On Mon, 28 Apr 2008 07:22:57 -0700, dpierce.cartchunk.org wrote:
> On Apr 28, 8:39 am, Edmund <nom...@[EMAIL PROTECTED]
> wrote:
>> On Sat, 19 Apr 2008 22:09:32 +0100, Eeyore wrote: Long time ago, not
>> with young people but older where tested.
>
> Who? When? How? Can you site a particular reference, or is this just
> memory of an anecdotal tale?
It is something I read 20 or even 30 years ago, don't ask
me for details :-)
>
>> They where tested and all where complete deaf over 14000 Hz. So they
>> played music and switched a filter on and off at 16000Hz, ALL of them
>> could tell the difference!
>
> There are ANY number of reasons why this could easily be the case,, none
> of which involve ANY requirement of bandwidth beyond 20 kHz. Arny
> mentioned the issue that the filkter itself can easily have an audible
> effect.
>
> Another case is where a standard laboratory function generator is used
> as the stimulus. The frequency was set to 7 kHz, and the output varied
> between a sine wave and a square wave. It's almost certain that in quick
> switch comparisons, MOST people will reliably hear a difference. But the
> reason has NOTHING to do with bandwidth.
>
> Most function generators have outputs calibrated such that the peak
> voltage of the waveforms is the same. That means that if you set the
> square wave output to 1 volt P-P, the sine wave output will also be 1
> volt P-P. Unfortunately, the 7 kHz component of the square wave is not
> 1volt P-P, but some 1.3 or so dB lower. The result is that the listeners
> end up hearing the difference betweem the different amplitude of the 7
> kHz sine wave in both, and NOT the presence or absence of the 21 kHz, 35
> kHz and so on components in the square wave.
>
> In other words, we have two examples of tests which have a fundamental
> flaw in the methodology, and do NOT, as you might think, illustrate the
> significance of ultrasonic bandwidth.
One detail I remember was the individual testing was done by
sine waves, which showed that non of the members heard anything
(at given level ) over 14000 Hz, most of them did a lot worse.
The filter test was done while playing music.
>
> Unitl you can site a test which accounts for these and other issues,
> including small in-band amplitude differences and more, the results are
> suspect, at best.
>
>> I myself have not yet done much listening tests between the SACD and
>> CD, however once I attended a session where SACD was played
>> and I made a copy of the CD layer of an SACD. I clued a new CD sticker
>> on both the SACD and the CD quality copy. We played both and everybody
>> could tell the difference. ( only I knew which was which )
>
> And can you prove conclusively that the CD layer and the SACD layer
> were, to begin with, exactky the same save for the bandwidth? For
> example, it is not infrequently the case where the CD layer and SACD
> layers can be made with two different sets of masters.
No I don't know about that, I assume they used the very same,
why would they use or even make different masters?
>
>> Some time ago :-) I have compared the CD with good vinyl, no doubt the
>> vinyl sounded better although the vinyl had the additional vinyl
>> problems.
>
> That, if you're at all serious, is an interesting comment,
> since, for the most part, the practical bandwidth of vinyl is less than
> that of CD, due to mechnical limitations in medium. And that's even more
> curious in light of the fact that most recordings have significant
> bandwidth limitations before they even hit the final stage.
If that is a fact and there are no exceptions
I wonder what kind of effects are at work then.
Edmund
|
