Is human hearing really sensitive to it? Specifically in the range? I don't know. Can You cite any scientific research?
are you serious? there is literally hundreds of citations like this:
"Within this range, the human ear is most sensitive between 2 and 5 kHz, largely due to the resonance of the ear canal and the transfer function of the ossicles of the middle ear."
so the data shows that humans can hear quieter sounds @ 3kHz.
What does that have to do with phase overlap?
What does that have to do with phase overlap?
nothing or everything...if you believe we are sensitive to phase shift, than placing the crossover to the region where human hearing is most sensitive is not a good idea...I have the feeling we are going in circles here...
Yes: Dynamically changing waveforms are a different matter. In particular, it’s not only reasonable, but easy to demonstrate (at least under artificially produced conditions) that musical transients (pluck, ding, tap) can be severely damaged by phase shift. Many frequencies of short duration combine to produce a transient, and phase shift smears their time relationship, turning a “tock!” into a “thwock!”.
Because music is a dynamic waveform, the answer has to be “yes”—phase shift can indeed affect the sound. The second part is “how much?” Certainly, that is a tougher question. It depends on the degree or phase error, the area of the spectrum it occupies, and the music itself. Clearly we can tolerate phase shift to a degree. All forms of analog equalization—such as on mixing consoles—impart significant phase shift. It’s probably wise, though, to minimize phase shift where we can.
Yes: Dynamically changing waveforms are a different matter. In particular, it’s not only reasonable, but easy to demonstrate (at least under artificially produced conditions) that musical transients (pluck, ding, tap) can be severely damaged by phase shift. Many frequencies of short duration combine to produce a transient, and phase shift smears their time relationship, turning a “tock!” into a “thwock!”.
Because music is a dynamic waveform, the answer has to be “yes”—phase shift can indeed affect the sound. The second part is “how much?” Certainly, that is a tougher question. It depends on the degree or phase error, the area of the spectrum it occupies, and the music itself. Clearly we can tolerate phase shift to a degree. All forms of analog equalization—such as on mixing consoles—impart significant phase shift. It’s probably wise, though, to minimize phase shift where we can.
Oh yes, but the question discussed here is different - it is whether human hearing is:
[bolds mine]
is it?
that's the question
if You believe
Ok, question is why should anybody believe?
Is this belief substantiated with any scientific evidence?
I don't think the problem here is the global phase shift, but rather the audibility of a bad summation between not coincident drivers off axis (and even horizontally if polars are not properly matched and/or slopes are too steep).
Such bad summation will cause lobes and variable comb filtering as you move around.
Such bad summation will cause lobes and variable comb filtering as you move around.
There is substantial data to show that excess group delay is most audible in the critical 1 - 6 kHz region. I myself have done some of tis work and my papers reference the works of Moore, etc. that substantiate the claims.
Can phase mismatch cause elevated odd harmonic distortion, specially 3rd? I have seen this many times, either caused by nonsymmetrical slopes or delay mismatch. I use minidsp for my 4-ways and REW for measuring. It is easy to toggle between response and distortion, and change delay between measurements. I first check inverted phase and then fiddle with distortion graph. Digitally it is easy to be one (or even more) cycle over optimal with tweeter xo frequency.
The problem lies with 3-dimensional sound production - one plane and direction-optimization might lead you to the swamp.
The problem lies with 3-dimensional sound production - one plane and direction-optimization might lead you to the swamp.
I'm with adason on this. Vocals start to sound unintelligible with drivers that aren't optimally phase aligned. The greater they are misaligned, the worse it gets. Intelligibility gets better as the alignment stays on track well through the xo point. What that means to me is that as each driver hands off/receives the frequency duties to the next driver they stay aligned till the other is way below 24db in amplitude in its xo slope. When that happens even a 4-way design can sound like a point source @ 1m out.
"unintelligible" - that's a bit extreme don't you think? Especially when it isn't even clear what amount of phase distortion is even audible.
Sorry Graaf......I misspoke in this regard. The research is there to suggest what I've stated but isn't definitive. I could add subjectively speaking of course that I am sensitive to it. I have yet to build/voice or listen to a system who's crossover is in the mentioned range that sounds as clear as a mid covering 400hz to 4.5khz. I've tried a few 3-3.5" mid/Fullrange drivers and have yet to be disappointed. Your experiences may of course be different.
If you'd like to give it a try, the Vifa TC9 is a very inexpensive driver to experiment with. The Vifa NE series and Scan Disco 10Fs are better.....but more expensive options.
True, not in phase is less intelligible.
I call it true time alignment, 6db slopes with drivers spaced for equal arival times of the drivers, similar to an older Thiel.
Playing with a 24db crossover, a 5khz crossover point 2 way (6.5" crossed to a piezo) was as intelligible as a full range driver but with super clean detail, that suprised me. Meanwhile a 2khz 24db crossover was less intelligible than a 2khz crossed, a full range driver, or 2 different older thiels I had. I think a 24db is close to back into alignment (or at least +/- 90 degrees) 2 octaves on either side of the crossover point. Better than 12db or 18, but not as good as a 6db.
I'm not totally solid on this stuff yet, my opinion is still forming.
Norman
I call it true time alignment, 6db slopes with drivers spaced for equal arival times of the drivers, similar to an older Thiel.
Playing with a 24db crossover, a 5khz crossover point 2 way (6.5" crossed to a piezo) was as intelligible as a full range driver but with super clean detail, that suprised me. Meanwhile a 2khz 24db crossover was less intelligible than a 2khz crossed, a full range driver, or 2 different older thiels I had. I think a 24db is close to back into alignment (or at least +/- 90 degrees) 2 octaves on either side of the crossover point. Better than 12db or 18, but not as good as a 6db.
I'm not totally solid on this stuff yet, my opinion is still forming.
Norman
Last edited:
To clarify a degree of intelligibility would be difficult in the medium we're conversing in right here. Unless I was to gradually decrease the font size to an infinitesimal point ... one in which you'd be hard pressed to make out what exactly I've written especially if a slight blur was introduced as well.
This is the difference I hear in the nuances of vocals and accompaniment in a musical piece. Some might call this smearing. It's subtle and perhaps not noticeable until a direct comparison is made. One system .. even a well done system ... may present a particular piece of music that sounds more like "noise" in the back round when if fact there is a lot of coherent things going on. Some of radioheads' work is a good example.
The thread is about a range of frequencies that the ear/brain is most sensitive to. Why would being able to hear an even subtle misaligned phase between two or more drivers in this range surprise you?
When I started AINOgradient Project (4-way dipole with dsp) I found that it had a very airy soundstage, and I was happy! When I gradually got the crossovers acoustically symmetric (24dB/oct) this airiness mostly diappeared!
Earl said that time misalignment does not elevate harmonic distortion. However I see it changing very much near xo's when I adjust delay for one driver. Generally, the deeper null, the less distortion, 2nd and 3rd specially. It is also easy to have delay set one or more cycles offset between M and T. This will result in spl response dip about half an octave off the xo. Sorry but I don't have time to make a series of measurements to demonstrate this now.
Earl said that time misalignment does not elevate harmonic distortion. However I see it changing very much near xo's when I adjust delay for one driver. Generally, the deeper null, the less distortion, 2nd and 3rd specially. It is also easy to have delay set one or more cycles offset between M and T. This will result in spl response dip about half an octave off the xo. Sorry but I don't have time to make a series of measurements to demonstrate this now.
Pretty soon we will all being hear signal degradations because of the phase of the moon. Of course a full moon will elevate the "intelligibility" while a waning moon makes the sound "dark".
That only affects polar response.
Full moon is omnidirectional, while waning/waxing moon, being a waveguide (think about those lunar HOMs !) gradually makes sound directional. Some people even claim to distinctly hear new moon sounds...
Back OT: while I fail to see how non-coincident phase could really increase 2nd/3rd harmonics, I could imagine it might, under specific circumstances, make them more audible/detectable by summing them in an unfavorable way (their relative phase might end up different that the relative phase of their fundamentals and they might even cancel out if one gets lucky)
However, that's only a speculation. I don't know of any evidence of such a phenomenon and I fail to see why exactly conincident phase should yield a harmonic distortion cancellation.
ok, everything is most audible in the range of the human hearing's highest sensitivity, that's logical, still:
that's my point
- Status
- Not open for further replies.