Ultrasonic Hearing in Music Recording & Reproduction

[kent walker <kent.walker@xxxxxxxxxxxxxx>]

Quoting AUDITORY automatic digest system <LISTSERV@xxxxxxxxxxxxxxx>:
>
> Date:    Thu, 24 Nov 2005 07:50:56 +0100
> From:    =?iso-8859-1?Q?Lars_Bramsl=F8w?= <lab@xxxxxxxxx>
> Subject: Re: ultrasonic hearing via bone conduction
>
> Hi Rob,
>
> I think the audio equipment manufacturers are trying to sell us
> something we don't need with 96 and 192 kHz sampling frequency.  When it comes
> to air-conducted sound I haven't seen a blind study where subjects could
> discriminate a 20 kHz and a 40 kHz band limited signal.
>
Dear Lars,

At the risk of repeating myself, there are possible advantages to high sample
rates which have nothing to high frequency perception.  One such advantage is
less pass-band ripple in the range of human hearing (although it should be
noted that using higher sample-rates are not the only solution to this problem).

Differences in higher sample rate recordings (DSD, 192 kHz, 96 kHz) when
compared to their lower resolution counter-parts can be quite striking.  If you
came to the studio at the University it would be realtively easy to demonstrate
that these media sound very different (even if you had a hearing aid). But are
such differences due to the higher sample rates themselves, filtering
differences, or something else?  Indeed, it is hard to construct scientific
evaluations of high-resolution audio in such a way that the results have
ecological validity to the real world of making records; there are too many
variables to hold constant.

In order for you to state that "audio equipment manufacturers are trying to sell
us something we don't need", you need to be able to produce studies that show
that there are no advantages to high sample-rate music recording.  Can you? 
Studies of hearing deficiencies may not have ANY validity to sound recording
whatsoever.

As I said before, there is lots of BS out there, but there is also some truth.

Regarding studies of high-frequency perception, one study (though not
peer-reviewed) that has reported that subjects might POSSIBLY be able to
perceive such frequencies is:

Ando, Akio; Hamasaki, Kimio; Nisiguchi, Toshiyuki; Ono, Kazuho, "Perceptual
Discrimination of Very High Frequency Components in Musical Sound Recorded with
a Newly Developed Wide Frequency Range Microphone", 2004 Audio Engineering
Society Conference preprint 6298.

Abstract:
Subjective evaluation tests on perceptual discrimination between musical sounds
with and without very high frequency (above 20 kHz) components have been
conducted. To make a precise evaluation, the test system is designed to exclude
any influence from very high frequency components in the audible frequency
range. Moreover, various sound stimuli are originally recorded by a newly
developed very wide frequency range microphone, in order to contain enough
components in very high frequency range. Tests showed that some subjects might
be able to discriminate between musical sounds with and without very high
frequency components. This paper describes these subjective evaluations, and
discusses the possibility of such discrimination as well as the high resolution
audio recording of music.
>
> Date:    Thu, 24 Nov 2005 15:17:27 -0500
> From:    Pawel Kusmierek <p.kusmierek@xxxxxxxxxxxxx>
> Subject: Re: Ultrasonic Hearing in Music Recording & Reproduction

> Digital equipment which records at 24/96 is easily available, and some can
> use 192kHz. Digital processing is usually done in a computer, which does
> not care about sampling frequency used. I think that it's harder to get a
> proper mic rather than equipment.
>
Dear Pawel,

While high sample rates are becoming "ubiquitous"  a very large number of
professional studios are still using recording equipment with sample rates only
as high as 48 kHz.  There are larger format digital consoles working at 2 x FS
and more, however, to use these sample rates means doubling the processing,
compromising other features such as number of channels, etc.  So higher sample
rates are often left aside because they can be more difficult to work with.  It
is extremely expensive to upgrade all of this equipment and the past 10 years
have not been kind to the pocket-books of most studios; they are just trying to
survive.  Many (most?) pop records are still recorded using lower sample rates.
Where high resolution has really "caught-on" is in Classical & Acoustic music
recording, and as we have both pointed out, the large majority of the mics used
to record this music "do not go up there".  Creating such a mic often means
sacrificing other features of a microphone which sound engineers and producers
have become acustomed to.

> Yes, but SACD and DVD-A, which slowly gain popularity, are capable of
> going beyond 20 kHz, and, as said by Rob Maher, speaker manufacturers
> continue introducing products with frequency range up to 40-50 kHz.

There are a number of SACD and DVD-A titles on the market which were recorded
using 44.1 kHz and 48 kHz sample rates and have been "up-sampled".  Many
high-profile pop recordings were/are recorded, mixed, and processed using
analogue gear, allowing for a "true" high-sample rate disc to enter the
customers' hands.  It does seem unlikely that there is much ultrasonic content
on most of these discs - although I won't say for certain until I can produce
some data :-)

regards,
-- 
Kent Walker
Ph.D. Candidate
McGill University
Schulich School of Music
Sound Recording Program
Instructor MUSR 300D