I was starting a debate on a well known hifi forum where I was about to
argue that the criticisms of digital sampling were flawed due to the
fact that human auditory perception is far from perfect (ask any dog)
and that even if a graph can show the difference between a smooth curve
and a 44,100 sampled curve that humans could not perceive it. Well we
cannot as far as I know perceive the difference between a 44.1KHz
sampled wave and a 48KHz sampled wave, can we ?

I then got to thinking what are the limits of human hearing ( I know
about frequrency ranges) but what is the lowest volume, shortest
duration, shortest time between 2 signals and so on that we can
actually perceive. The human ear is a mechanical system and there must
be a latency, i.e the ear cannot respond instantaneously ? Do we have
any psychophysics experts here ?

Cheers

Jim

wrote in message
oups.com

I was starting a debate on a well known hifi forum where
I was about to argue that the criticisms of digital
sampling were flawed due to the fact that human auditory
perception is far from perfect (ask any dog) and that
even if a graph can show the difference between a smooth
curve and a 44,100 sampled curve that humans could not
perceive it. Well we cannot as far as I know perceive the
difference between a 44.1KHz sampled wave and a 48KHz
sampled wave, can we ?

There have been many listening tests that have compared

An original analog signal with
The same signal converted to 16 bits 44,00 Hz and back to analog

No audible differences were found.

There have been many listening tests that have compared

An original analog signal converted to 24 bits 96,000 Hz and back to analog
The same signal converted to 16 bits 44,00 Hz and back to analog

No audible differences were found.

wrote in news:1148323711.912961.228090
@j33g2000cwa.googlegroups.com:

I was starting a debate on a well known hifi forum where I was about to
argue that the criticisms of digital sampling were flawed due to the
fact that human auditory perception is far from perfect (ask any dog)
and that even if a graph can show the difference between a smooth curve
and a 44,100 sampled curve that humans could not perceive it.

What kind of graph are you talking about here? Any differences in the
signals are miniscule, so any normal graph of the signal will show no
differences whatsoever.

Well we
cannot as far as I know perceive the difference between a 44.1KHz
sampled wave and a 48KHz sampled wave, can we ?

I certainly can't, and I don't know of anyone that can.

I then got to thinking what are the limits of human hearing
[snip]
Do we have any psychophysics experts here ?

Not me. It's a large field of study, but there are many resources
available on the net. Do a google search for psychoacoustics and enjoy the
next month or so reading various papers.

Murray Peterson wrote:
wrote in news:1148323711.912961.228090
@j33g2000cwa.googlegroups.com:

I was starting a debate on a well known hifi forum where I was about to
argue that the criticisms of digital sampling were flawed due to the
fact that human auditory perception is far from perfect (ask any dog)
and that even if a graph can show the difference between a smooth curve
and a 44,100 sampled curve that humans could not perceive it.

What kind of graph are you talking about here? Any differences in the
signals are miniscule, so any normal graph of the signal will show no
differences whatsoever.

Well we
cannot as far as I know perceive the difference between a 44.1KHz
sampled wave and a 48KHz sampled wave, can we ?

I certainly can't, and I don't know of anyone that can.

I then got to thinking what are the limits of human hearing
[snip]
Do we have any psychophysics experts here ?

Not me. It's a large field of study, but there are many resources
available on the net. Do a google search for psychoacoustics and enjoy the
next month or so reading various papers.

Wikipedia is a decent place to start:
http://en.wikipedia.org/wiki/Psychoacoustics

There are also probably fewer fields where there's such a big
difference between blind and non-blind tests. I'm sure
there are people that can convince themselves their
ears have 50 kHz bandwidth - if they know what they're
listening to.

I found one hilarious article by a guy arguing against audio
compression.
He claimed that double-blind tests weren't really important; it was
enough to *know* that some fidelity had been lost. Unfortunately,
I've lost the link. Can anyone help?

-jc