"liquidator" <mikeh@[EMAIL PROTECTED]
> wrote in message
news:fuvevg$lut$1@[EMAIL PROTECTED]
> OK- may have misunderstood your point- so can I start
> over?
> The assertion is ears can hear 1 dB.
Under certain cir***stances that are not uncommon, yes.
> Ignore everything else- I still say you cannot tell the
> diff between 139 and 140 dB. I do not recommend trying
> the test.
In general, the ear is most sensitive to differences at SPLs around 85
dB.
Pidking potentially harmful SPLs like 140 dB is an example of an
excluded-middle argument.
> Once you are well into pain, can you tell 1 dB
> differences? I say Hell no.
Here's a standard reference that was published in the JAES:
http://www.aes.org/e-lib/browse.cfm?elib=11982
You can find just the frequency/level match graph at:
http://www.provide.net/~djcarlst/abx_crit.gif
This graph shows that our ability to hear level mismatch depends on the
frequency range over which the mismatch exists. If you have a mismatch
over
a 3 octave range, and those 3 octaves are within the range from 200 to
8000
Hz, and the mismatch is in excess of about 0.3 dB, then you might hear a
difference.
> Can you tell 1 dB differences at conversational levels in
> the human voice range. Almost certainly unless you
> performed the test above that I didn't recommend.
At 85 dB SPL, then you should probably hear a difference.
> Ok, then can you tell 1 dB difference at 20 Hz?
If you study
http://www.provide.net/~djcarlst/abx_crit.gif
you will find that the answer is that you are unlikely to hear a
difference.
> So all test conditions have to be qualified...who is your
> listening sample, frequency range, volume range...and on
> and on.
Agreed.
> Otherwise blanket statements are meaningless.
A common sense approach is to ask yourself, "Do I hear a difference when I
move a fader enough to cause a 1 dB change, according to the calibrations
on
the fader?. I believe I that I often do.
|
