> >
>
> While not disputing the phenomena cited by Ken, I'm still
> surprised that "ordinarily" the cut-off is as low as 8,000 Hz. Of
> course, it all depends on what is meant by "ordinarily". I'd
> suggest an operational definition would be a test while watching
> TV at normal volume. Would a subject notice a tone inserted
> during the programme at the same volume but above 8,000 Hz?
>
> Data, Ken?
>
Here is another operational definition of "everyday" and some data.
I have attached a wav file to this reply. It contains 2 versions of
the same segment of music from a selection by the David Grisman
Quintet. One is the original wav segment directly from the CD and the
other is the same segment after run through the MP3 encoding process.
Play this with a program that will show you the spectral frequencies.
Listen first and judge their similarity. Now watch the frequency
display to see how much was removed from the upper end.
The MP3 process is based on masking. See the figure on p. 182 of
Coren, Ward, and Enns. It shows the results by Zwicker (1958) on pure
tone masking. The important aspect to notice is that the effects of a
masker are asymmetrical, the effect is stronger on higher frequency
sensitivity.
How's that for data?
Ken
----------------------
Kenneth M. Steele, Ph.D. [EMAIL PROTECTED]
Dept. of Psychology
Appalachian State University
Boone, NC 28608
dawg_w128.wav
