yes, I've read them. Heyser did a good job yet I don't recall them being particularly useful, maybe I'm just too superficial of a reader. I subscribed to Audio from ~1973 until they folded. Guess I mostly read magazines for interest and entertainment and largely ignore them when purchasing equipment.
Are you saying that you described a temporal metric that had good alignment with what listeners experienced? If so maybe I WILL have to spring for a copy.
Regards, Rick
Nothing however about temporal metrics -- saving that for a later day.
clark
The definitive statement on polarity written 20 years ago. It ignores simple impulse tests, it ignores many EIA standards in regards to microphones and headphones already in publication, it ignores the more recent AES 26 and other standards. It ignores new media such as laser discs, MP-3, and such, as one would expect from a relatively ancient and not well researched document. Now we are to await further details in an upcoming article, details of which can not be shared until publication. Clark's attempt to enlighten the general public shows his true colors: money before sharing of knowledge.
Such hubris
Stu
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: