We found micro placement beneficial :http://www.essex.ac.uk/csee/research/audio_lab/malcolmspubdocs/C78 Loudspeaker placement.pdf
Actually, I slipped a decimal place. 3mm is the right number. At that, if you can position your speakers and your head to within 3mm distance to each speaker, you're talking Clockwork Orange headgear to maintain that at the listening position. All moot, of course- keeping interchannel delay from the electrical source to near-zero is trivially easy.
I would love to reference any paper that does exactly that... Unfortunately, look at the typical stuff being done..(Joachim linked to some )btw, ""Can't buy me love"" by the King singers??? For a second there, I was thinking Beatles..
Note it's all about PAN pot and pan law...sheesh..If you're going to all that trouble with the reference speaker and physical movement, the least that could have been done was calculate the ITD as well for the SUT..sigh.
So lets see...by your math, if you are able to move your head to within 6 inches of the sweetspot, and reliably repeat that experiment, then you are able to discern 10 uSec interchannel without reference?? Sy, ya hearin dat??
Thanks for the link. As I pointed out, there was no ITD introduced, merely IID panning..nice setup, but not complete.
Cheers, jn
Hi,
Nope, it ain't.
Surely you know how delay and level differences correlate (e.g. how much delay produces the same subjective shift of a sound source as a given level difference), I mean I learend this s..t over two decades ago in the evening courses for sound engineers in east germany...
Ciao T
Nope, it ain't.
Surely you know how delay and level differences correlate (e.g. how much delay produces the same subjective shift of a sound source as a given level difference), I mean I learend this s..t over two decades ago in the evening courses for sound engineers in east germany...
Ciao T
At what frequency? You do know that the correlation breaks when frequency is introduced, no?
I learned that three decades ago, in the late night courses for sound engineers in east oshkosh...
Seriously, Greisinger produced data showing how the pan law diverges based on frequency. Not sure which paper it was, I recently changed computers so the directories are kinda messed up.. I can find it if you wish...
The gist is, if you use a pan pot to move a wide spectra voice, different harmonics of the voice will shift differentially causing a tight image center stage to begin to become wider as more pan is introduced.
Cheers, John
ps. Found it.
Greisinger, David "Stereo and surround panning in practice", AES 112th convention, Munich, Germany, 2002, May 10-13.
Textually, the concept is elaborated on in page 5, partial copied here for educational use only (in accordance with the fair use laws of the USA. Should the moderators belive this is inconsistent with copywrite law, please remove.)
Per the cited reference: note fig 6 is incorrectly footnoted as 5...
Note the high degree of frequency dependence of the curves.
At 300Hz the ITD follows the sine law precisely. At about
600Hz, the ITD is well predicted by the sine-cosine law.
Above 600Hz, the apparent image position moves rapidly
toward the left loudspeaker. This happens because the sound
diffracting around the head is interfering destructively with
the direct sound.
Last edited:
Digitizing Error
Trying to put the topic behind me, I guess the piece of information I am missing here is: how much dither is necessary to remove quantizing error as a function of bit depth.
I can't find a definite answer in any of my references, so if someone could shoot me a quick answer to this, I'll stop ragging on the subject
Howie
Howard Hoyt
CE - WXYC-FM
UNC Chapel Hill, NC
www.wxyc.org
1st on the internet
Trying to put the topic behind me, I guess the piece of information I am missing here is: how much dither is necessary to remove quantizing error as a function of bit depth.
I can't find a definite answer in any of my references, so if someone could shoot me a quick answer to this, I'll stop ragging on the subject
Howie
Howard Hoyt
CE - WXYC-FM
UNC Chapel Hill, NC
www.wxyc.org
1st on the internet
Hi,
The correlation does not break down when frequency is introduced, but it is arguably frequency dependent. I never claimed it was not, I merely enquired as to if you where aware of it...
Good on you. Maybe you shoulde use what you learned more frequently.
Please do. The stuff I am aware of goes back to WW2 era research in germany, so I'm happy to agree, it is wee bit out of date, I'm always keen to extend the the reach of my knowedge...
Unless we take care of frequency response differences so make the panning frequency dependent, like one may with a stereo pair of 1960's W88 Eckmiller faders...
Ciao T
The correlation does not break down when frequency is introduced, but it is arguably frequency dependent. I never claimed it was not, I merely enquired as to if you where aware of it...
I learned that three decades ago, in the late night courses for sound engineers in east oshkosh...
Good on you. Maybe you shoulde use what you learned more frequently.
Please do. The stuff I am aware of goes back to WW2 era research in germany, so I'm happy to agree, it is wee bit out of date, I'm always keen to extend the the reach of my knowedge...
Unless we take care of frequency response differences so make the panning frequency dependent, like one may with a stereo pair of 1960's W88 Eckmiller faders...
Ciao T
What, take it out for a spin so that it can wear out?? No way dude.
Oh, and by the way..if you had said three decades, I'd have said 4..
twas a joke..And it was from Greisinger's 2002 paper anyway, so it's actually less than a decade for me..
Did...cross posts..
Cheers, jn
Illegal values
Illegal values cause non-correctable aliasing. This, like the quantization caused by lack of dither, is a real distortion. Real modulation, real distortion products, measurable, verifiable by anybody.
If you find these distortion products preferable to "fuzzy distortion" (whatever that is intended to mean) that's your choice - it's a free country - but let's not sugar-coat it. It's distortion, heavily correlated to signal and strongly anti-monotonic.
Thanks,
Chris
"It's ffrr for me!
I've an opera here that you shan't escape
On miles and miles of recording tape." -F&S, about 1957
Illegal values cause non-correctable aliasing. This, like the quantization caused by lack of dither, is a real distortion. Real modulation, real distortion products, measurable, verifiable by anybody.
If you find these distortion products preferable to "fuzzy distortion" (whatever that is intended to mean) that's your choice - it's a free country - but let's not sugar-coat it. It's distortion, heavily correlated to signal and strongly anti-monotonic.
Thanks,
Chris
"It's ffrr for me!
I've an opera here that you shan't escape
On miles and miles of recording tape." -F&S, about 1957
SY,
To get an interchannel difference you need two channels. So if your loudspeakers are ten feet away and spaced ten feet apart, how far do you have to move so that the path length difference changes by 10 usec x 1130 ft/sec x 12 in/ft = .1356 inch (.344424 cm?)
D1 = sqrt(120 x 120 + 60 x 60) =D2initial = 134.16" D2dif = 134.16 - .1356 = 134.028
Sqrt (134.16 x 134.16 - 134.028 x 134.028) = 5.948"
ES
To get an interchannel difference you need two channels. So if your loudspeakers are ten feet away and spaced ten feet apart, how far do you have to move so that the path length difference changes by 10 usec x 1130 ft/sec x 12 in/ft = .1356 inch (.344424 cm?)
D1 = sqrt(120 x 120 + 60 x 60) =D2initial = 134.16" D2dif = 134.16 - .1356 = 134.028
Sqrt (134.16 x 134.16 - 134.028 x 134.028) = 5.948"
ES
Last edited:
Hi,
Well, if you don't give it some wear and tear, how do you know it is any good...?
Appreciated. I normally do not engage in Brinkmanship, unless I feel it will save me or the other side unnecessary damage (remember, apart from messing with girls and electronics I trained for around ten years of my childhood every weekend as what some may call a killer/terrorist/merc) and when I do I'm (literally) deathly serious.
You must forgive my German superiority complex, which is based on the simple fact that (excluding digital audio) much that is still debated or in fashion in these days (including stealth) was worked out there in the late 30's (there was little real progress in fundamentals but much in practicals after WW2 started) and it seems very little real progress has happened since. Without that stupid war (as well as the one that followed) the world would be at least 50 Years more advanced.
Well, spoilt milk.
Ciao T
Well, if you don't give it some wear and tear, how do you know it is any good...?
Oh, and by the way..if you had said three decades, I'd have said 4..
Appreciated. I normally do not engage in Brinkmanship, unless I feel it will save me or the other side unnecessary damage (remember, apart from messing with girls and electronics I trained for around ten years of my childhood every weekend as what some may call a killer/terrorist/merc) and when I do I'm (literally) deathly serious.
You must forgive my German superiority complex, which is based on the simple fact that (excluding digital audio) much that is still debated or in fashion in these days (including stealth) was worked out there in the late 30's (there was little real progress in fundamentals but much in practicals after WW2 started) and it seems very little real progress has happened since. Without that stupid war (as well as the one that followed) the world would be at least 50 Years more advanced.
Well, spoilt milk.
Ciao T
SY, that is where the 10u comes from, tests with a speaker. It was done by rotating the speaker around a listener to establish what angle of displacement could be discerned; from memory about 1 degree. From that, the sensitivity of the ear to time shifts was calculated. It is a binaural phenomena, that is, it describes the sensitivity to relative time differences between two ears.
I am sure I have a book here on the shelf somewhere that describes the experiment and its outcomes, I just don't know which one and my book collection does not have an automated search function. But I'll try to find it later on,
vac
- Status
- Not open for further replies.