Oh, one other thing, I've noticed my impulse response is different at every angle. Is this normal?
Dan
Dan
Everyone has the right to disagree and that is fine. Then there is reality of published works.
A few phase and time related references concerning loudspeakers. This list could easily be expanded considerably but what would be the point?
"Audibility of time misalignment of acoustic signals" -M. Kunchur
"Temporal resolution by bandwidth restriction" -M. Kunchur
"The wood effect" -R.C. Johnson
"Crossover networks & phase response" S.K. Pramanik B&O
"An active crossover with phase correctors" R. Ballard
"Phase coherent crossover networks" Nelson Pass
"Passive crossover networks for non-coincident drivers" -S. Lipschitz
"Power response for non-coincident drivers" -J. Vanderkooy & S. Lipschitz
"In phase linearization of loudspeaker crossover networks possible with time offset and EQ" - J. Vanderkooy & S. Lipschitz
"Use of frequency overlap & EQ to produce high slope linear phase loudspeaker crossover networks" -S. Lipschitz & J. Vanderkooy
Bringing the importance of phase and time to the front again it is very clear, as I have been told by many, time and phase are poorly or not understood at all by most designers. Am trying to help that situation again. Keeping coherence high and signals in phase is the centerpiece of good sound and number one on my list of what is important to measure and design for.
Of course others may disagree but the thread is exactly about this subject of what is important to measure, why, and how. I have answered all three with lots of references and links to equipment.
To clarify, there are three domains: time; frequency; and energy. All need to be considered in a complete design. To limit testing to the single domain of frequency is to toss out the other domains. Two out of three not important? Maybe to someone else but not here.
Anybody else care to join in the fun on such a complete level? Please do as I am always ready to learn!
=SUM
A few phase and time related references concerning loudspeakers. This list could easily be expanded considerably but what would be the point?
"Audibility of time misalignment of acoustic signals" -M. Kunchur
"Temporal resolution by bandwidth restriction" -M. Kunchur
"The wood effect" -R.C. Johnson
"Crossover networks & phase response" S.K. Pramanik B&O
"An active crossover with phase correctors" R. Ballard
"Phase coherent crossover networks" Nelson Pass
"Passive crossover networks for non-coincident drivers" -S. Lipschitz
"Power response for non-coincident drivers" -J. Vanderkooy & S. Lipschitz
"In phase linearization of loudspeaker crossover networks possible with time offset and EQ" - J. Vanderkooy & S. Lipschitz
"Use of frequency overlap & EQ to produce high slope linear phase loudspeaker crossover networks" -S. Lipschitz & J. Vanderkooy
Bringing the importance of phase and time to the front again it is very clear, as I have been told by many, time and phase are poorly or not understood at all by most designers. Am trying to help that situation again. Keeping coherence high and signals in phase is the centerpiece of good sound and number one on my list of what is important to measure and design for.
Of course others may disagree but the thread is exactly about this subject of what is important to measure, why, and how. I have answered all three with lots of references and links to equipment.
To clarify, there are three domains: time; frequency; and energy. All need to be considered in a complete design. To limit testing to the single domain of frequency is to toss out the other domains. Two out of three not important? Maybe to someone else but not here.
Anybody else care to join in the fun on such a complete level? Please do as I am always ready to learn!
=SUM
I know Kunchur's work has nothing to do with music and doesn't demonstrate anything new that the link I gave you didn't say. Are there any in particular the show the audibility of phase in a room with music? I'll check out what the web will allow for free.
Thanks,
Dan
Thanks,
Dan
It doesn't seem the Wood effect is pertinent. Can you or anyone briefly explain how it might be?
The evidence I know of for this on music doesn't hold true when listening to music:
ABX Double Blind Tests: Polarity
Thanks,
Dan
PS. I'll keep looking unless they are all just on special test tones.
Thanks again
The evidence I know of for this on music doesn't hold true when listening to music:
ABX Double Blind Tests: Polarity
Thanks,
Dan
PS. I'll keep looking unless they are all just on special test tones.
Thanks again
Note that those tests do not prove that phase is not important in music, it only shows that that test was unable to show that it is audible.
dave
Gentlemen and Ladies- am not trying to "prove" anything or "convince to lobby" anyone to follow what I have found as most important to least important. The question was what, how and why. Have tried to be somewhat clear and precise which is why I bother with all those links and references. (even if you do not want any particular one) Please test and design as you see fit. As stated in an earlier post #38 there are very different goals for different applications. Certainly the person under the balcony has little use for a coherent speaker which is why there are planar bipolar radiators. My goal has always been to have the sound at the position of the conductor. To move toward that goal the items as listed are the ones found to be of importance here in my lab and in the custom installs done and by those who are trying to reproduce an acoustic signal which matches the electrical drive signal. If someone believes frequency response is the end all, more power to them. Different goals based in different beliefs does not make one person wrong and another right. Both may be correct for their goals. I have tried to be fairly clear on my specific goals and the results with some guidance to achieve those goals. For other goals other method and a different path leads to that different goal. If we all wanted the exact same things there would be no discussion as that would likely have been reached long ago. But we do not. PLEASE do not try and make this the "one size fits all" discussion conclusion. That is impossible as anyone can plainly see. Find what suits you and follow that guide.
=SUM😀
=SUM😀
Thanks for the correction there Dave. I should have said when listening to music through loudspeakers it has been demonstrated to be inaudible.
Bass is absolutely critical - according to Toole its a full 20% of the variance of preference. But bass is more of a room measurement and setup than it is a loudspeaker one. Bass and everything else need to be dealt with seperately and tend to have entirely different requirements. So my list did not include anything regarding bass - basically thats another topic altogether.
As to people disagreeing, its important to realize that there probably aren't two correct answers and its amazing how when people hear a really good system the opinions don't vary, they unanimously agree whats good and what is not. Its NOT a personal thing.
As to people disagreeing, its important to realize that there probably aren't two correct answers and its amazing how when people hear a really good system the opinions don't vary, they unanimously agree whats good and what is not. Its NOT a personal thing.
Last edited:
It has to be to some extent, the less so the better. Thats what Constant directivity is all about - you could say: Same impulse response in every direction, its the same thing.
You mean one size does fit all? Now you have really made me laugh! Therefore everyone given a choice between two speakers the same speaker will always be chosen? Really? Thanks for that.🙂
That is absolutely what I meant, make no mistake about it. Laugh all you want, just remember that I'm not joking.
Yes your goals are different, and I understand them. I was hoping that you had some evidence of audibility for your "why" on things other than test tones. I'm weighting my efforts in design on accurately reproducing those aspects which have been demonstrated to be audible during listening to music. I don't expect anyone else to have those goals no matter how reasonable they may seem to be for me. Others may think differently that's why I appreciate the fact that you are willing to state your ideas. Now power response is audible and those aspects make good sense to me but since I don't think anyone is using strict crossover typology in their design it seems that the final response measurements would take care of that. Same goes for lobing. Please don't think I don't find crossovers(filters) crucial.
Thanks again,
Dan
Anyone else have any input?
Interesting. Thanks! They do resemble each other. I wish I knew exactly what I was looking at when I see them. I just don't get the concept, but I also haven't done much work on doing so. I used to think it was essentially the braking power of the speaker. Is that correct on a basic level or way off? This speaker has a very good directivity pattern. Look at the impulse response. Can you, Dr. Geddes, or anyone who has a good comprehension of what's going on here explain what we are looking at in this impulse response courtesy of http://www.ai-audio.com/products_esp15.html:
An externally hosted image should be here but it was not working when we last tested it.
If this can't be posted here just let me know and it's removed immediately.
Oh and yes, bass is a room measurement, but if your loudspeaker can't reproduce the tone then there's not much you can do about it. I know you realize that.
Dan
Last edited:
Still not right.
"When listening to music through loudspeakers phase was inaudible in these tests."
ABX is not strong enuff to say anything more.
dave
I think I understand what you're saying, but not the connection to the "strength" of abx? You certainly can bungle any experiment. Try an experiment to see if dogs are invisible, but forget to open your eyes during the test. You'll report that they are indeed invisible. When the experiment is conducted by someone you respect or with a track record of performing his experiments carefully you should take more notice but it's still not proof. When the experiment has been repeated by many careful experimenters and they all get the same result, it's still not proof, but it's the closest thing we have in science. Every time you've ever looked in your basement it looked normal but that's not "proof" that ghosts don't come out and have a tea party when you're not down there. I'm not trying to equate the audibility of absolute phase with the existence of ghosts, I have no personal experience with either subject, I'm only making a comment about the scientific method and wondering if you're referring to abx as being flawed due to this or something else I don't know about.
For what it's worth Argument from ignorance - Wikipedia, the free encyclopedia
Last edited:
I've switched polarity on both speakers multiple times in my life. Never been audible for me. Try it for yourselves.
Last edited:
It was not so noticeable to me until I got the drivers to decay on an average 10db more than the original drivers in the first 0.4ms or so.
It is statistically unable to prove the result that there is no difference.
It can only be used to prove that things are different.
dave
Here'a another analogy to ponder: "If a tree falls in the forest and no one is there to witness the event, does it make a noise?".... It wasn't recorded, so there's no data to show it made a noise.. an argument from ignorance. 😀
- Status
- Not open for further replies.