janneman said:
Ah, very nice...it is unfortunate you are talking about JND for monophonic systems...do you have any understanding whatsoever with respect to source localization in humans?
janneman said:
If one completely disregards localization, then, you are correct..
You clearly have absolutely no understanding of what I said...my apologies...I guess I didn't keep it simple enough...
As long as you stick to your " 5% system" approach, with no regard to how we actually hear and need to measure, then of course, you will learn nothing..luckily, the researchers I discuss this with do not share your sentiments...although they are some of the best in the world with regard to human hearing, I guess they should listen to you...and stop with their foolish research...
It's unfortunate that there are some who blindly believe they know it all..rather than engage in reasonable discussion..
Cheers, John
PS..Perhaps Jan, your efforts would be better spent discussing at an intellectual level, rather than the trivial little snipes, such as "any other 'engineers' out there". I look forward to seeing if you actually understand what I'm talking about...your rather poor "showing" affords no indication of your 'expertise'..
SY said:
Hi SY
That, I most certainly will..
Does him being a smart guy mean he can say things like ""Any other 'engineers' out there?"", while dumb guys like me cannot respond to provocation?
As I said in response, I would prefer an intellectual discussion..petty little snipes are not that..
Cheers, John
jneutron said:
John,
Maybe I am the one who kept it too simple. Of course it is quite interesting to know how we hear, how we localise. No doubt about that. But my point is that that type of research is totally superfluous in the context of the influence of speaker cables.
I know that you know what you are talking about. So you are also aware that all that localisation etc can be completely thrown off course if you move your head an inch or so. And the normal variation in parameters of the system (speaker, xover, all that I mentioned) will introduce more uncertainties and variablities.
When the speaker cable influences are less than all those other variables, you can pull your conclusion, and save yourself lots of research that in itself is interesting, but has nothing to do in THIS thread.
As an example, you state somewhere that 1 uSec differential delay (or something similar) can be heard. Apart from the fact that a - you do not provide any backup for that statement, and b - I bet that is NOT tested with normal music, and that is what this is about; this is a golden opportunity for you. Just show us that speaker cables can give more than 1uSec differential delay and we will beat a path to your door to congratulate you!
Jan Didden
PS Sorry for that little snipe. I got carried away.
janneman said:
No, it is not useless..
It is rather easy to describe how cable LRC affects the signal that gets to the load, I'm sure you've done so.....but then what next??
.5 dB across 20 to 20K? 4 degree phase shift across the band, to JND levels?
Those measurements do not consider how we localize sounds..just how we hear them monophonically..
Current hearing research uses SAM, modulated SAM, dithered signals, even, lord knows why, rectified SAM. Some even combine ITD variations against IID references, and vice versa (sp), with various tones and combos.
But none provide any reasonable clue as to what we should measure for speaker wires...
I'm pushing the envelope w/r to human differential localization criteria...how the overall RLC will affect the system, that's easy...what's not is the cirteria we hear..
Cheers, John
jneutron said:
John,
Let me try to save you some time
. With the speed of sound some 300 m/Sec, moving your head an inch changes the differential localisation by some 840 uSec. Huge. If I would have a pair of speaker cable that has comparable DIFFERENCE between the individual cables of the same order of magnitude, I would agree that such a cable might have an influence on the localisation.I have not ever encountered such a cable, not even those 'engineered' cables with boxes with C's and L's on each end to make the sound more colored. So, unless such a cable turns up, I'l devote my energy elsewhere.
Jan Didden
Konnichiwa,
Is in case of RFI from the speaker cable the demodulation of an AM Radio station in the power amp section of a Marantz PM-66 with clear audibility during daytime (sounded like Kreplakish to me, not a UK station I think) using generic "speaker cable" sufficient?
Clip On ferrites largely cured this, as did later speaker cables (PC style ribbon, interleaved connection). As for amp's oscillating into various capacitive loads and blowing up as a result (and needing fixing), come on!
Sayonara
pinkmouse said:
Is in case of RFI from the speaker cable the demodulation of an AM Radio station in the power amp section of a Marantz PM-66 with clear audibility during daytime (sounded like Kreplakish to me, not a UK station I think) using generic "speaker cable" sufficient?
Clip On ferrites largely cured this, as did later speaker cables (PC style ribbon, interleaved connection). As for amp's oscillating into various capacitive loads and blowing up as a result (and needing fixing), come on!
Sayonara
Even assuming an absolutly terrible amplifier output impedance for the cable capacitacne to drain it's charge into of, say, one ohm, you would still need to have CABLES with a capacitance of around a micro Farad or so before the RC time constant of the cable will cause an audible effect (asuming a threshold of around 1uS as the limit for distinction).
So, even using the worst speaker cables and amplifer you could find, I doubt you would ever get anything this bad. So, it would seem a moot point about the sound staging caused by speaker cables in this sense when actual effects of stored energy are likely to many orders of magnitude lower than this..
jneutron,
The ITD and ILD analisys you made is very interesting, but just one fault. You've completely forgotten about an important part, the HRTF.
If you start to analyze how sensitive is the ear for very small directional changes, and you conclude numbers using geometrical analisys for the ITD and ILD in the us or even ns range, you are on the wrong track.
The ear is really sensitive for very small directional changes ( 0.5deg in the horizontal frontal plane) but only when mid-hf content is present. This huge accuracy comes from the HRTF and one eared direction finding mechanism of the ear above around 700 Hz, where the head shadowing effect starts to spoil the ITD and ILD. The small head movements together with the comb filtering of the pinna create differences in the mid-hf freq response for every direction. This is responsible for the accuracy of direction sensing mechanism, not the sensitivity for 1us or .01dB ITD or ILD differences.
BTW. from this point of wiew expecting hi-fi reproduction from a stereo speaker pair is laughable. Stereo relies heavily in ILD, hopes that ITD will be recreated at least in the LF at listening, and completely ignores the HRTF.
The ITD and ILD analisys you made is very interesting, but just one fault. You've completely forgotten about an important part, the HRTF.
If you start to analyze how sensitive is the ear for very small directional changes, and you conclude numbers using geometrical analisys for the ITD and ILD in the us or even ns range, you are on the wrong track.
The ear is really sensitive for very small directional changes ( 0.5deg in the horizontal frontal plane) but only when mid-hf content is present. This huge accuracy comes from the HRTF and one eared direction finding mechanism of the ear above around 700 Hz, where the head shadowing effect starts to spoil the ITD and ILD. The small head movements together with the comb filtering of the pinna create differences in the mid-hf freq response for every direction. This is responsible for the accuracy of direction sensing mechanism, not the sensitivity for 1us or .01dB ITD or ILD differences.
BTW. from this point of wiew expecting hi-fi reproduction from a stereo speaker pair is laughable. Stereo relies heavily in ILD, hopes that ITD will be recreated at least in the LF at listening, and completely ignores the HRTF.
Konnichiwa,
BTW, this thread provides an excellent possibility to demonstrate something. Please bear with me, but I have a important question for the thread starter:
The key here is, would you kindly elaborate on how exactly "High resistance thin cable killed the bass". I want to avoid assuming what you meant, so please respond detainling how the thin the Cable "killed the bass". Do you mean there was less bass (which is what I would consider correct for the term "kill")?
Finally, Twin Mains Ring & Earth is actually quite audiophile in terms of cables (Solid Core Oxygen Free Copper Speaker Cable is what you have, if you do not call mains ring twin & earth), though you get more milage from Cat5 Network cable.
Sayonara
BTW, this thread provides an excellent possibility to demonstrate something. Please bear with me, but I have a important question for the thread starter:
MikeG said:The measurements matched the listening tests. High resistance thin cable killed the bass, but QED and 15A ring mains Twin and Earth sounded the same. I used latter, £10 total from Wickes.
The key here is, would you kindly elaborate on how exactly "High resistance thin cable killed the bass". I want to avoid assuming what you meant, so please respond detainling how the thin the Cable "killed the bass". Do you mean there was less bass (which is what I would consider correct for the term "kill")?
Finally, Twin Mains Ring & Earth is actually quite audiophile in terms of cables (Solid Core Oxygen Free Copper Speaker Cable is what you have, if you do not call mains ring twin & earth), though you get more milage from Cat5 Network cable.
Sayonara
First order of business...the most important items..
If you re-read my posts, you will see that I clearly (well, to me at least
) stated up front....It is not possible to consider the problem in terms of absolute localization, as we do not hold our head in an accurate, stable position...that being the case, this problem has to be approached from a differential perspective...not where one source is w/r to the listener, but rather, where two sources are, relative to one another...
Even with one source, female vocal, does the spectra remain in one fixed position?...a good example would be sibilance and an eq..I have listened to a system that was incapable of centering the fundamental tones and the sibilence at the same time, her voice was center stage, and the s'ss were about a foot to the right...try as I could, those darn eq knobs were incapable of presenting a single presence..the sibilance was disembodied...the system shifted either ITD or IID, or both, of the higher frequency content to the extent that the vocal could not be focussed with any knob available..
In light of the fact that I am talking about differential localization, perhaps that needs to be re-considered..you quashed an entity I was not talking about...the old head in a vice thingy..(wish there was a head in the vice smiley...)
The statements 1 uSec and .01dB were examples of setting criteria to measure...I could have provided any numbers as examples of criteria...I am suprised you focussed on 1 uSec instead of .01 dB...I think the dB level is more difficult and outrageous..
That said, I've attached a graph of actual date, this simply because I already scanned it and converted to jpeg...the other data I have is in hardcopy form, the scanner is at home..
As is evident, differences at the 1.5 uSec level were reliably detected..this graph being simple tones..
Actual music will have far more information which we use to localize, so I certainly expect our sensitivity to be on par with what is on the graph...
Um, Jan..have you ever seen a test setup that is capable of measuring accurately 1 uS response stuff at 4 to 8 ohms?..at 20 Khz, even a 6 inch long piece of 500mcm generates enough magnetic field rate of change to produce 300 mV error. Nevermind making a darn load resistor capable of flat response to 106 hz..
My biggest gripe, is that we only hear to 20k (well, some can, I can't), but we localize to 1.5 to 5 uSec timeframes, which is entirely incongruous with the 20K bandwidth..
BTW, it'll be a virtual door..this is after all, just a hobby..
Cheers, John
I also apologize..janneman said:
It's rather easy to talk past one another..janneman said:
If you re-read my posts, you will see that I clearly (well, to me at least
) stated up front....It is not possible to consider the problem in terms of absolute localization, as we do not hold our head in an accurate, stable position...that being the case, this problem has to be approached from a differential perspective...not where one source is w/r to the listener, but rather, where two sources are, relative to one another...Even with one source, female vocal, does the spectra remain in one fixed position?...a good example would be sibilance and an eq..I have listened to a system that was incapable of centering the fundamental tones and the sibilence at the same time, her voice was center stage, and the s'ss were about a foot to the right...try as I could, those darn eq knobs were incapable of presenting a single presence..the sibilance was disembodied...the system shifted either ITD or IID, or both, of the higher frequency content to the extent that the vocal could not be focussed with any knob available..
janneman said:
In light of the fact that I am talking about differential localization, perhaps that needs to be re-considered..you quashed an entity I was not talking about...the old head in a vice thingy..(wish there was a head in the vice smiley...)
janneman said:
The statements 1 uSec and .01dB were examples of setting criteria to measure...I could have provided any numbers as examples of criteria...I am suprised you focussed on 1 uSec instead of .01 dB...I think the dB level is more difficult and outrageous..
That said, I've attached a graph of actual date, this simply because I already scanned it and converted to jpeg...the other data I have is in hardcopy form, the scanner is at home..
As is evident, differences at the 1.5 uSec level were reliably detected..this graph being simple tones..
Actual music will have far more information which we use to localize, so I certainly expect our sensitivity to be on par with what is on the graph...
janneman said:
Um, Jan..have you ever seen a test setup that is capable of measuring accurately 1 uS response stuff at 4 to 8 ohms?..at 20 Khz, even a 6 inch long piece of 500mcm generates enough magnetic field rate of change to produce 300 mV error. Nevermind making a darn load resistor capable of flat response to 106 hz..
My biggest gripe, is that we only hear to 20k (well, some can, I can't), but we localize to 1.5 to 5 uSec timeframes, which is entirely incongruous with the 20K bandwidth..
BTW, it'll be a virtual door..this is after all, just a hobby..
Cheers, John
Attachments
fcserei said:
The geometric analysis is only the beginning..it assumes nothing w/r to our frequency dependent or head geometry effects.
The differential localization most certainly has to be tested farther on down the line, and all your statements will apply..
ITD...in a recording? Don't they just use that darn "pan" pot during mixdown??
Cheers, John
janneman said:
John,
Let me try to save you some time
Jan Didden
How did you calculate that?..speed of sound is 13620 inches per second. (345 meters per second, consistent with your number..).invert that, I get 73.42 uSec/ inch. The distance between ears is 440 uSec..assuming 6 inches, of course.
Cheers, John
John,
I don't think I'll quote your lengthy post # 54. I know there is a lot interesting stuff in it. But precisely BECAUSE our heads are not in a vice, we have huge varying differential delays, all over the place, unpredictable, continuously. Milliseconds of it. Show me a cable that can dominate the localisation at those levels.
And I also do NOT agree that measurements with pure, jittered tones can be transplanted to music 1:1. Available measurements with THD levels, phase shifts, delays etc that I am aware of (but I may have missed a few), ALL show that you can hear these things, sometimes, with specially constructed signals, but NOT with music.
Jan Didden
I don't think I'll quote your lengthy post # 54. I know there is a lot interesting stuff in it. But precisely BECAUSE our heads are not in a vice, we have huge varying differential delays, all over the place, unpredictable, continuously. Milliseconds of it. Show me a cable that can dominate the localisation at those levels.
And I also do NOT agree that measurements with pure, jittered tones can be transplanted to music 1:1. Available measurements with THD levels, phase shifts, delays etc that I am aware of (but I may have missed a few), ALL show that you can hear these things, sometimes, with specially constructed signals, but NOT with music.
Jan Didden
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.