14 Apr
2010
14 Apr
'10
5:40 p.m.
It's all ones and zeroes. Given the same inputs, the same output should be obtained.
Acoustics is physics. And if I can't measure it, it doesn't exist.
yeah well that quite nicely works for machines
but with human sensory systems you'll find they are quite non-linear
and hence the field of psycho-acoustics
which can be interpreted as voodoo by some
14 Apr
14 Apr
6:11 p.m.
New subject: digital voodoo: master fader should be set at 0db
Hi,
On Wednesday 14 April 2010 21:40:08 Kim Cascone wrote:
You are mixing something up here.
If machines can't measure it, it doesn't exist (at least to Ken)[*]. Does that
doesn't mean that if your 0.50€ mic can't record anything from mosquitos
making out, they don't exist?
No, it means that if no mic in the world can record the mosquitos, they don't
make any sound.
The human reception is different then machines reception (not just the non-
linearity!). But it isn't able to detect anything machines can't. Your ears
are worse then any half-decent microphone regarding SNR and frequency-range.
Your eyes are worse then microscopes, telescopes and slower then fast cameras.
Your nose is a lot worse in detecting smells then any mass-spectrometer is.
What makes the human different (some call it superior) is combining this
sensory input not only into facts but also into feelings. And it can combine
thoughts to create new ideas much better then machines combine their input to
even foresee the future, let alone transfer knowledge of one thing onto
predicting behaviour of some other thing.
Of course there are legions of scientists and nerds working on making machines
better in these parts too.
Going into psychoacoustics is not really contradicting the "machines can't
measure it, still it exists". Machines can measure the frequencies the human
ear can't hear but which still have an effect how we perceive the sound. Only
the effects aren't looked into as deep as the frequencies below 20kHz are. To
the result that most scientific research wasn't able to give reliable results.
Which in turn makes most audio people discard frequencies >22kHz light-
heartedly. And they are right as the scientific (thus neutral) proof of the
effect of the frequencies below 22kHz is _much_ greater then above. That
doesn't deny the psycho-acoustic effects, it only ignores them for the sake of
bandwidth, reliability and cost...
Have fun,
Arnold
[*] Might also be that today's machines aren't good enough. Just compare the
knowledge about the atom of today with that of 250 years ago...
It's all
ones and zeroes. Given the same inputs, the same output should
be obtained. Acoustics is physics. And if I can't measure it, it doesn't
exist.
yeah well that quite nicely works for machines
but with human sensory systems you'll find they are quite non-linear
and hence the field of psycho-acoustics
which can be interpreted as voodoo by some 
7:06 p.m.
New subject: digital voodoo: master fader should be set at 0db
On 04/15/2010 06:06 AM, Arnold Krille wrote:
Hi,
On Wednesday 14 April 2010 21:40:08 Kim Cascone wrote:
You are mixing something up here.
If machines can't measure it, it doesn't exist (at least to Ken)[*]. Does that
doesn't mean that if your 0.50€ mic can't record anything from mosquitos
making out, they don't exist?
No, it means that if no mic in the world can record the mosquitos, they don't
make any sound.
The human reception is different then machines reception (not just the non-
linearity!). But it isn't able to detect anything machines can't. Your ears
are worse then any half-decent microphone regarding SNR and frequency-range.
Your eyes are worse then microscopes, telescopes and slower then fast cameras.
Your nose is a lot worse in detecting smells then any mass-spectrometer is.
What makes the human different (some call it superior) is combining this
sensory input not only into facts but also into feelings. And it can combine
thoughts to create new ideas much better then machines combine their input to
even foresee the future, let alone transfer knowledge of one thing onto
predicting behaviour of some other thing.
Of course there are legions of scientists and nerds working on making machines
better in these parts too.
Going into psychoacoustics is not really contradicting the "machines can't
measure it, still it exists". Machines can measure the frequencies the human
ear can't hear but which still have an effect how we perceive the sound. Only
the effects aren't looked into as deep as the frequencies below 20kHz are. To
the result that most scientific research wasn't able to give reliable results.
Which in turn makes most audio people discard frequencies>22kHz light-
heartedly. And they are right as the scientific (thus neutral) proof of the
effect of the frequencies below 22kHz is _much_ greater then above. That
doesn't deny the psycho-acoustic effects, it only ignores them for the sake of
bandwidth, reliability and cost...
Have fun,
Arnold
[*] Might also be that today's machines aren't good enough. Just compare the
knowledge about the atom of today with that of 250 years ago...
Currently we only measure in upto 4 dimensions but if there are 12 as
string theory suggests then we are most likely missing a lot of
information from our scientific measurements.
Patrick Shirkey
Boost Hardware Ltd
It's
all ones and zeroes. Given the same inputs, the same output should
be obtained. Acoustics is physics. And if I can't measure it, it doesn't
exist.
yeah well that quite nicely works for machines
but with human sensory systems you'll find they are quite non-linear
and hence the field of psycho-acoustics
which can be interpreted as voodoo by some
7:07 p.m.
New subject: digital voodoo: master fader should be set at 0db
On Wed, Apr 14, 2010 at 10:06:42PM +0200, Arnold Krille wrote:
On Wednesday 14 April 2010 21:40:08 Kim Cascone
wrote:
Absolutely.
And let me add that psychoacoustics is not the same as
subjectivism - the methods used are what they should be
to qualify as science, results can be verified and
falsified, and there is no voodoo involved at all.
If psychoacoustic experiments show not once but over
and over again that humans can't detect e.g. a sound
in the 2 kHz band if there is another one 50 dB louder
in the 1 kHz band (the numbers are just an example, I
don't claim them to be exact), then for any subjective
result that is in contradiction with this the burden of
proof is on those claiming the subjective result.
You can't just say 'I do hear this' - you have to
explain it as well, and prove it in tests that are
as least as strict as the ones leading to the psycho-
acoustic results which you seem to invalidate. Usually
subjectivist claims fail on both accounts.
Ciao,
--
FA
O tu, che porte, correndo si ?
E guerra e morte !
yeah well that quite nicely works for machines
but with human sensory systems you'll find they are quite non-linear
and hence the field of psycho-acoustics which can be interpreted
as voodoo by some
You are mixing something up here. 
7:17 p.m.
New subject: digital voodoo: master fader should be set at 0db
On Wed, Apr 14, 2010 at 3:06 PM, Arnold Krille <arnold(a)arnoldarts.de> wrote:
Going into psychoacoustics is not really contradicting
the "machines can't
measure it, still it exists". Machines can measure the frequencies the
human
ear can't hear but which still have an effect how we perceive the sound.
Only
the effects aren't looked into as deep as the frequencies below 20kHz are.
To
the result that most scientific research wasn't able to give reliable
results.
Which in turn makes most audio people discard frequencies >22kHz light-
heartedly. And they are right as the scientific (thus neutral) proof of the
effect of the frequencies below 22kHz is _much_ greater then above. That
doesn't deny the psycho-acoustic effects, it only ignores them for the sake
of
bandwidth, reliability and cost...
Sorry to pick on you, but this is the piece of audio voodoo that irks me
most of all. There is no evidence I'm aware of that supports the idea that
ultrasound has any effect whatsoever on human perception of anything. This
point is marketed to no end in the audiophile world, whether it's the
supposed 100k frequency response of virgin vinyl, or the 50k response of
SACD. I believe it's 100% bunk.
If you doubt this, as many people do, try the experiment I used to prove it
to myself:
1) Set up some good synthesizer software (e.g. SuperCollider, csound,
PureData, etc.)
2) Generate pure sine tones of increasing frequency until you reach the
point where you can't hear it (for me this point is just over 16k; for an
older colleague of mine it's just under 14k) (warning: this revelation can
be depressing)
3) Start cranking up the volume. Your VUs will tell you it's there . . .
your amp clipping lights will tell you it's there . . . your dog that runs
screaming from the room will tell you it's there . . . your exploding
tweeters will tell you it's there . . . but you won't hear it. Ever.
4) Now that you've got this horrifyingly loud sound blasting throughout the
room, live your life. Watch some TV. Listen to some music. You won't
notice anything. Better yet, have someone else turn it off and on
periodically, and you note when it's on. You won't reliably be able to.
So . . . if an ungodly pure tone at, say, 18k has no impact on my senses at
all, ain't *no way* that the "air" of a symphony at 30k will. Until some
unbiased A/B/X tests start showing up that prove otherwise, ultrasound is
pointless in music.
I'd love to be proven wrong here.
Footnote: Yes, there's a valid argument for higher than 44.1k when
recording, which involves the filters in A/D converter hardware. There is
no argument for it in the listener space.
7:40 p.m.
New subject: digital voodoo: master fader should be set at 0db
Hi,
On Wednesday 14 April 2010 23:17:05 micromoog wrote:
On Wed, Apr 14, 2010 at 3:06 PM, Arnold Krille
<arnold(a)arnoldarts.de> wrote:
I am not into that field of science but as far as I know there is research in
that there are effects when >22kHz frequencies are in a normal sound, it makes
a different effect then when the signal is clipped frequency-wise at 22kHz.
Difference between a violin played inside the room you are and a violin-
recording you hear from CD.
But I don't actually know if that is proven enough or if its voodoo...
Going into psychoacoustics is not really
contradicting the "machines
can't measure it, still it exists". Machines can measure the frequencies
the human
ear can't hear but which still have an effect how we perceive the sound.
Only
the effects aren't looked into as deep as the frequencies below 20kHz
are. To
the result that most scientific research wasn't able to give reliable
results.
Which in turn makes most audio people discard frequencies >22kHz light-
heartedly. And they are right as the scientific (thus neutral) proof of
the effect of the frequencies below 22kHz is _much_ greater then above.
That doesn't deny the psycho-acoustic effects, it only ignores them for
the sake of
bandwidth, reliability and cost...
Sorry to pick on you, but this is the piece of
audio voodoo that irks me
most of all. There is no evidence I'm aware of that supports the idea that
ultrasound has any effect whatsoever on human perception of anything. If you doubt this, as many people do, try the
experiment I used to prove it
to myself:
1) Set up some good synthesizer software (e.g. SuperCollider, csound,
PureData, etc.)
2) Generate pure sine tones of increasing frequency until you reach the
point where you can't hear it (for me this point is just over 16k; for an
older colleague of mine it's just under 14k) (warning: this revelation can
be depressing)
Back-nitpicking: You have to go down with the frequency if you want to test
your upper frequency limit. Otherwise your ear will claim that it hears
something after you actually stopped hearing it. Or switch the frequencies
randomly... Similar to phantom-images in your eyes, your ears can hear
phantom-frequencies...
3) Start cranking up the volume. Your VUs will tell
you it's there . . .
your amp clipping lights will tell you it's there . . . your dog that runs
screaming from the room will tell you it's there . . . your exploding
tweeters will tell you it's there . . . but you won't hear it. Ever.
4) Now that you've got this horrifyingly loud sound blasting throughout the
room, live your life. Watch some TV. Listen to some music. You won't
notice anything. Better yet, have someone else turn it off and on
periodically, and you note when it's on. You won't reliably be able to.
What if you you realize that you feel better with that frequency on?
If its only you out of 100 people, its voodoo. If its 99 out of 100 people,
its a fact proven by science.
Might be I feel better with such a tone around me. Because I don't like dogs
that much. :-P
So . . . if an ungodly pure tone at, say, 18k has no
impact on my senses at
all, ain't *no way* that the "air" of a symphony at 30k will. Until some
unbiased A/B/X tests start showing up that prove otherwise, ultrasound is
pointless in music.
I'd love to be proven wrong here.
Might be you are proven wrong here. As others claimed in this thread, the ear
is not linear (not even logarithmic in every aspect). Might be that 18kHz
don't affect you while 30kHz do:-)
Footnote: Yes, there's a valid argument for higher
than 44.1k when
recording, which involves the filters in A/D converter hardware. There is
no argument for it in the listener space.
Actually I am not defending the "you have to record at 192kHz, everything else
is crap". From my experience, 48kHz is enough. And cheaper on disk-space...
Have fun,
Arnold
8:58 p.m.
New subject: digital voodoo: master fader should be set at 0db
Excerpts from Arnold Krille's message of 2010-04-14 23:39:44 +0200:
Hi,
On Wednesday 14 April 2010 23:17:05 micromoog wrote:
The violin likely sounds different than speakers, but I guess it's
more likely due to its radiation pattern and things like that which are,
I guess, very hard to capture and reproduce accurately. Just a thought.
Sorry to pick on you, but this is the piece of
audio voodoo that irks me
most of all. There is no evidence I'm aware of that supports the idea that
ultrasound has any effect whatsoever on human perception of anything.
I am not into that field of science but as far as I know there is research in
that there are effects when >22kHz frequencies are in a normal sound, it makes
a different effect then when the signal is clipped frequency-wise at 22kHz.
Difference between a violin played inside the room you are and a violin-
recording you hear from CD.
But I don't actually know if that is proven enough or if its voodoo... 
9:33 p.m.
New subject: digital voodoo: master fader should be set at 0db
On 15 April 2010 06:57, Philipp <hollunder(a)lavabit.com> wrote:
Excerpts from Arnold Krille's message of
2010-04-14 23:39:44 +0200:
Yes, it is actually part voodoo, and (probably) part documented
(lecturers will include it in their 5 minutes of knowledge-sharing).
The science of the particular effect or effects resulting from the
addition of frequencies to > 20kHz to those below that range and
within the audible spectrum have yet to be set in stone (or a thesis).
That is the exact opposite of the advice to handle subsonic
frequencies with great care; you can fly supersonic as far as your
ears cannot handle. Basically, even when you're in an Anechoic
Chamber, you will not be able to benchmark the two cases, but the
signal which has the "enhanced" additions (think exciters in a loose
way) will work its voodoo on your ears and brain receptors, leading to
a more pleasing result of the two. Yes, average Joe will "feel" it
too.
Those kinds of experiments are hard to conclude, as there are very
little numbers aside from the amount of experienced industry guinea
pigs and multiple sittings across weeks or even months.
As you may or may not know, Science is part voodoo too.
--
GPG/PGP ID: B42DDCAD
Hi,
On Wednesday 14 April 2010 23:17:05 micromoog wrote:
The violin likely sounds different than speakers, but I guess it's
more likely due to its radiation pattern and things like that which are,
I guess, very hard to capture and reproduce accurately. Just a thought. Sorry to pick on you, but this is the piece of
audio voodoo that irks me
most of all. There is no evidence I'm aware of that supports the idea that
ultrasound has any effect whatsoever on human perception of anything.
I am not into that field of science but as far as I know there is research in
that there are effects when >22kHz frequencies are in a normal sound, it makes
a different effect then when the signal is clipped frequency-wise at 22kHz.
Difference between a violin played inside the room you are and a violin-
recording you hear from CD.
But I don't actually know if that is proven enough or if its voodoo...
9:59 p.m.
New subject: digital voodoo: master fader should be set at 0db
On 15 April 2010 07:33, Ray Rashif <schivmeister(a)gmail.com> wrote:
addition of frequencies to > 20kHz to those below
that range and
s/to//
:: addition of frequencies 20kHz to those below that range and
Now, with regards to the violin scenario, we all know that that is
close to unattainable at this point of time due to technical
limitations - mainly hardware (I assume we are referring to digital
reproduction). You have to admit, even if not from experience, but
from a theoretical standpoint, that there are or should be noticeable
differences between a live and recorded analogue signal (if it's a
music band/track; anything else is voodoo).
The live one sounds like crap.
How much you can capture to reproduce has and will always depend on
the tool(s) used for the capturing. And as I see it, Physical
Modelling is one workaround in the digital domain, such that in the
future (if it will still be called "Physical Modelling"), it
apparently "should" surpass transducers with the help of computing
prowess.
I don't find that hard to believe, because people are already
"printing" physical 3-dimensional objects from virtual images. Not a
good analogy, but you get it.
--
GPG/PGP ID: B42DDCAD
15 Apr
15 Apr
7:23 a.m.
New subject: digital voodoo: master fader should be set at 0db
On 14/04/2010, Arnold Krille <arnold(a)arnoldarts.de> wrote:
I am not into that field of science but as far as I
know there is research
in that there are effects when >22kHz frequencies are in a normal sound
Then tell us where the research is. It shouldn't be hard to find.
Especially when it's so easy to posit an experiment to do such testing
with DACs that can reproduce signals up to 192kHz. Simply generate,
additively, some interesting spectra with and without frequency
components above a person's established audibility threshold, and see
if they can be ABX'd. (You'll want to ensure that your DAC isn't
curtailing your frequency response prematurely, of course; you'll also
want to ensure that your speakers can properly reproduce those
frequencies, too, rather than only converting them into heat - which
will certainly affect the tonal nature of the speaker's reproduction
of audible frequencies... so decent crossover filtering for your
woofers is probably advised too.)
If a phenomenon can't be ABX'd reliably under advantageous conditions,
it has no objective existence - and anyone trying to claim otherwise
is peddling fairy stories.
What if you you realize that you feel better with that
frequency on?
If you haven't confirmed it objectively, you haven't "realised"
anything - you've merely given yourself a shiny new superstition.
If its only you out of 100 people, its voodoo. If its
99 out of 100 people,
its a fact proven by science.
And the prize for "most egregious misunderstanding of the scientific
process" goes to...
3:18 p.m.
New subject: digital voodoo: master fader should be set at 0db
On Thursday 15 April 2010 11:22:42 Gwenhwyfaer wrote:
On 14/04/2010, Arnold Krille
<arnold(a)arnoldarts.de> wrote:
Me? Because I say 99 out of 100 is scientific proof?
If 99 out of 100 is not enough for you, better not take any medicine. Many of
that is "proven" in studies where 3 out of 10 had an improvement... And yes,
they had a total test group of ten for that study.
Arnold
If its only you out of 100 people, its voodoo. If
its 99 out of 100
people, its a fact proven by science.
And the prize for "most egregious
misunderstanding of the scientific
process" goes to... 
4:07 p.m.
New subject: digital voodoo: master fader should be set at 0db
On Thu, Apr 15, 2010 at 5:22 AM, Gwenhwyfaer <gwenhwyfaer(a)gmail.com> wrote:
On 14/04/2010, Arnold Krille
<arnold(a)arnoldarts.de> wrote:
This is a very good idea save one possible problem. The use of 192kHz
DACs is not generally in order to get reproduction up to 96kHz
signals. It was to ease the analog output stage topology design so
you don't need to use a brick-wall rolloff that's full of ripples. At
least, this was the original motivation. It's entirely possible that
these days there are 192kHz DACs with output filters that have very
nice response well over 22kHz. But I'd test that on a trusted
spectrum analyzer first.
(For example, the old emagic boxes with 96kHz playback could indeed
reproduce up to 48kHz output with good flatness. So I'm sure the
beasts exist, I just wouldn't assume that was the case without
measuring it)
Monty
I am not into that field of science but as far as
I know there is research
in that there are effects when >22kHz frequencies are in a normal sound
Then tell us where the research is. It shouldn't be hard to find.
Especially when it's so easy to posit an experiment to do such testing
with DACs that can reproduce signals up to 192kHz.
4:27 p.m.
New subject: digital voodoo: master fader should be set at 0db
On Thu, Apr 15, 2010 at 10:22:42AM +0100, Gwenhwyfaer wrote:
Especially when it's so easy to posit an
experiment to do such testing
with DACs that can reproduce signals up to 192kHz. Simply generate,
additively, some interesting spectra with and without frequency
components above a person's established audibility threshold, and see
if they can be ABX'd. (You'll want to ensure that your DAC isn't
curtailing your frequency response prematurely, of course; you'll also
want to ensure that your speakers can properly reproduce those
frequencies, too, rather than only converting them into heat - which
will certainly affect the tonal nature of the speaker's reproduction
of audible frequencies... so decent crossover filtering for your
woofers is probably advised too.)
And you need to be sure that your ADs, amps and speaker are
perfectly linear and don't produce products in the < 20 kHz
band when given ultrasonic signals only. For the speakers
that will be near to impossible.
There is at least one 'sound spotlight' system that uses
the nonlinear effects of air to produce audible sound at
the place where two narrow ultrasonic beams cross. IIRC
one of the installations in the ZKM lobby (the one with
the spotlight following a random visitor) uses such a
system. It requires insanely high levels.
Ciao,
--
FA
O tu, che porte, correndo si ?
E guerra e morte !
4:32 p.m.
New subject: digital voodoo: master fader should be set at 0db
And you need to be sure that your ADs, amps and
speaker are
perfectly linear and don't produce products in the < 20 kHz
band when given ultrasonic signals only. For the speakers
that will be near to impossible.
Ohhh, yes, IMD and air itself will introduce IMD. I forgot all about that.
Monty
5327
days inactive
5328
days old
linux-audio-user@lists.linuxaudio.org
13 comments
9 participants
participants (9)
-
Arnold Krille -
fons＠kokkinizita.net -
Gwenhwyfaer -
Kim Cascone -
micromoog -
Monty Montgomery -
Patrick Shirkey -
Philipp -
Ray Rashif