Oh, I don't know..... just a silly whim, I suppose.
It is what it is, it's not short hand for anything.
Bass guitar is IMO easily the *best* example with a wide variety of music that includes that instrument. If you don't like that example - then look elsewhere.
I would restate "we localize only on sound > 700 Hz" to "we localize principally on sound > 700 Hz." but otherwise all the statements are correct and do not indicate a complete disregard for < 700 Hz. Because it is the most important does not mean that is the only thing that is important. Can we now cease with the "disregard" claims?
We 2 . . .
Of course those of us who have loudspeakers that maintain pattern control through the fundamental music and voice range already knew . . .
I think the brain learns localization much faster than you can figure out what it is doing. From a design point of view, we just try to take an iterative approach to gradually minimize various design problems. Trying to figure out analytically what is more critical is fine as long as you follow up with quantifying and prioritizing design goals, otherwise, all the research goes to waste.
Just listened to some sweeps. I don't get any distinct localization. Localization fluctuates a lot with varying frequency.
I can hear the comb filter at certain frequencies. There's sound at one ear but not on the other.
My Geddes speakers are currently set up with a 56° toe-in.
You do get sharp localization of single sines? Are you sure it's not distortion artefacts?
"The problem" is far more complex than that, and begins way back with the (already failed) capture of the information.
Of course it's not. But if a speaker cannot meet that basic minimum (sound coming from the same apparent source, which it doesn't if some part of the frequency range includes reflections and another part doesn't) it will certainly "distort interchannel signal information".
It's not an issue in an anechoic or sufficiently large environment. But in a small room the effect of frequency dependent first reflections is to move the apparent source. That has been a cause for criticism of "tweeter bloom", and has been a cause for criticism of the "woofer bloom" ("baffle step") exhibited by almost all "box" loudspeakers. It sounds "wrong" at 1500 Hz. (ORION), and it sounds wrong at 700 Hz (amply demonstrated by Pano) and at 350 Hz. and even at 175 Hz., (at least in some cases . . . Pano thinks even lower, and I'm not inclined to disagree).
IMO the only "safe" argument is that for the entire frequency range over which we can "localize" a sound source our speakers must radiate uniformly . . . so that the sum of direct and reflected sound presents uniform directional information. If the speakers can't do that there is little chance that directional clues (stereo information) embedded in the recording will be presented in a uniform and convincing manner by them.
Indeed, in case of a phase shift, which virtually all speakers have, even FR, the phase of the harmonics will be different from the fundamental. Take the case of a 4th order xover at 2KHz playing a 1500 Hz tone. The first harmonic will be delayed compared to the fundamental and I guestimate that to be at least by 180 degrees. Obviously, when you change the fundamental, e.g. bring it up to 2Khz, the phase relationship with its harmonic will change also, because the phase shift between 1.5 and 3 KHz will be different from the phase shift between 2 and 4 KHz.
Don't judge my personality, but my arguments, please. I don't attribute your attention to scientific qualifications to your nationality.
Last edited:
That's what happens with diffraction and reflections . . . and especially when the reflections vary with frequency (diffraction always does). You might try speakers with less diffraction and better directivity control . . .
You might want to talk to Earl why his waveguide creates so much diffraction and has such poor directivity
Why would a uniformly radiating speaker perform better? The room distorts the information, not the speaker (hopefully).
I believe that kind of phase distortion is completely meaningless when it is the same in both speakers. The real problem is phase distortion happening in one transmission channel but not in the other. Such distortion is caused by the room and to a lesser extend by the speaker (if your speakers aren't well matched).
Yes, it's way in front of all these requirements being mentioned here if you give it decent data to work with. And that data doesn't be to be directional or anything else - just high quality ...
I thought that this goes without saying. Harmonics can be "out of phase" in the original sound already. But this phase relation is fixed and does not change that much on the run like it does in Panos experiment.
I try to understand the person behind the postings, because that can help me to understand his argumentation. But I certainly don't "judge" personalities.
Rudolf
I do not. I've stated this many time as the reason why I test with noise, speech and music. I have a very hard time localizing pure tones of any frequency.