[Audio Empire]

[ I fumbled the formatting on the last message here. The
corrected attribution appears on the next three lines. -- dsr ]

In Message-ID:
Edmund wrote:

On Wed, 20 Apr 2011 21:57:11 +0000, Audio Empire wrote:

On Wed, 20 Apr 2011 07:13:16 -0700, Arny Krueger wrote (in article
):
=20
"Edmund" wrote in message

Hi Guys,
=20
Since I now finally have a HD player I like to buy HD material. There
is a possibility to buy HD tracks on line. However when I open such a
file with "Audacity" I don't SEE any frequency above 20 kHz, but I
must admit I did not totally analyze this data I just looked at it
in the graphics of Audacity. But I still wonder if any of these track=
s
actually contains frequencies above 20 kHz. What do you guys make of
this?
=20
Many so-called hi definition recordings are actually 24/96 or 24/192
transcriptions of analog tapes and 16/44 or 16/48 masters. Analog tape=
s
were ususally made at 15 ips which pretty well eliminates response
above 25-30 KHz.
=20
Actually, it's difficult to maintain even a pro 15 ips analog tape
recorder to much above 15 kHz. Self erasure, poor head contact at
ultra-short wavelengths all make analog tape "iffy" at much above 15
khz. Back in analog days, most studios only maintained their tape decks
to 15Khz. More than that took too long and wasn't practical. Before I
recorded the SF Symphony back in the 1970's I would carefully align bot=
h
Otari MX5050s for head azimuth, bias, frequency response, and Dolby 'A'
level AFTER I set up the equipment (the tape recorder was not moved
after that). I never tried to get response beyond 15 Khz, but I
scrupulously maintained the recorders to that level.
=20
Any of you have a source for HD which has music information above
20kHz?
=20
Not really. Let's say that the recording was made at 24 or 32- bit and
192 KHz (or perhaps DSD). The recording equipment might have that kind
of bandwidth, but I'll guarantee that the microphones used don't. Most
condenser mikes fall-off like a rock above the resonant frequency of
their diaphragms - and that is usually 16 -18 Khz. Not much there above
those frequencies. Doesn't matter, though. You likely wouldn't be able
to hear it anyway.

Well we can assume anything, if technicians in the past would have assume=
d
we could not hear improvements, we would still recording music on a drum=20
made of wax.
=20
=20
As Jwvm correctly points out, nature abhors the creating of true hi de=
f
recordings. His list of reasons is good and relevant.
=20
Let me make clear that I don't want to start whether or not higher
frequencies are audible, only that HD music should have recorded
higher frequencies in the first place.
=20
Why would it? A chain is only as strong as its weakest length. In the
case of recording, as elsewhere in audio, the weakest length is always
the transducers - speakers, phono cartridges, microphones. In this case
it's microphones.
=20
I agree that it is very nice when so-called hi def tracks are actually
hi def! ;-)
=20
The hi-def doesn't refer to recordings filled with extraneous and
inaudible ultra-high-frequency information, it refers to the fact that
the bandwidth of the medium is wide enough that only gentle slope
filters, located high above the highest audible frequency, need to be
employed on either end of the the chain (recording and playback) to
satisfy Nyquist.=20

I do not think you have in fact a gentle slope filter if one use a mic
that filters like a brick wall.

I was talking about the recording process not the signal source. And mikes
don't filter at all, they roll-off after reaching their moving mass's
resonant peak like all transducers.

Many insist that this sounds better. Certainly,
recording information that only small dogs can hear is of no use to mos=
t
humans in and of itself. The advantage (if any) is in the process of
recording and playing back with a very high sampling rate, not the
information that such a sampling rate is capable of quantizing. Hint, i=
f
any part of high-resolution recording is beneficial, its the word lengt=
h
(24 or 32-bits) not the sampling rate.

Actually both word length and sample rate interact.

Actually, word length and sample rate do not "interact". Word length gives
you a specific dynamic range and the longer the word, the greater the
(theoretical) dynamic range. As far as I can see, somewhere around 130 dB is
about the limit of current technology. But with high dynamic range comes
increased recording headroom and this is high bit-rate's most important
contribution.

OTOH, sample rate defines the highest frequency that can be quantized.
Nyquist says that this frequency is half the sampling frequency. With 192
KHz, that's a Nyquist frequency of 96 KHz which, except for the advantage
that this frequency is so far out of the pass-band of human hearing that any
anti-ailaising filtering would have, essentially, no affect on the signal
being quantized, really is meaningless in any practical sense. Notice, I'm
not saying that such a high Nyquist frequency has no sonic advantages (though
some here will be quick to give that opinion), I'm just saying that any
"intelligence" that lies that far above audibility, is rather pointless in
and of itself.

Again I do NOT want to start a discussion whether or not HD audio is=20
audible, first I am looking for recordings that actually recorded=20
high frequencies.
Suggestions for material/recordings are appreciated.

Well, Edmund, it looks like you are pretty-much out of luck. I doubt
seriously, if any high-resolution recordings have anything on them much above
30 KHz, if that. And I'll guarantee you there's certainly nothing at all
above 50 KHz on ANY recording. If the master was analog, like I said
yesterday, you can figure that 15-16 KHz is about the uppermost limit.