Significance of Inadible High Frequencies to Musical Enjoyment and Relaxation
Harry Lavo
posted
in The Journal of Neurophysiology. Nobody on this forum should skip this
unless you are already familiar with the findings. I haven't seen it
mentioned here.
This reports the results of a scientific test that shows pretty
conclusively that the ultrasonic components of music play a significant role
in musical enjoyment, as recorded both subjectively and by objective brain
activity. This directly contradicts accepted "hearing" theory as it
involves frequencies above 22khz, and it seems to be related directly to
music as opposed to just "hearing".
http://jn.physiology.org/cgi/content/full/83/6/3548
In addition, I call your attention to a link in the 10/21/2002 edition of
ProSoundNews wherein Rupert Neve recalls an incident involving inaudible
frequencies as well as a visit to the laboratory of one of the participants
in the study, one Dr. Tsutomu Oohashi.
http://www.prosoundweb.com/chat_psw/transcripts/rupert.php
While the implications for the SACD and DVD-A vs. CD argument appear obvious
it seems to me to be another piece of evidence that music is "hardwired"
into
the brain and dealt with differently than just sound or hearing per se.
Within the last year I have seen an article (but alas do not have it or its
sources for reference) that scientists have determined seemingly innate
human responses to certain rhythmic patterns and not to others, cutting
across many cultures. It was the scientists conducting this study that
concluded that the brain was "hardwired" in some important ways when it came
to music. The above cited test may be another piece of "evidence" if it
stands up, for it seems to show that ultrasonic hearing by itself is not
important or even "heard" but that when removed from music it becomes a
disabling factor, leading to reduced enjoyment and reduced brain activity
associated with this enjoyment and a reduced sense of happiness when
listening.
Some discussions here recently for which this may be relevant:
* CD's reported lack of emotional "triggering" as a medium
* Importance of listener-controlled-switching in comparison tests
* Ability of ultrasonic behavior to influence perceived audible reproduction
(beyond foldback per se) - eg. cables, amplifiers, loudspeakers, cd formats
* Possible reality of long term musical memory
etc. etc.
It is fascinating how little we really know about how the brain works with
our senses and especially with regard to music.
Harry. Lavo
"it don't mean a thing if it ain't got that swing"
Arny Krueger
news:aqn5u3$uou$1...@bourbaki.localdomain
> http://www.prosoundweb.com/chat_psw/transcripts/rupert.php
Well, here's the quote:
"We discovered that there was a 3 dB peak at 54kHz Geoff's golden
ears had perceived that there was a difference."
What is not known is the order and damping of this resonance.
Depending on the actual parameters, there could be significant
response variations below 20 KHz.
Harry Lavo
news:aqn9cs$ru$1...@bourbaki.localdomain...
True, it is possible. But you are ignoring all the clinical evidence
included in the links.
Denis Sbragion
"Arny Krueger" <ar...@hotpop.com> wrote in news:aqn9cs$ru$1
@bourbaki.localdomain:
...
> What is not known is the order and damping of this resonance.
> Depending on the actual parameters, there could be significant
> response variations below 20 KHz.
furthermore in the Oohashi paper there's nothing about the non linear
properties of the speakers and the 1 bit recording system used. May be the
differences found are caused by distortion folded in the audible range by
the non linear behaviour of the equipment used. Anyway, really an
interesting paper.
Regards,
--
Denis Sbragion
InfoTecna
Tel: +39 0362 805396, Fax: +39 0362 805404
URL: http://www.infotecna.it
dangling entity
to claims of ultrasonic benefits (I'm not bashing the theory-
seriously I'm wondering what speakers). Are there speakers that
really go far beyond 20kHz cleanly? Is this an exclusive capability
to certain ribbon or panel-type transducers? Can conventional dome
tweeters ever go that high?
Gary Rosen
Reported by whom? I get plenty of "emotional triggering" from CDs,
including ones that are remastered versions of songs I originally
heard many times on vinyl.
- Gary Rosen
Arny Krueger
news:aqn5u3$uou$1...@bourbaki.localdomain
> This reports the results of a scientific test that shows pretty
> conclusively that the ultrasonic components of music play a
> significant role in musical enjoyment, as recorded both
> subjectively and by objective brain activity. This directly
> contradicts accepted "hearing" theory as it involves frequencies
> above 22khz, and it seems to be related directly to music as
> opposed to just "hearing".
>
> http://jn.physiology.org/cgi/content/full/83/6/3548
Key to the success of this test would be a lack of changes to the
"low pass" file other than removal of signals above 22-26 KHz.
Interested parties should download this file:
http://jn.physiology.org/content/vol83/issue6/images/large/9k06010470
01.jpeg
Then, use a graphical utility program to select out the images
entitled FRS and HCS which are the full range signal and the
low-passed signal used in the study.
With a simple graphics utility (I used Paintshop Pro 4) it is
possible to get two full-page images that are the same size. I
printed them out. When I superimpose these printed images and hold
them up to the light, I find many areas of significant amplitude
difference in the 0-20 KHz range.
IOW I disagree with the following claim in the paper:
"The spectrum of FRS was essentially the same as that of the source
and contained both LFCs below and HFCs above 22 kHz."
I regretfully am forced to conclude that the study was critically
flawed because the spectral content of the low-pass HCS signal in the
0-20 KHz range was different from the spectral content of the FRS
full-range signal in the 0-20 KHz range.
Arny Krueger
>>> http://www.prosoundweb.com/chat_psw/transcripts/rupert.php
The two links are really not comparable. The first link is an
anecdote, and the second is a peer-reviewed scientific paper. The
first link, the anecdote, suffers from a near-total lack of
documentation. We'll probably never really know what happened in
sufficient detail. The second link is well-documented, and that
documentation proves that the experiment had a serious failing.
Ironically, both of the links suffer from the same basic problem -
being that the effects of the respective filters below 20 KHz were
not fully evaluated by the writers. In the first paper we don't know
for sure what the characteristics of the filter were, but in the
second paper we know for sure that the filter affected frequency
response below 20 KHz.
A fundamental key of any experiment is controlling relevant
variables. When you do an experiment with filters that cause
clearly-observable frequency response differences below 20 KHz, it
seems presumptuous to attribute such observations as are made to
frequency response differences above 20 KHz. This is particularly
ironic given that there is no need in these times for a filter that
separates audio signals in the 22-26 KHz region to cause such big
frequency response differences below 20 KHz.
Thomas A
I've previously read that article and also their former article
regarding the "hypersonic" effect. I have several questions to this
article. First, have anyone else repeated their experiments and if so
do they get the same results? Second, is the high amount of HF unique
to this kind of music/instruments? Would there be an effect if they
made the same experiment with orchestral music? If so, why did they
not show this with ordinary music instruments? That would have
strengthen their conclusions. I've read that it is possible to hear
the effects of HF just above the hearing limit since it can modulate
the audible lower freqeuncies, IF the HF is high enough in level.
Don't know if that is published though.
Thomas
Harry Lavo
news:aqovkc$nip$1...@bourbaki.localdomain...
Read the full article...you'll have your answer.
Arny Krueger
news:aqovkc$nip$1...@bourbaki.localdomain
> I always wondered what sort of speakers are being used when it
> comes to claims of ultrasonic benefits (I'm not bashing the
> theory- seriously I'm wondering what speakers). Are there
> speakers that really go far beyond 20kHz cleanly? Is this an
> exclusive capability to certain ribbon or panel-type transducers?
No. IME many panel type speakers are pretty mediocre in this regard.
> Can conventional dome tweeters ever go that high?
I can substantiate the manufacturer's claim that the Vifa TX25 goes
up to about 40 KHz within a few dB.
Here's the spec sheet:
http://www.partsexpress.com/pdf/264-555.pdf
At least two other Vifa tweeters are specd go up to into the 30's.
http://www.partsexpress.com/pdf/264-566.pdf
http://www.partsexpress.com/pdf/264-578.pdf
At the highest frequencies these drivers are very directional!
Harry Lavo
news:aqord3$li4$1...@bourbaki.localdomain...
> Hello Arny,
>
> "Arny Krueger" <ar...@hotpop.com> wrote in news:aqn9cs$ru$1
> @bourbaki.localdomain:
>
> ...
> > What is not known is the order and damping of this resonance.
> > Depending on the actual parameters, there could be significant
> > response variations below 20 KHz.
>
> furthermore in the Oohashi paper there's nothing about the non linear
> properties of the speakers and the 1 bit recording system used. May be the
> differences found are caused by distortion folded in the audible range by
> the non linear behaviour of the equipment used. Anyway, really an
> interesting paper.
>
Not likely, since the scientists anticipated this and took great pains to
prevent it. Their attention to detail is quite remarkable compared to the
loosy-goosiness of the tests presented and discussed here.
Arny Krueger
>> What is not known is the order and damping of this resonance.
>> Depending on the actual parameters, there could be significant
>> response variations below 20 KHz.
>
> furthermore in the Oohashi paper there's nothing about the non
> linear properties of the speakers and the 1 bit recording system
> used. May be the differences found are caused by distortion
> folded in the audible range by the non linear behaviour of the
> equipment used. Anyway, really an interesting paper.
I think this point is very important. Key to the success of this
experiment is controlling variables so that the two musical samples
differed ONLY in that one had spectral content above 22-26 KHz. It is
also important to show that the ENTIRE reproduction system was free
of nonlinear distortion, particularly at high frequencies.
Freedom from nonlinear distortion is very important because
relatively small amounts of nonlinear distortion can cause distortion
products that will be appear in the range where the human ear is
known to be highly sensitive.
In fact the http://jn.physiology.org/cgi/content/full/83/6/3548 paper
shows that the two music samples differed significantly both above
and below 22-28 KHz. There is no evidence showing that there weren't
other differences as well. Furthermore, there is no objective
evidence that the reproduction system was free of nonlinear
distoriton.
Arny Krueger
> "Denis Sbragion" <d.sbr...@infotecna.it> wrote in message
> news:aqord3$li4$1...@bourbaki.localdomain...
>> "Arny Krueger" <ar...@hotpop.com> wrote in news:aqn9cs$ru$1
>> @bourbaki.localdomain:
>> ...
>>> What is not known is the order and damping of this resonance.
>>> Depending on the actual parameters, there could be significant
>>> response variations below 20 KHz.
>> furthermore in the Oohashi paper there's nothing about the non
>> linear properties of the speakers and the 1 bit recording system
>> used. May be the differences found are caused by distortion
>> folded in the audible range by the non linear behavior of the
>> equipment used. Anyway, really an interesting paper.
> Not likely, since the scientists anticipated this and took great
> pains to prevent it.
Where's the evidence to support this claim? I've gone over the paper
and in another post I've clearly shown that the scientists mustn't
have even done a trivial comparison between two of their own
graphics.
BTW, I performed a similar comparison among a sampling of the 24/96
..wav files that people can download from
http://www.pcabx-pro.com/technical/sample_rates/index.htm and below
20 KHz, the spectral analysis lines fell right on top of each other
almost entirely pixel-for pixel, even when blown up to double the
vertical pixels and about 1/3 more horizontal pixels.
> Their attention to detail is quite
> remarkable compared to the loosy-goosiness of the tests presented
> and discussed here.
At best, that would be a matter of opinion.
Let's compare http://jn.physiology.org/cgi/content/full/83/6/3548 to
the standard that subjectivists have been demanding around here:
Peer-reviewed: yes
Test performed using exclusively commercial high end audio equipment:
NO!
Test performed using exclusively experienced high end audiophiles:
NO!
Interesting how *the rules* change when a subjectivist is presenting
the evidence!
Arny Krueger
news:aqovr...@enews4.newsguy.com
> I've previously read that article and also their former article
> regarding the "hypersonic" effect. I have several questions to
> this article. First, have anyone else repeated their experiments
> and if so do they get the same results?
Anybody with some kind of PC, mac or Unix box, a 24/96 sound card and
appropriate speakers and amps can compare music with and without > 22
KHz content by downloading files from
http://www.pcabx-pro.com/technical/sample_rates/index.htm .
>Second, is the high amount of HF unique to this kind of
music/instruments?
No. Two other references show that common Western musical instruments
can have very substantial spectral output above 22 KHz. One reference
showing this is:
http://www.cco.caltech.edu/~boyk/spectra/spectra.htm
and the other is:
http://www.pcabx-pro.com/technical/sample_rates/index.htm
>Would there be an effect if they made the same experiment with
> orchestral music?
That depends which parts of which musical selections you pick, and
whether or not the experiment can possibly have a positive outcome if
correctly done.
>If so, why did they not show this with ordinary
> music instruments?
Good question.
>That would have strengthen their conclusions.
> I've read that it is possible to hear the effects of HF just
> above the hearing limit since it can modulate the audible lower
> frequencies, IF the HF is high enough in level. Don't know if
> that is published though.
I'd like to see the article if it was published.
Harry Lavo
> I am attaching below a link to a clinical study conducted in Japan and
> posted
> in The Journal of Neurophysiology. Nobody on this forum should skip this
> unless you are already familiar with the findings. I haven't seen it
> mentioned here.
>
>snip<
..
>
> http://jn.physiology.org/cgi/content/full/83/6/3548
>
>snip<
After rereading the Oohashi et al article "Inaudible High-Frequency Sounds
Affect Brain Activity: Hypersonic Effect " for a second time at my leisure
last night (after posting the above link here on RAHE) I was struck how in
one giant research leap, this effort (if supported by confirming research)
can be reasonably by interpreted to support much of the subjectivist points
of view expressed here on RAHE and challenges much of the dogma thrown out
by the objectivist camp. If the purpose of that camp is to convert RAHE to
objectivist hegemony, than a large boulder has just been placed smack dab in
the middle of the road.
Everybody should read this article as soon as possible. The test basically
was done using a three minute piece of Balinese gamelan music, rich in
inaudible overtones reaching up to 50-100khz. The audible music and
inaudible overtones were recorded and played back using a special recorder
able to reach this high, and were presented using separate sound files and
loudspeakers for the audible and inaudible parts, so there was no foldback
effect or interaction. They were synced so as to appear as one source. And
care was made in recording the inaudible overtones that no audible artifacts
appeared. This was later confirmed by the fact that ambient noise and the
presence of inaudible overtones recorded exactly the same brain response and
psychological evaluation.
The actual number of test subjects, which tests were given to what group,
etc. are in the article and I will not attempt to summarize them hear. What
is important is that the tests were double blind and the results had to be
significant with less than 5% chance of being wrong (by chance). In
reality, most of the results were significant at the 1% level. The four
variables tested were music without overtones, music with overtones,
overtones without music, and ambient noise. Physiological measurements were
made using multiple location EEG measurements and blood flow measurements
within the brain. Psychological questions were self-administered relating
to the listeners state of well being, and to their perception of the music
being played (when appropriate). Even when subjects were "wired", great
care was exerted to keep the surroundings as pleasant and homelike as
possible. A great deal of sophisticated statistical work was done to
"normalize" the results.
For those who have not already read the study, I will try to explain the
basic findings in laymen's terms.
1) Music with inaudible overtones intact registered a far different reaction
than either music without overtones, plain ambient noise, or overtones w/o
audible music. Music with overtones was self-rated to sound different,
sound richer, and be more relaxing) and physiological measurement showed an
increase in brain activity and blood flow involving known "pleasure centers"
and "mechanisms" of the brain.
2) Music without the overtones showed no difference from ambient background
noise as far as brain activity was concerned.
3) Inaudible overtones by themselves did not register...they only registered
when combined with the audible music as a complete bundle. By themselves
they registered only the same as ambient background noise.
4) It took awhile for the brain to react to the stimulus....the build up
took "tens of seconds" (20-40?) to register, and a like time to unregister
following a switch back to overtoneless music, ambient noise, or overtones
without music.
Now with that behind us, let's look at the implications for the debates
raging here on RAHE. I will start with the most definitive and move on to
the more speculative implications.
1) Do SACD and/or DVD-A really sound better (e.g.. more natural, "real-er")
than CD or is that all in the imagination? The research clearly indicates
(again, if eventually repeated and verified) that SACD and or DVD-A *SHOULD*
sound better ("clearer","richer") and that CD's inherent frequency
limitation turns out to be a very bad thing, perception-of-music-wise.
2) Can SACD and/or DVD-A really "relax" us more than CD? Again, the
research pretty clearly states that this is a concomitant of wider bandwidth
when reproducing music.
3) Is Quick-Switch ABX testing a la Arnie the best way to evaluate music?
The CCIR protocols oft cited by Arnie as the "bible" for quick-switch ABX
testing are directly challenged by Oohashi and the nine other authors. They
flatly suggest that the research conducted in the early eighties and given
as support for the CD 22khz cutoff were wrong (because both the
quick-switching and the short excerpts used "masked" the true reaction of
the subjects by not allowing enough time for the brain to react to any but
the audible portion of the signal.. Oohashi et al devote a summary section
specifically to this, which I've included as a footnote to this post.(1).
4) Is hearing music the same thing as hearing anything else? The evidence
presented hear pretty conclusively suggests not. Music has its own "rules"
for the human physiology. One can properly infer, therefore, that what may
be appropriate for discerning audible artifacts may not *necessarily* be
appropriate for discerning/evaluating music.
4) Is evaluating music in a relaxed sighted, state using "pleasure" as a
measure of correctness an subjectivist audiophile fantasy? This test
suggests that it may be a highly significant "indicator", especially of the
system's ability to present the full overtone structure of music without
distortion, attenuation, or time-smear.
5) Is their any support for the notion that evaluators should be able to
control switching? Not directly, but indirectly the fact that it takes
20-40 seconds for a "pleasure state" to emerge and the old one to disappear
suggests that users who know a piece, using their own music, on their own
systems should be able to anticipate their response and evaluation of the
upcoming segment of music, and compare the response and evaluation they
actually get (from the other piece of gear switched in) against this fading
"memory". But this will only work if they control the switching and are
focused on the emotional and sound evaluation reaction of this familiar
piece of music. This is obviously a hypothesis to be pursued, it certainly
cannot be dismissed out of hand as in the past by many objectivists, based
on this research presented in this article.
6) Can cables or amplifiers with similar electrical characteristics "sound
different"? The scientists here hypothesized that the inaudible overtones
of music may be sensed directly through the skull or other parts of the body
(as in some other creatures). If this is true, then what happens in the
inaudible high frequency part of the spectrum that affects proper
reproduction of the sound (frequency deviation, time-smear, rfi
interference, etc.) could conceivable affect sound, not by fold-back into
the audible part of the spectrum, but directly.
There are probably many more implications possible arising from this
research; my guess it will become a *landmark* in the field of audio
reproduction and hearing physiology.
*****FOOTNOTES*****
(1) The following quote is one subsummary from the source article:
"EXPLANATION OF THE DESCREPANCY BETWEEN THE PRESENT AND PREVIOUS STUDIES"
"The fact that we used an entire piece of natural music lasting 200 s as
sound stimuli instead of short fragments of sounds might explain the
discrepancy between our findings and those of previous studies carried out
around 1980 to determine the format for digital audio CDs (e.g., Muraoka et
al. 1978; Plenge et al. 1979), which concluded that the presence of sounds
containing a frequency range above 15 kHz was not recognized as making a
difference in sound quality. The CCIR (1978), and the current International
Telecommunication Union-Radio communication sector (ITU-R 1997), have
recommended that sound samples used for the comparison of sound quality
should not last longer than 15-20 s (CCIR 1978; ITU-R 1997), and that
intervals between sound samples should be about 0.5-1 s (CCIR 1978) because
of short-term human memory limitations. Most of the previous psychological
experiments, including the studies by Muraoka et al. (1978) and Plenge et
al. (1979), were carried out using, essentially, the sound presentation
method recommended by the CCIR. We also examined the psychological
evaluation using the same material and sound presentation system as was used
for the present study, but followed the presentation method recommended by
the CCIR, and confirmed that the results were in agreement with the studies
by Muraoka et al. (1978) and Plenge et al. (1979).
"In our EEG and PET experiments, we focused on physiological brain responses
and objectively evaluated the effect of the combination of audible sounds
and inaudible HFCs on brain activity, independent of a subjective evaluation
of sound quality. According to the EEG measurements, the occipital alpha-EEG
gradually increased over several tens of seconds after the exposure to FRS
began, and this increase persisted for several tens of seconds after FRS
ended. These findings suggest that the phenomenon that we call the
hypersonic effect may involve some neuronal mechanisms that can be
characterized by delay and persistence for as long as several tens of
seconds. It seems, therefore, that an exposure to FRS shorter than 20 s, as
recommended by the CCIR and ITU-R, may be insufficient to introduce a
physiological effect. By the same token, a short exposure to HCS following
FRS with a short interval of 0.5-1 s may not be enough to withdraw
physiological effects, if any, induced by the preceding FRS. Based on this
physiological consideration, we performed our psychological experiment with
sound materials of longer duration. The results showed a significant
difference between FRS and HCS in some elements of sound quality. That
difference was evident despite the fact that a long presentation time should
make it more difficult to detect subtle differences between two materials
due to the limitation of short-term auditory memory. Our findings suggest
the possibility that the results of the previous psychological studies may
not be valid in a situation where humans are continuously exposed to
auditory stimuli such as music or environmental sounds. "
SOURCE: "Inaudible High-Frequency Sounds Affect Brain Activity: Hypersonic
Effect ",
Tsutomu Oohashi, Emi Nishina, Manabu Honda, Yoshiharu Yonekura, Yoshitaka
Fuwamoto, Norie Kawai, Tadao Maekawa, Satoshi Nakamura, Hidenao Fukuyama,
and Hiroshi Shibasaki, The Journal of Neurophysiology Vol. 83 No. 6 June
2000, pp. 3548-3558 Copyright ©2000 by the American Physiological Society
Harry Lavo
Sorry, Arnie....you are the one who has decided this is a problem...the
authors appear to be fully cognizant of this problem and to have addressed
it. Have you written to them about the techniques used; obtained copies of
their sound samples, reproduced them using separate syncronized systems as
per their efforts? If not, you are just reacting defensively
(understandable, since the study badly undercuts many of your beliefs).
Also, do you have any hypothesis about why then these "audible" folddown
byproducts don't create the same response when played as "only audible
frequencies" as when played with the inaudible overtones present? You don't
and can't --- it was controlled as a variable in the test...if you had read
it carefully you would have noted that they claim to have used the *EXACT
SAME* audible clip...with the overtones added by a completely separate
synchronized system.
Harry Lavo
news:aqp6r...@enews2.newsguy.com...
What a herring! This was a research study. Even we lowly subjectivists
know the difference between a research study and home evaluation of audio
components.
Harry Lavo
Arnie, instead of a knee-jerk, read the study in its completeness. They
(the authors) anticipated just this criticism and actually modified the
cutoff and devised a means of presenting the sound specifically to get
around this problem.
Harry Lavo
news:aqp5sl$ql8$1...@bourbaki.localdomain...
The test *CONTROLLED* for these variables. You claim to have read it. I
doubt it. They played the SAME FILE with and without overtones. The file
*WITHOUT* overtones showed no heightened brain activity, no increase in the
pleasure centers, no subjective improvement in sound or reduction in stress.
The test *WITH* overtones did. If they used the same file (and they say
they did) then the "audible artifacts" that you claim exist, exist in both
cells of the test. Now explain to me how your hypothesis would give this
result!
You are creating an obfuscatory hypothesis that doesn't stand up to the
facts, in order to protect your beliefs. The exact thing you have always
accused the subjectivists of doing.
Harry Lavo
>>
>snip<
> The actual number of test subjects, which tests were given to what group,
> etc. are in the article and I will not attempt to summarize them hear.
What
> is important is that the tests were double blind and the results had to be
> significant with less than 5% chance of being wrong (by chance). In
> reality, most of the results were significant at the 1% level. The four
> variables tested were music without overtones, music with overtones,
> overtones without music, and ambient noise. Physiological measurements
were
> made using multiple location EEG measurements and blood flow measurements
> within the brain. Psychological questions were self-administered relating
> to the listeners state of well being, and to their perception of the music
> being played (when appropriate). Even when subjects were "wired", great
> care was exerted to keep the surroundings as pleasant and homelike as
> possible. A great deal of sophisticated statistical work was done to
> "normalize" the results.
>
I should have added here that according to the authors the exact same
audible sound file was used for both cells of the test featuring music. So
fold-down as proposed by Arnie (an minimized in any case by the authors)
could not affected the outcome...if present, it was present in both
cells...with and without overtones...yet the results were different at the
1% significance level.
Steve G
(...)
>
> With a simple graphics utility (I used Paintshop Pro 4) it is
> possible to get two full-page images that are the same size. I
> printed them out. When I superimpose these printed images and hold
> them up to the light, I find many areas of significant amplitude
> difference in the 0-20 KHz range.
>
> IOW I disagree with the following claim in the paper:
>
> "The spectrum of FRS was essentially the same as that of the source
> and contained both LFCs below and HFCs above 22 kHz."
>
Well, with this eyeballing you are assuming that the graphs are
produced accurately; grabbing data from the print may or may not be
sufficient. I wouldn't bet my hat on it.
However, notwithstanding this, the effects could be explained by
nonlinearities in the ear. An auditory psychophysicist of my
acquaintance writes:
"Whilst it is likely that any frequency component above 20 kHz is
inaudible, the inner ear generates nonlinear distortion products -
called combination tones - from "primaries" in the original sound.
These are lower in frequency than the primaries (e.g. cubic
difference tone 2f1-f2). Therefore, two inaudible primaries may
generate an audible primary. I suspect that this is what is happening
- rather than any "magic" or "subliminal perception" (which falls
under "magic" anyway)"
Indeed, I recently went to a talk where the speaker demonstrated such
effects by playing a tune using distortion products - i.e., he didn't
play any of the notes we heard! Even if the audio sample filtering was
carried out correctly - and Arny's investigation suggests maybe not -
this might allow for discrimination between the conditions (i.e., test
tunes), without invoking any dark magic.
These distortion products were first discovered back in the 1800s,
BTW.
Some further comments:
The authors did not carry out any formal discrimination tests on the
audio samples. (It is not sufficient to just say that 'blindfolded
subjects could not hear any differences'). Therefore discrimination
between samples may have been possible. Indeed, this is extremely
likely given that different subjective ratings were given to the
different samples - it's axiomaic that any time a rating-difference
exists, a difference in discrimination will also be found.
- Or more simply: It's not subliminal!
Next, there is no reason to accept that these effects are restricted
to music. If the effects are distortion products, then they will occur
with or without 'music' per se. And if Arny's correct, then the whole
effect is void anyway - but I think the actual raw data are really
needed to make sure of this.
Some facetious soul might say that the authors have a vested interest
in finding (or emphasising the existence of) high-frequency effects,
given that they have links to Philips! A facetious soul, but not me.
An interesting paper - thanks for pointing it out, Harry.
Steve.
Harry Lavo
Reread the article. They built the reproduction system carefully to prevent
any artifacts, and raised the cutoff frequency for the high frequency filter
to specifically prevent audible artifacts. They also controlled for this
variable in the test design itself.
Arny Krueger
That means that they played two different files, right?
> The file *WITHOUT* overtones showed no heightened
> brain activity, no increase in the pleasure centers, no
> subjective improvement in sound or reduction in stress.
So far so good.
> The test *WITH* overtones did.
The problem here is that the file with overtones differed
substantially BELOW 20 KHz from the file without overtones. This is
based on graphical comparsions of the spectral content of the
low-pass HCS signal in the 0-20 KHz range and the spectral content of
the FRS full-range signal in the 0-20 KHz range.
>If they used the same file (and they say
> they did) then the "audible artifacts" that you claim exist,
> exist in both cells of the test. Now explain to me how your
> hypothesis would give this result!
Harry you seem to be contradicting yourself and the available facts.
(1) There were two files.
(2) The file with overtones differed substantially BELOW 20 KHz from
the file without overtones.
> You are creating an obfuscatory hypothesis that doesn't stand up
> to the facts, in order to protect your beliefs. The exact thing
> you have always accused the subjectivists of doing.
I believe that this is an empty claim without proof of error or
obfuscation. My efforts have been to shed light by performing
comparisons that anybody should be able to duplicate for themselves.
Harry Lavo
news:aqpc41$t2v$1...@bourbaki.localdomain...
>snip reply to Arny<
> However, notwithstanding this, the effects could be explained by
> nonlinearities in the ear. An auditory psychophysicist of my
> acquaintance writes:
>
> "Whilst it is likely that any frequency component above 20 kHz is
> inaudible, the inner ear generates nonlinear distortion products -
> called combination tones - from "primaries" in the original sound.
> These are lower in frequency than the primaries (e.g. cubic
> difference tone 2f1-f2). Therefore, two inaudible primaries may
> generate an audible primary. I suspect that this is what is happening
> - rather than any "magic" or "subliminal perception" (which falls
> under "magic" anyway)"
>
> Indeed, I recently went to a talk where the speaker demonstrated such
> effects by playing a tune using distortion products - i.e., he didn't
> play any of the notes we heard! Even if the audio sample filtering was
> carried out correctly - and Arny's investigation suggests maybe not -
> this might allow for discrimination between the conditions (i.e., test
> tunes), without invoking any dark magic.
>
> These distortion products were first discovered back in the 1800s,
> BTW.
>
Yep, and that's why the scientists went as far as they did tto prevent any
interatction that was not "natural'...if you read their test and
reproduction setup carefully, you will see that.
> Some further comments:
>
> The authors did not carry out any formal discrimination tests on the
> audio samples. (It is not sufficient to just say that 'blindfolded
> subjects could not hear any differences'). Therefore discrimination
> between samples may have been possible. Indeed, this is extremely
> likely given that different subjective ratings were given to the
> different samples - it's axiomaic that any time a rating-difference
> exists, a difference in discrimination will also be found.
>
> - Or more simply: It's not subliminal!
The discrimination tests were in the physiological measurements...clearly
there was a significant difference (at <1% error) that their *WAS* a
discrimination. But it was not aural, it was subliminal.
>
> Next, there is no reason to accept that these effects are restricted
> to music. If the effects are distortion products, then they will occur
> with or without 'music' per se. And if Arny's correct, then the whole
> effect is void anyway - but I think the actual raw data are really
> needed to make sure of this.
>
Not necessarily restricted to music, but not generated by random noise,
either. Neither the ambient background noise nor reproduction of the
non-audible high-frequency overtones without an audible musical base
provided any discimination or differences whatsoever. So it only happens
when the brain "makes sense" of something and expects that "high frequency
content" to be there as it occurs in nature.
Do you remember past discussions here? For example, if presented with a
"snippet" of the sound of an auto crash, would you know what it was? With
no context?
The author's actually mention that the rainforests contain an abundance of
inaudible sounds and that they would expect that human evolution would have
taken care of some ability to differentiate if they were important to
survival.
> Some facetious soul might say that the authors have a vested interest
> in finding (or emphasising the existence of) high-frequency effects,
> given that they have links to Philips! A facetious soul, but not me.
>
Actually support and funding was listed by the authors:
" ACKNOWLEDGMENTS "
"We thank the staff of the Kyoto University PET Center for valuable
contributions to this work; Dr. Yoshio Yamasaki, Waseda University, for the
use of his recently developed signal processing system; the Yamashiro
Institute of Science and Culture for recording the sound sources; Dr.
Norihiro Sadato, National Institute for Physiological Sciences, for valuable
comments on an early version of the manuscript; and Dr. Masako Morimoto,
Japan Society for the Promotion of Science, for valuable technical support.
"
"This work was supported in part by the Japan Ministry of Education, Science
and Culture, through the Grants-in-Aid for Scientific Research (A)
(09490031) to T. Oohashi, on Priority Areas to H. Shibasaki, and for
International Scientific Research Program (10041144) to T. Oohashi, and by
the Japan Society for the Promotion of Science through the Research for the
Future Program JSPS-RFTF 97L00201 to H. Shibasaki. "
I don't see Phillips (or Sony) listed....I was expecting perhaps they funded
it. But maybe they did indirectly through the government agencies. On the
other hand, the DVD-A people would have an equally strong interest. Perhaps
it is a neutral government entitiy.
> An interesting paper - thanks for pointing it out, Harry.
>
> Steve.
>
My pleasure, but I need to give credit to Takeshi Andoh who originally
brought it to my attention in another forum.
Arny Krueger
> "Arny Krueger" <ar...@hotpop.com> wrote in message
> news:aqp6r...@enews2.newsguy.com...
>> Let's compare
>> http://jn.physiology.org/cgi/content/full/83/6/3548 to the
>> standard that subjectivists have been demanding around here:
>> Peer-reviewed: yes
>> Test performed using exclusively commercial high end audio
>> equipment: NO!
>> Test performed using exclusively experienced high end
>> audiophiles: NO!
>> Interesting how *the rules* change when a subjectivist is
>> presenting the evidence!
> What a herring! This was a research study.
So were most of the papers published in the JAES.
> Even we lowly
> subjectivists know the difference between a research study and
> home evaluation of audio components.
What this mean about Ludovic Mirabel, who has made demands based on
these key points?
Arny Krueger
> My pleasure, but I need to give credit to Takeshi Andoh who
> originally brought it to my attention in another forum.
I think that it should be pointed out that substantially the same
paper is in the AES archives:
High-Frequency Sound Above the Audible Range Affects Brain Electric
Activity and Sound Perception
Author(s): Oohashi, Tsutomu; Nishina, Emi; Kawai, Norie; Fuwamoto,
Yoshitaka; Imai, Hiroshi
Publication: Preprint 3207; Convention 91; October 1991
Abstract: The authors have developed a new system for sound
presentation and a novel technique to measure brain electric
activity, and used them to determine if high frequency components
above the audible range can influence sound perception in ways not
discerned by the method of paired comparisons. The authors report
that high frequency sound induces activation of electroencephalogram
rhythms that persist in the absence of high frequency stimulation,
and can affect perception of sound quality.
Please notice the date of this conference preprint - 1991. In over 10
years has apparently not been approved by the AES review board.
Is it the same article? The AES paper's opening paragraph is:
It is generally accepted that audio frequencies above 20 kHz do not
affect human sound perception since they are beyond the audible
range[I]-[7]. Consequently, most audio equipment now in use is not
designed to accommodate such frequencies.
Compare that to http://jn.physiology.org/cgi/content/full/83/6/3548
which says
"It is generally accepted that audio frequencies above 20 kHz do not
affect human sensory perception since they are beyond the audible
range Thus for example, most of the conventional commercial digital
audio formats have been standardized to a frequency range that does
not allow such high-frequency components of sounds to be included."
Obviously this paper has been written and re-written for about a
decade for the purpose of being published in some refereed journal
some place.
Harry Lavo
What it means is when we are talking about home audio evaluations, we are
talking about home audio evaluations.
When we are talking about scientific experiments, we are talking about
scientific experiments.
When we are talking about music, we are talking about music.
When we are talking about components we are talking about components.
Ludovic has always tried to focus the discussion on the use of ABX'ng for
determining differences (before preferences) of audio components. That's
why he has focused on a) small-scale testing, b) testing actual high end
audio components. You guys have dragged HK's research testing into the
fold...Ludovic's only comments there are to listener sensitivity using ABX,
not to whether the tests were components or not.
More herring!
Arny Krueger
>> http://jn.physiology.org/cgi/content/full/83/6/3548
> After rereading the Oohashi et al article "Inaudible High-
> Frequency Sounds Affect Brain Activity: Hypersonic Effect "
( http://jn.physiology.org/cgi/content/full/83/6/3548)
> for a
> second time at my leisure last night (after posting the above
> link here on RAHE) I was struck how in one giant research leap,
> this effort (if supported by confirming research) can be
> reasonably by interpreted to support much of the subjectivist
> points of view expressed here on RAHE and challenges much of the
> dogma thrown out by the objectivist camp.
I was struck by the irony of a leading RAHE subjectivist's newfound
excitement over this paper that is basically over 11 years old. It
has been apparently been languishing for years in its search for a
professional journal editorial review committee that presumably would
condescend to approve it. Two years ago it found a home. It appears
that as of yesterday or so, Mr. Lavo finally discovered it.
Of course, many of us have been aware of the Oohashi paper for years,
tracing back to its presentation 11 years ago at an AES Convention.
An example of a RAHE post I authored mentioning the Oohashi AES
article can be found at
http://groups.google.com/groups?selm=68thf2%24qnj%40jamesv.Warren.MEN
TORG.COM dated 1/6/1998, almost 5 years ago. Truth be known, my RAHE
post was anticipated by about 4 years by a RAHE post by one Mike
Coleman available online at
http://groups.google.com/groups?selm=MC.92Jun18222404%40soul.tv.tek.c
om
For time immemorial, sound quality has been related to the
(conscious) enjoyment of music. Objectivists and subjectivists alike
have agreed on this point for years, if only this point.
It appears that in the name of a higher form of subjectivism, sound
quality has been redefined by Mr. Lavo in terms of indications on
electroencephalogram machines, e.g. meters.
Stewart Pinkerton
wrote:
>I should have added here that according to the authors the exact same
>audible sound file was used for both cells of the test featuring music. So
>fold-down as proposed by Arnie (an minimized in any case by the authors)
>could not affected the outcome...if present, it was present in both
>cells...with and without overtones...yet the results were different at the
>1% significance level.
Total speculation, based on a HUGE assumption that the 'HF' playback
system produced no LF artifacts.
--
Stewart Pinkerton | Music is Art - Audio is Engineering
Stewart Pinkerton
wrote:
>"Harry Lavo" <harry...@rcn.com> wrote in message
>news:aqn5u3$uou$1...@bourbaki.localdomain...
>> I am attaching below a link to a clinical study conducted in Japan and
>> posted
>> in The Journal of Neurophysiology. Nobody on this forum should skip this
>> unless you are already familiar with the findings. I haven't seen it
>> mentioned here.
>>
>
>>snip<
>..
>>
>> http://jn.physiology.org/cgi/content/full/83/6/3548
>>
>
>>snip<
>
>After rereading the Oohashi et al article "Inaudible High-Frequency Sounds
>Affect Brain Activity: Hypersonic Effect " for a second time at my leisure
>last night (after posting the above link here on RAHE) I was struck how in
>one giant research leap, this effort (if supported by confirming research)
Confirming research, of course, is the key. Interesting that this
paper has *never* passed peer review by the AES, despite its great
interest to the audio engineering community - IF TRUE. The paper was
certainly *presented* at an AES convention more than ten years ago -
but never published by the AES - and it *should* have created great
excitement and lots of follow-on studies, especially now that wideband
distribution media are commercially available. However, back here in
the real world, there is absolutely *no* corroborating evidence - even
from Sony!
>can be reasonably by interpreted to support much of the subjectivist points
>of view expressed here on RAHE and challenges much of the dogma thrown out
>by the objectivist camp. If the purpose of that camp is to convert RAHE to
>objectivist hegemony, than a large boulder has just been placed smack dab in
>the middle of the road.
The purpose is to improve the sound of reproduced music in the home,
without the aid of wallet-lightening voodooo and snake-oil products.
>Everybody should read this article as soon as possible. The test basically
>was done using a three minute piece of Balinese gamelan music, rich in
>inaudible overtones reaching up to 50-100khz. The audible music and
>inaudible overtones were recorded and played back using a special recorder
>able to reach this high, and were presented using separate sound files and
>loudspeakers for the audible and inaudible parts, so there was no foldback
>effect or interaction.
You are now assuming that the 'HF' system produced no intermodulation
artifacts in the LF range. BIG assumption...................
<snip of great swathes of speculations and extrapolations>
Harry Lavo
Arny, its a paper with a doubled set of authors, additional tests, many
up-to-date references. It is not the same paper, and to suggest it is
smacks of intellectual dishonesty. Moreover, you do not have a clue as to
why or how it was revised..the first one was an audio convention paper...the
second one is a medical paper in a medical journal. I have pointed this
out to you in another post, which you have so far ignored. This is your
second attempt to "smear" the paper by innuendo. So much for character.
> For time immemorial, sound quality has been related to the
> (conscious) enjoyment of music. Objectivists and subjectivists alike
> have agreed on this point for years, if only this point.
>
And this paper and series of tests shows that sound recorded with
frequencies chopped off fails to deliver the goods.
> It appears that in the name of a higher form of subjectivism, sound
> quality has been redefined by Mr. Lavo in terms of indications on
> electroencephalogram machines, e.g. meters.
>
It appears to protect his "bizness" Mr. Kreuger will resort to inuendo when
he cannot challenge the test itself. Again, RAHE readers can (and should)
judge for themselves.
Harry Lavo
news:aqqf36$c3k$1...@bourbaki.localdomain...
This is the first time this work has been published in a peer reveiw
journal..only two years ago. For a large, rigorous study such as this to be
replicated will take time...and often in science the test itself is not
replicated...but its findings are corraborated as part of the results of
additional studies exploring slightly different topics. That process has, I
am sure, started.
> >can be reasonably by interpreted to support much of the subjectivist
points
> >of view expressed here on RAHE and challenges much of the dogma thrown
out
> >by the objectivist camp. If the purpose of that camp is to convert RAHE
to
> >objectivist hegemony, than a large boulder has just been placed smack dab
in
> >the middle of the road.
>
> The purpose is to improve the sound of reproduced music in the home,
> without the aid of wallet-lightening voodooo and snake-oil products.
>
That's fine...but along the way "scientific orthodoxy" as propounded by its
proponents has been put forth as the unassailable standard. And as usual
the case with science, "orthodoxy" often has a way of evolving...
> >Everybody should read this article as soon as possible. The test
basically
> >was done using a three minute piece of Balinese gamelan music, rich in
> >inaudible overtones reaching up to 50-100khz. The audible music and
> >inaudible overtones were recorded and played back using a special
recorder
> >able to reach this high, and were presented using separate sound files
and
> >loudspeakers for the audible and inaudible parts, so there was no
foldback
> >effect or interaction.
>
> You are now assuming that the 'HF' system produced no intermodulation
> artifacts in the LF range. BIG assumption...................
>
I'm not assuming it...it was designed out of the test. READ THE TEST.
Here, I'll make it easy for you.....
Arny Krueger
> "Arny Krueger" <ar...@hotpop.com> wrote in message
> news:aqovj...@enews4.newsguy.com...
>> "Harry Lavo" <harry...@rcn.com> wrote in message
>> news:aqnb94$1l2$1...@bourbaki.localdomain
>>> "Arny Krueger" <ar...@hotpop.com> wrote in message
>>> news:aqn9cs$ru$1...@bourbaki.localdomain...
>> Ironically, both of the links suffer from the same basic problem
>> - being that the effects of the respective filters below 20 KHz
>> were not fully evaluated by the writers. In the first paper we
>> don't know for sure what the characteristics of the filter were,
>> but in the second paper we know for sure that the filter
>> affected frequency response below 20 KHz.
>> A fundamental key of any experiment is controlling relevant
>> variables. When you do an experiment with filters that cause
>> clearly-observable frequency response differences below 20 KHz,
>> it seems presumptuous to attribute such observations as are made
>> to frequency response differences above 20 KHz. This is
>> particularly ironic given that there is no need in these times
>> for a filter that separates audio signals in the 22-26 KHz
>> region to cause such big frequency response differences below 20
>> KHz.
> Sorry, Arnie....you are the one who has decided this is a
> problem...the authors appear to be fully cognizant of this
> problem and to have addressed it.
If you think they addressed it Harry, why don't you inform us with
some appropriate direct quotes from the article. One would be an
explicit admission that there were significant differences between
the two spectral analysis below 33-36 KHz, and the other would be an
explicit discussion of the means by which they addressed the problem.
> Have you written to them about
> the techniques used; obtained copies of their sound samples,
> reproduced them using separate synchronized systems as per their
> efforts?
I have detailed the method of analysis in another post. Briefly, I
created two full-page images of the paper's graph's FRS and HCS that
are the same size. When I superimpose these printed images and hold
them up to the light, I find many areas of significant amplitude
difference in the 0-20 KHz range.
I also did an analysis of similar data that I prepared using a more
precise technique. My spectral analysis plots portrayed significantly
greater detail. These spectral analysis lines fell right on top of
each other almost entirely pixel-for pixel. Thus, I've shown that
this experiment can be done precisely and accurately with readily
available and reasonably-priced audio tools.
>If not, you are just reacting defensively
> (understandable, since the study badly undercuts many of your
> beliefs).
The study has enough flaws that it can't possibly be compared with
the accurate results that careful workers can obtain for themselves
using the .wav files they can freely download from
http://www.pcabx-pro.com/technical/sample_rates/index.htm .
<snip irrelevant discussion of audible folddown products that Harry
gratuitously added>
Denis Sbragion
"Harry Lavo" <harry...@rcn.com> wrote in
news:aqqnis$f4c$1...@bourbaki.localdomain:
...
> I'm not assuming it...it was designed out of the test. READ THE TEST.
> Here, I'll make it easy for you.....
>
> http://jn.physiology.org/cgi/content/full/83/6/3548
I've reread the paper with a closer look to the intermodulation problem.
Well, the authors say that intermodulation distortion have been taken
into account, but the procedure they describe is not enough to ensure
that no intermodulation problem was present. They simply say they used
two different drivers (a 2 way speaker for the audible frequency range
and a supertweeter for the ultrasonic signal), but they don't provide
evidence that there was no intermodulation distortion when both were
playing. Don't know if any intermodulation distortion may be produced by
this kind of setup, but a direct measure of it I think is mandatory in
this kind of experiment.
Furthermore I took a look at the power spectrums measured with the
supertweeter on and off (the FRS and HCS graphs in the first picture) as
suggested by Arny Krueger, and there are cleraly visible differences well
below the 20 Khz limit. May be these are just measuring or scanning
errors, but of course to test differences that are above 20 Khz in this
kind of setup you have to guarantee that anything below doesn't change in
the two situations. This is of paramount importance so even a measuring
or scanning error this big can't be accepted.
Anyway the subject is really interesting for me. If you have any
other reference on this subject please let me know. The other reference
to the inner intermodulation distortion of the ear itself reported few
message ago is also really interesting.
Regards,
P.S. In my previous msg I stated that intermodulation distortion in the
A/D system may be a problem as well. Considering the setup used I was
clearly wrong, any intermodulation in the A/D system simply becomes part
of the test signal.
--
Denis Sbragion
InfoTecna
Tel: +39 0362 805396, Fax: +39 0362 805404
URL: http://www.infotecna.it
Stewart Pinkerton
wrote:
>"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
>news:aqqf36$c3k$1...@bourbaki.localdomain...
>> On Mon, 11 Nov 2002 22:04:08 GMT, "Harry Lavo" <harry...@rcn.com>
>> wrote:
>> Confirming research, of course, is the key. Interesting that this
>> paper has *never* passed peer review by the AES, despite its great
>> interest to the audio engineering community - IF TRUE. The paper was
>> certainly *presented* at an AES convention more than ten years ago -
>> but never published by the AES - and it *should* have created great
>> excitement and lots of follow-on studies, especially now that wideband
>> distribution media are commercially available. However, back here in
>> the real world, there is absolutely *no* corroborating evidence - even
>> from Sony!
>>
>This is the first time this work has been published in a peer reveiw
>journal..only two years ago.
However, it was presented at an AES convention in 1991! It failed to
get through peer review at that time, and took eight years to find a
home in a field entirely unrelated to audio.
>replicated will take time...and often in science the test itself is not
>replicated...but its findings are corraborated as part of the results of
>additional studies exploring slightly different topics. That process has, I
>am sure, started.
It's had eleven years to 'get started', and plenty of pressure from
those pushing the commercial acceptance of wideband media such as SACD
and DVD-A - but there is *no* replication or corroboration.
>> The purpose is to improve the sound of reproduced music in the home,
>> without the aid of wallet-lightening voodooo and snake-oil products.
>>
>That's fine...but along the way "scientific orthodoxy" as propounded by its
>proponents has been put forth as the unassailable standard. And as usual
>the case with science, "orthodoxy" often has a way of evolving...
Indeed it has - which is why, as you yourself pointed out, we need
*confirming* evidence from other sources. It's been more than a decade
since that paper was presented - where's the beef?
>> You are now assuming that the 'HF' system produced no intermodulation
>> artifacts in the LF range. BIG assumption...................
>>
>I'm not assuming it...it was designed out of the test. READ THE TEST.
I already did.
>Here, I'll make it easy for you.....
>
>http://jn.physiology.org/cgi/content/full/83/6/3548
You seem to have completely missed the point. Where is the evidence
that the HF system does not produce artifacts in the range of the LF
system?
This paper is *very* old news Harry, and it has simply *not* been
verified by anyone else. With all the hooraw about SACD and DVD-A,
you'd think that *someone* would have confirmed their findings by now
- if indeed they were valid findings........
Andrew Haley
My thoughts, too. However, even if the HF system is producing
intermodulation products, the possibility that this apparatus is doing
something significant is still interesting, even if not for the
reasons mentioned.
Also, as mentioned, given that it is well known that the ear is
nonlinear and therefore produces its own intermodulation products,
perhaps some audibility of supersonic multitones at high amplitudes is
not very surprising.
Andrew.
Thomas A
I do believe that they tried that. However, peer reviewed articles
need not imply that they are "true". Many articles with so-called
astonishing results have been published and later shown to be caused
by flaws in the test setup or the intepretations. Since I do both
write and review articles in the field of Microbiology, I know that
this happens. So to be solid, the experiments need to be repeated by
an independent lab to be meaningful. The lack of articles showing a
negative result in this area may be due to that it is difficult to
publish negative results; however, that does not imply that this
article can be challenged since it is now published. But as far I as
know, this is the only group that has found this result.
If the intepretations are correct, another group should do it with a
different method, perhaps making DBTs in combination with mismatch
negativity response measurements of the brain. This method measures a
special signal in the brain which respond to changes in sound.
Thomas
Harry Lavo
Total ignorance of the facts. READ THE PAPER, Stewart. And not just (or
even) the abstract. No such assumption is made...in fact guarding against
it was done purposively and removing it via test variable was a part of the
test design. READ THE PAPER.
Howard Ferstler
(snips)
> I'm not assuming it...it was designed out of the test. READ THE TEST.
> Here, I'll make it easy for you.....
>
> http://jn.physiology.org/cgi/content/full/83/6/3548
I looked it over. Here are some problems.
First, how many recording microphones can respond high enough up (I
mean high enough up with any degree of flatness and strength) for this
to matter with typical program sources?
Second, how many loudspeakers can respond high enough up (I mean high
enough up with any degree of flatness and strength) for this to matter
with typical program sources? Most tweeters crap out at the ranges
discussed, and those that do not often have peaks above the standard
20 kHz point. It looks as if those tweeters would be distorting the
signals enough that high up to cause more problems than solutions - if
the paper is correct.
If the SACD and DVD-A recordings that seem to turn people on
(including a number of reissues made with microphones that are
anything but smooth and flat into the ultrasonic range) do not have
signals from microphones and speakers that are flat and smooth into
the ultrasonic range, then the findings of this paper (assuming they
are correct) become rather meaningless.
Going retro for a moment, how many LP cutting lathes can respond high
enough up for these signals to be present on the LP recordings that so
many subjectivists claim are superior to the compact disc for reasons
that involve ultrasonic frequencies? Those kinds of signals do not
exist on the LP record, and so even if this paper is on the mark it
does not explain the so-called superior sound many individuals claim
for the LP record. And how many lionized phonograph cartridges can
respond high enough up (I mean high enough up with any degree of
flatness and strength) for this to matter?
Finally, just how many conventional musical instruments can generate
signals appreciably strong enough into the ultrasonic range for this
to matter, even if the above criteria were satisfied?
This paper strikes me as a rather overdone hunt for a way to justify
the new SACD and DVD-A technologies. When listening to typical musical
sources for pleasure most enthusiasts would never, ever be able to
link up with the kind of nuances that this paper outlines - assuming
those nuances exist at all. That some individuals will latch on to it
as a way to justify having spend big bucks on borderline fantasy
technologies is no surprise.
Howard Ferstler
Nousaine
ArnyK wrote:>> For time immemorial, sound quality has been related to the
>> (conscious) enjoyment of music. Objectivists and subjectivists alike
>> have agreed on this point for years, if only this point.
>>
>
>And this paper and series of tests shows that sound recorded with
>frequencies chopped off fails to deliver the goods.
Interestingly, if we completely accept the papers findings improvement in
reproduced sound in your home would have to be assigned a 'future possible
improvements' status because high frequency transducers either in the
recordings chain or home loudspeakers with the required bandwidth haven't yet
become commonplace. I, for one, remain hopefull but skeptical.
One also has to recall that the special instrument (which seemed to be recorded
solo) is one of few that produces ultra high frequencies.
Bob Marcus
>
> After rereading the Oohashi et al article "Inaudible High-Frequency Sounds
> Affect Brain Activity: Hypersonic Effect " for a second time at my leisure
> last night (after posting the above link here on RAHE) I was struck how in
> one giant research leap, this effort (if supported by confirming research)
> of view expressed here on RAHE and challenges much of the dogma thrown out
> by the objectivist camp.
This strikes me as a bit premature, even with your proviso. Let's just
agree that it stirs the pot.
> If the purpose of that camp is to convert RAHE to
> objectivist hegemony, than a large boulder has just been placed smack dab in
> the middle of the road.
Um, a single, as-yet-unreplicated study has reached a remarkably
different conclusion. If it's replicated, and if the effects are
explained, then there's a lot of rewriting of science to be done.
Those are still big ifs, especially since these effects seem to run
counter to a lot of studies that HAVE been replicated.
> Everybody should read this article as soon as possible. The test basically
> was done using a three minute piece of Balinese gamelan music, rich in
> inaudible overtones reaching up to 50-100khz. The audible music and
> inaudible overtones were recorded and played back using a special recorder
> able to reach this high, and were presented using separate sound files and
> loudspeakers for the audible and inaudible parts, so there was no foldback
> effect or interaction. They were synced so as to appear as one source. And
> care was made in recording the inaudible overtones that no audible artifacts
> appeared. This was later confirmed by the fact that ambient noise and the
> presence of inaudible overtones recorded exactly the same brain response and
> psychological evaluation.
>
> The actual number of test subjects, which tests were given to what group,
> etc. are in the article and I will not attempt to summarize them hear. What
> is important is that the tests were double blind and the results had to be
> significant with less than 5% chance of being wrong (by chance). In
> reality, most of the results were significant at the 1% level. The four
> variables tested were music without overtones, music with overtones,
> overtones without music, and ambient noise. Physiological measurements were
> made using multiple location EEG measurements and blood flow measurements
> within the brain. Psychological questions were self-administered relating
> to the listeners state of well being, and to their perception of the music
> being played (when appropriate). Even when subjects were "wired", great
> care was exerted to keep the surroundings as pleasant and homelike as
> possible. A great deal of sophisticated statistical work was done to
> "normalize" the results.
>
> For those who have not already read the study, I will try to explain the
> basic findings in laymen's terms.
>
> 1) Music with inaudible overtones intact registered a far different reaction
> than either music without overtones, plain ambient noise, or overtones w/o
> audible music. Music with overtones was self-rated to sound different,
> sound richer, and be more relaxing) and physiological measurement showed an
> increase in brain activity and blood flow involving known "pleasure centers"
> and "mechanisms" of the brain.
If you say so, although the term "pleasure centers" appears nowhere in
the text.
>
> 2) Music without the overtones showed no difference from ambient background
> noise as far as brain activity was concerned.
This would be a shocking result, but I suspect it's a misprint. Don't
you mean that the "overtones," as you call them, were no different
from ambient background in their physiological effect?
>
> 3) Inaudible overtones by themselves did not register...they only registered
> when combined with the audible music as a complete bundle. By themselves
> they registered only the same as ambient background noise.
A surprising result, for which the researchers offered a few
hypotheses but no real explanation. This suggests that a lot more work
needs to be done.
>
> 4) It took awhile for the brain to react to the stimulus....the build up
> took "tens of seconds" (20-40?) to register, and a like time to unregister
> following a switch back to overtoneless music, ambient noise, or overtones
> without music.
This in the physiological testing.
>
> Now with that behind us, let's look at the implications for the debates
> raging here on RAHE. I will start with the most definitive and move on to
> the more speculative implications.
>
> 1) Do SACD and/or DVD-A really sound better (e.g.. more natural, "real-er")
> than CD or is that all in the imagination? The research clearly indicates
> (again, if eventually repeated and verified) that SACD and or DVD-A *SHOULD*
> sound better ("clearer","richer") and that CD's inherent frequency
> limitation turns out to be a very bad thing, perception-of-music-wise.
A possibility, though "clearly" is, again, a bit premature. Someone
who wanted to convey the real import of this article, rather than use
it to support his pre-existing point of view, would word this much
more carefully.
>
> 2) Can SACD and/or DVD-A really "relax" us more than CD? Again, the
> research pretty clearly states that this is a concomitant of wider bandwidth
> when reproducing music.
Please note that neither SACD nor DVD-A were used in this research. So
we really do not know what their effect would be.
>
> 3) Is Quick-Switch ABX testing a la Arnie the best way to evaluate music?
> The CCIR protocols oft cited by Arnie as the "bible" for quick-switch ABX
> testing are directly challenged by Oohashi and the nine other authors. They
> flatly suggest that the research conducted in the early eighties and given
> as support for the CD 22khz cutoff were wrong
"Flatly suggest" is an oxymoron. They note that their results differ
from earlier ones. A stronger statement would be premature (which is
why the researchers didn't make one).
>(because both the
> quick-switching and the short excerpts used "masked" the true reaction of
> the subjects by not allowing enough time for the brain to react to any but
> the audible portion of the signal..
You're taking two separate findings here and trying to smash them
together into one conclusion. Yes, there was a delay in brain
recognition of the HFC. And yes, the subjects tended to prefer music
with the HFC added. That doesn't mean there's a relationship between
the two.
>Oohashi et al devote a summary section
> specifically to this, which I've included as a footnote to this post.(1).
>
> 4) Is hearing music the same thing as hearing anything else? The evidence
> presented hear pretty conclusively suggests not. Music has its own "rules"
> for the human physiology. One can properly infer, therefore, that what may
> be appropriate for discerning audible artifacts may not *necessarily* be
> appropriate for discerning/evaluating music.
This is baseless. The researchers used only one recording. How can you
conclude anything about human reaction to different types of sounds
from a test using only one recording?
>
> 4) Is evaluating music in a relaxed sighted, state using "pleasure" as a
> measure of correctness an subjectivist audiophile fantasy? This test
> suggests that it may be a highly significant "indicator", especially of the
> system's ability to present the full overtone structure of music without
> distortion, attenuation, or time-smear.
Too bad the researchers didn't do the obvious: an ABX test using the
longer musical passages, to see if subjects could identify a
difference. Had they done so, this would have offered direct evidence
that the short snippet approach is flawed. (And had they gotten a
negative result, this would have called into question the results of
their preference test.)
>
> 5) Is their any support for the notion that evaluators should be able to
> control switching? Not directly, but indirectly the fact that it takes
> 20-40 seconds for a "pleasure state" to emerge and the old one to disappear
> suggests that users who know a piece, using their own music, on their own
> systems should be able to anticipate their response and evaluation of the
> upcoming segment of music, and compare the response and evaluation they
> actually get (from the other piece of gear switched in) against this fading
> "memory". But this will only work if they control the switching and are
> focused on the emotional and sound evaluation reaction of this familiar
> piece of music. This is obviously a hypothesis to be pursued, it certainly
> cannot be dismissed out of hand as in the past by many objectivists, based
> on this research presented in this article.
Interesting you should bring this up, given that this was a test in
which the subjects did NOT control switching. (But that doesn't mean
there's anything wrong with it!)
>
> 6) Can cables or amplifiers with similar electrical characteristics "sound
> different"? The scientists here hypothesized that the inaudible overtones
> of music may be sensed directly through the skull or other parts of the body
> (as in some other creatures). If this is true, then what happens in the
> inaudible high frequency part of the spectrum that affects proper
> reproduction of the sound (frequency deviation, time-smear, rfi
> interference, etc.) could conceivable affect sound, not by fold-back into
> the audible part of the spectrum, but directly.
Sure, if cables can produce these "hypersonic" effects. Can they?
>
> There are probably many more implications possible arising from this
> research; my guess it will become a *landmark* in the field of audio
> reproduction and hearing physiology.
And that's all it is at this point--a guess. (Nevertheless, a
fascinating article. Thanks for bringing it to our attention.)
bob
Steve G
> "Steve G" <ne...@stevethepsycho.co.uk> wrote in message
> news:aqpc41$t2v$1...@bourbaki.localdomain...
>
> >snip reply to Arny<
>
> > However, notwithstanding this, the effects could be explained by
> > nonlinearities in the ear. An auditory psychophysicist of my
> > acquaintance writes:
> >
> >
> > These distortion products were first discovered back in the 1800s,
> > BTW.
> >
>
> Yep, and that's why the scientists went as far as they did tto prevent any
> interatction that was not "natural'...if you read their test and
> reproduction setup carefully, you will see that.
I think you might be misunderstanding what I say? I'm not saying that
the effect uncovered is 'unnatural', but that it is probably a
concomitant of the ear's nonlinear (amplification) behaviour. This
nonlinear behaviour is well-established. It is - by definition -
'natural'!
BTW, these ear-based distortion products are absolutely independent of
any distortions in the sound-playing kit.
>
> > Some further comments:
> >
(...)
> >
> > - Or more simply: It's not subliminal!
>
> The discrimination tests were in the physiological measurements...clearly
> there was a significant difference (at <1% error) that their *WAS* a
> discrimination. But it was not aural, it was subliminal.
No. The authors showed significant differences in hedonic (pleasure)
*ratings*, as well as significant differences in cortical activity. As
I say, it is axiomatic that significant differences in ratings
absolutely *ensure* discriminative performance. Note that they did not
explicitly test discriminative performance either.
Incidentally, when carrying out sensory threshold tests, it is very
common for the subject to feel that they are behaving 'randomly'. This
is expected, since the sensory signals are on the edge of
perceptibility. If you asked those subjects if they were perfoming
better than chance, there's every chance they'd say 'no'; however they
do, and this performance is not 'subliminal'.
Note also - p<.01 is not synonymous with '<1% error'! Minor point.
(...)
> >
>
> Not necessarily restricted to music, but not generated by random noise,
> either. Neither the ambient background noise nor reproduction of the
> non-audible high-frequency overtones without an audible musical base
> provided any discimination or differences whatsoever.
They did not formally test discrimination performance, so we cannot
make any strong statements here.
Distortion products could (and can) be generated by (amusical)
'noise'. However, it is possible that these products are important in
musical enjoyment, as the cited paper may suggest.
> So it only happens
> when the brain "makes sense" of something and expects that "high frequency
> content" to be there as it occurs in nature.
Well, this is not true of distortion products - they happen
continually. However, the perceptions they lead to may only have
'meaning' in some circumstances, perhaps.
(...)
ome facetious soul might say that the authors have a vested interest
> > in finding (or emphasising the existence of) high-frequency effects,
> > given that they have links to Philips! A facetious soul, but not me.
> >
>
> Actually support and funding was listed by the authors:
And actually, I was not only facetious, but wrong! I might've been
thinking of the speaker system they had built by Pioneer! Never
produce a hasty reply at the end of the day, eh...
Steve.
Steve G
(...)
> I've previously read that article and also their former article
> regarding the "hypersonic" effect. I have several questions to this
> article. First, have anyone else repeated their experiments and if so
> do they get the same results?
This publication is quite recent, so few follow-ups are to be expected
yet. The closest approximation appears to be Fujioka et al (2002), who
studied evoked cortical magnetic fields (another measure of cortical
function) to single-frequency tones of up to 40kHz. No cortical
activity was demonstrated with ultrasounds. The authors suggest that
their delivery system involved only air-conducted sounds (i.e., no
bone-conducted sound), although I don't see how they guaranteed that.
Despite these results, they did not specifically disagree with Oohashi
et al.
> Second, is the high amount of HF unique to this kind of music/instruments?
No.
> Would there be an effect if they
> made the same experiment with orchestral music?
If the results hold, one would expect so.
> If so, why did they not show this with ordinary music instruments? That would
> have strengthen their conclusions.
What's normal?! I don't think it would strengthen their conclusions,
although pursuing the role of ear-based distortion products would;
indeed, they may explain the effect entirely, and would be consistent
with the findings of Fujioka et al, too.
> I've read that it is possible to hear
> the effects of HF just above the hearing limit since it can modulate
> the audible lower freqeuncies, IF the HF is high enough in level.
> Don't know if that is published though.
That's those damn distortion products again! There are references
going back to the late 1800s, though I'm not sure if any are available
online, without a subscription.
If I knew my tweeter could perform >20kHz, I would play around myself,
to see if I could obtain clear audible dist prods from one (or more)
inaudible primaries. However, if the tweeter doesn't have signficant
response way oop there, it'd be all a bit futile...!
Steve.
Fujioaka et al (2002). Neuroscience 112, 367-381.
Stewart Pinkerton
I read the paper. All they said was that the HF section on its own was
inaudible. Absolutely *zero* evidence was presented that the combined
signal contained no intermodulation artifacts. A *proper* scientific
study would have included a B&K microphone trace of the combined
output compared with the 'HF only' output.
This would be one reason why the study failed to pass peer review
after presentation to the AES in 1991.
Harry Lavo
news:aqrg9...@enews2.newsguy.com...
> "Harry Lavo" <harry...@rcn.com> wrote in message
news:<aqqnis$f4c$1...@bourbaki.localdomain>...
>
> (snips)
>
> > I'm not assuming it...it was designed out of the test. READ THE TEST.
> > Here, I'll make it easy for you.....
> >
> > http://jn.physiology.org/cgi/content/full/83/6/3548
>
> I looked it over. Here are some problems.
>
> First, how many recording microphones can respond high enough up (I
> mean high enough up with any degree of flatness and strength) for this
> to matter with typical program sources?
>
A B&K 2804 lab instrumentation mike, good out to about 50khz. Similar mikes
are often used for orchestral recordings
> Second, how many loudspeakers can respond high enough up (I mean high
> enough up with any degree of flatness and strength) for this to matter
> with typical program sources? Most tweeters crap out at the ranges
> discussed, and those that do not often have peaks above the standard
> 20 kHz point. It looks as if those tweeters would be distorting the
> signals enough that high up to cause more problems than solutions - if
> the paper is correct.
>
Seems to be a big move in the direction of more extended high frequency
response....and it is not at all unusual to have response well above 20khz.
> If the SACD and DVD-A recordings that seem to turn people on
> (including a number of reissues made with microphones that are
> anything but smooth and flat into the ultrasonic range) do not have
> signals from microphones and speakers that are flat and smooth into
> the ultrasonic range, then the findings of this paper (assuming they
> are correct) become rather meaningless.
>
Not at all. New recordings made with recent high-rez equipment (listen to
Allison Krauss, for example) sound absolutely spectacular, much better than
the older stuff. It's just that the older stuff sounded better on SACD
thatn on CD.
>
BTW, My 4 track 30ips Otari MX-12 goes out to 35Hz. High enough for you?
>
> Going retro for a moment, how many LP cutting lathes can respond high
> enough up for these signals to be present on the LP recordings that so
> many subjectivists claim are superior to the compact disc for reasons
> that involve ultrasonic frequencies? Those kinds of signals do not
> exist on the LP record, and so even if this paper is on the mark it
> does not explain the so-called superior sound many individuals claim
> for the LP record. And how many lionized phonograph cartridges can
> respond high enough up (I mean high enough up with any degree of
> flatness and strength) for this to matter?
>
This has been covered her ad infinitum
> Finally, just how many conventional musical instruments can generate
> signals appreciably strong enough into the ultrasonic range for this
> to matter, even if the above criteria were satisfied?
>
Who is to say how strong is strong enough? You?
>
> This paper strikes me as a rather overdone hunt for a way to justify
> the new SACD and DVD-A technologies. When listening to typical musical
> sources for pleasure most enthusiasts would never, ever be able to
> link up with the kind of nuances that this paper outlines - assuming
> those nuances exist at all. That some individuals will latch on to it
> as a way to justify having spend big bucks on borderline fantasy
> technologies is no surprise.
Well, if you think work started around 1989 was due to SACD, you may be
right, but I don't thing so.
Harry Lavo
I read your post, Arnie, and replied to it. The reason that they look
identical is that they are identical...they use the same audible frequency
file. Only one was tested without the inaudible frequencies and the other
without...and the difference in response was significant at the <1% level in
favor of the full-frequency cell. And the one without full-frequency got
the same response as ambient noise. Why do you keep trying to obfuscate
this fact. Your "variable" is not a "variable" it was tested away.
READ THE TEST!
> I also did an analysis of similar data that I prepared using a more
> precise technique. My spectral analysis plots portrayed significantly
> greater detail. These spectral analysis lines fell right on top of
> each other almost entirely pixel-for pixel. Thus, I've shown that
> this experiment can be done precisely and accurately with readily
> available and reasonably-priced audio tools.
>
> >If not, you are just reacting defensively
> > (understandable, since the study badly undercuts many of your
> > beliefs).
>
> The study has enough flaws that it can't possibly be compared with
> the accurate results that careful workers can obtain for themselves
> using the .wav files they can freely download from
> http://www.pcabx-pro.com/technical/sample_rates/index.htm .
>
> <snip irrelevant discussion of audible folddown products that Harry
> gratuitously added>
I want to respond b---sh-t but it probably won't be allowed. But I find
your assertion that your tests are better than this experiment laughable.
Steve G
> On Tue, 12 Nov 2002 11:11:03 GMT, "Harry Lavo" <harry...@rcn.com>
> wrote:
>
> >"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
> >news:aqqf36$c3k$1...@bourbaki.localdomain...
> >> On Mon, 11 Nov 2002 22:04:08 GMT, "Harry Lavo" <harry...@rcn.com>
> >> wrote:
>
> >>
> >This is the first time this work has been published in a peer reveiw
> >journal..only two years ago.
>
> However, it was presented at an AES convention in 1991! It failed to
> get through peer review at that time, and took eight years to find a
> home in a field entirely unrelated to audio.
Although now in a considerably more esteemed journal (impact factor J
Neuroscience ~3.5, impact factor JAES ~0.5), and one with much wider
circulation.
(...)
>
> It's had eleven years to 'get started', and plenty of pressure from
> those pushing the commercial acceptance of wideband media such as SACD
> and DVD-A - but there is *no* replication or corroboration.
Now it has been published in a less esoteric journal, people may
attempt to replicate. Indeed elsewhere in this thread I note a related
paper dealing with ultrasonic perception (or not!).
(...)
> You seem to have completely missed the point. Where is the evidence
> that the HF system does not produce artifacts in the range of the LF
> system?
Well, the driver systems appear unconnected, although I don't know if
this prevents 'physical' intermodulation?
Regardless, ear-based distortion products provide a mechanism for the
effects, if they truly exist. However, there is no convincing evidence
that the effect is 'subliminal', or indeed that 'subliminal' exists at
all in the layman's sense.
Another possible mechanism might include bone conduction.
There is also no evidence that the effect - if replicable - is limited
to musical material.
Note also that the FRS sound they refer to is not the FRS per se, but
HFC+LFC. The fact that this 'FRS'<>FRS (i.e., the real FRS) is not
important, though it does confuse a tad on a first read through.
Steve.
BEAR
> "Denis Sbragion" <d.sbr...@infotecna.it> wrote in message
> news:aqord3$li4$1...@bourbaki.localdomain
> > Hello Arny,
> >
> > "Arny Krueger" <ar...@hotpop.com> wrote in news:aqn9cs$ru$1
> > @bourbaki.localdomain:
> ....
> >> What is not known is the order and damping of this resonance.
> >> Depending on the actual parameters, there could be significant
> >> response variations below 20 KHz.
> >
> > furthermore in the Oohashi paper there's nothing about the non
> > linear properties of the speakers and the 1 bit recording system
> > used. May be the differences found are caused by distortion
> > folded in the audible range by the non linear behaviour of the
> > equipment used. Anyway, really an interesting paper.
>
> I think this point is very important. Key to the success of this
> experiment is controlling variables so that the two musical samples
> differed ONLY in that one had spectral content above 22-26 KHz. It is
> also important to show that the ENTIRE reproduction system was free
> of nonlinear distortion, particularly at high frequencies.
>
> Freedom from nonlinear distortion is very important because
> relatively small amounts of nonlinear distortion can cause distortion
> products that will be appear in the range where the human ear is
> known to be highly sensitive.
>
> In fact the http://jn.physiology.org/cgi/content/full/83/6/3548 paper
> shows that the two music samples differed significantly both above
> and below 22-28 KHz. There is no evidence showing that there weren't
> other differences as well. Furthermore, there is no objective
> evidence that the reproduction system was free of nonlinear
> distoriton.
Odd to hear this sort of objection coming from the learned Arny K!
Quote: "It is also important to show that the ENTIRE reproduction system was
free
of nonlinear distortion, particularly at high frequencies."
When in another discussion about 'flawed ABX tests' I suggested that one of the
problems with the published documentation proffered so far is that the
equipment itself has not been measured and quantified as being low in non-linear
distortion, or any other type of distortion for that matter, no one seemed to
agree!
This was just *one* of a multitude of issues that I raised. (and was ignored)
Which, in turn is just one factor in what I suggested would make up a SOTA
system -
a basic requirement for a definitive test...
_-_-bear
--
_-_-bearlabs
http://www.bearlabsUSA.com
- Silver Lightning Interconnects & Custom Mods -
Harry Lavo
> Hello Harry,
>
> "Harry Lavo" <harry...@rcn.com> wrote in
> news:aqqnis$f4c$1...@bourbaki.localdomain:
>
> ...
> > I'm not assuming it...it was designed out of the test. READ THE TEST.
> > Here, I'll make it easy for you.....
> >
> > http://jn.physiology.org/cgi/content/full/83/6/3548
>
> I've reread the paper with a closer look to the intermodulation problem.
> Well, the authors say that intermodulation distortion have been taken
> into account, but the procedure they describe is not enough to ensure
> that no intermodulation problem was present. They simply say they used
> two different drivers (a 2 way speaker for the audible frequency range
> and a supertweeter for the ultrasonic signal), but they don't provide
> evidence that there was no intermodulation distortion when both were
> playing. Don't know if any intermodulation distortion may be produced by
> this kind of setup, but a direct measure of it I think is mandatory in
> this kind of experiment.
The test setup they used should have eliminated any electro-mechanical
folddown into the audible frequency range. If their were "natural"
intermodulations from including the inaudible fequency material, that is
different. "Natural" is what we are looking for, I believe.
> Furthermore I took a look at the power spectrums measured with the
> supertweeter on and off (the FRS and HCS graphs in the first picture) as
> suggested by Arny Krueger, and there are cleraly visible differences well
> below the 20 Khz limit. May be these are just measuring or scanning
> errors, but of course to test differences that are above 20 Khz in this
> kind of setup you have to guarantee that anything below doesn't change in
> the two situations. This is of paramount importance so even a measuring
> or scanning error this big can't be accepted.
If they are different then they are natural, because the LFC component of
the test was exactly the same file, played through exactly the same
electronics and loudspeakers.
> Anyway the subject is really interesting for me. If you have any
> other reference on this subject please let me know. The other reference
> to the inner intermodulation distortion of the ear itself reported few
> message ago is also really interesting.
>
I find the whole evolving area of understanding aspects of how the brain
works fascinating. It have a simple example of it close at hand...my right
ear has pretty severe hearing loss above 6k. When listening with both ears,
I perceive the world as a binaural place with both ears conveying the same
information. In musical terms, doesn't matter where the cymbols are , I
hear them full frequency. Put a pillow over one ear, however, and I might
as well be deaf on one side. The brain compensates.
> Regards,
>
> P.S. In my previous msg I stated that intermodulation distortion in the
> A/D system may be a problem as well. Considering the setup used I was
> clearly wrong, any intermodulation in the A/D system simply becomes part
> of the test signal.
>
Regards yourself, and thanks for reading the article.
Stewart Pinkerton
wrote:
>"Howard Ferstler" <fer...@attglobal.net> wrote in message
>news:aqrg9...@enews2.newsguy.com...
>> "Harry Lavo" <harry...@rcn.com> wrote in message
>news:<aqqnis$f4c$1...@bourbaki.localdomain>...
>>
>> (snips)
Harry is assuming that there were no audible intermodulation products
present in the test, caused by the 'HF' playback chain.
>> > I'm not assuming it...it was designed out of the test.
No, it wasn't. As you say - read the test.
>> > READ THE TEST.
>> > Here, I'll make it easy for you.....
>> >
>> > http://jn.physiology.org/cgi/content/full/83/6/3548
>>
>> I looked it over. Here are some problems.
>>
>> First, how many recording microphones can respond high enough up (I
>> mean high enough up with any degree of flatness and strength) for this
>> to matter with typical program sources?
>>
>A B&K 2804 lab instrumentation mike, good out to about 50khz. Similar mikes
>are often used for orchestral recordings
Er no, not 'often', in fact it's *extremely* rare to use such a
microphone, as it has much higher self-noise than conventional
large-capsule studio microphones.
>> If the SACD and DVD-A recordings that seem to turn people on
>> (including a number of reissues made with microphones that are
>> anything but smooth and flat into the ultrasonic range) do not have
>> signals from microphones and speakers that are flat and smooth into
>> the ultrasonic range, then the findings of this paper (assuming they
>> are correct) become rather meaningless.
>>
>
>Not at all. New recordings made with recent high-rez equipment (listen to
>Allison Krauss, for example) sound absolutely spectacular, much better than
>the older stuff. It's just that the older stuff sounded better on SACD
>thatn on CD.
In your humble opinion, of course! Others report little or no
difference - where it's even possible to find the same mix on both
layers.....................
>BTW, My 4 track 30ips Otari MX-12 goes out to 35Hz. High enough for you?
That's not even properly into the bass, never mind the treble! :-)
>> Going retro for a moment, how many LP cutting lathes can respond high
>> enough up for these signals to be present on the LP recordings that so
>> many subjectivists claim are superior to the compact disc for reasons
>> that involve ultrasonic frequencies? Those kinds of signals do not
>> exist on the LP record, and so even if this paper is on the mark it
>> does not explain the so-called superior sound many individuals claim
>> for the LP record. And how many lionized phonograph cartridges can
>> respond high enough up (I mean high enough up with any degree of
>> flatness and strength) for this to matter?
>>
>This has been covered her ad infinitum
Quite so - and never answered by the vinyl fans..........
>> This paper strikes me as a rather overdone hunt for a way to justify
>> the new SACD and DVD-A technologies. When listening to typical musical
>> sources for pleasure most enthusiasts would never, ever be able to
>> link up with the kind of nuances that this paper outlines - assuming
>> those nuances exist at all. That some individuals will latch on to it
>> as a way to justify having spend big bucks on borderline fantasy
>> technologies is no surprise.
>
>Well, if you think work started around 1989 was due to SACD, you may be
>right, but I don't thing so.
Interesting that in the intervening 13 years, there has been *zero*
corroborating evidence from other sources. Especially interesting
given the *huge* commercial pressure from promoters of SACD and DVD-A.
Harry Lavo
news:aqrgh...@enews2.newsguy.com...
One of the few that produces them in abundance at such high overtones.
Obviously chosen to make sure the overtones were present in abundance. But
so what.
Harry Lavo
Absolutely, you won't find an argument from me here. From my very first
post i said (if verified by future testing) now, didn't I. But the test was
published two years ago. So to duplicate it would require dedication,
money, time and for what end. To debunk the findings..only Arnie has the
motivation to do that. I don't think it will happen. What will happen are
other studies that push the boundries further...if the test results hold up,
or will run across contra-indicating facts as a result and these will be
published in due course.
> If the intepretations are correct, another group should do it with a
> different method, perhaps making DBTs in combination with mismatch
> negativity response measurements of the brain. This method measures a
> special signal in the brain which respond to changes in sound.
I'm not sure I understand what you mean here, but I suspect the same caveats
apply that I expressed above.
At this point it is just one finding and therefore a very tentative "fact".
But it is a fact that has enormous implications.....and certainly it should
create a newfound sense and caution here in this forum about what is "known"
about hearing and music.
Haven't seen evidence that it has started to have that effect, however. :-)
Harry Lavo
news:aqrglc$qo5$1...@bourbaki.localdomain...
>
> "Harry Lavo" <harry...@rcn.com> wrote in
> news:aqqnis$f4c$1...@bourbaki.localdomain:
>
> ...
> > I'm not assuming it...it was designed out of the test. READ THE TEST.
> > Here, I'll make it easy for you.....
> >
> > http://jn.physiology.org/cgi/content/full/83/6/3548
>
> I've reread the paper with a closer look to the intermodulation problem.
> Well, the authors say that intermodulation distortion have been taken
> into account, but the procedure they describe is not enough to ensure
> that no intermodulation problem was present. They simply say they used
> two different drivers (a 2 way speaker for the audible frequency range
> and a supertweeter for the ultrasonic signal), but they don't provide
> evidence that there was no intermodulation distortion when both were
> playing. Don't know if any intermodulation distortion may be produced by
> this kind of setup, but a direct measure of it I think is mandatory in
> this kind of experiment.
If there was intermodulation distortion when both were playing, given the
care of their setup, then it was the intermodulation distortion of the
natural sound itself, heard full range. And that is the only plausible
explanation that would suggest why this effect created subjectively clearer
and richer sound and a relaxation effect on the subject.
> Furthermore I took a look at the power spectrums measured with the
> supertweeter on and off (the FRS and HCS graphs in the first picture) as
> suggested by Arny Krueger, and there are cleraly visible differences well
> below the 20 Khz limit. May be these are just measuring or scanning
> errors, but of course to test differences that are above 20 Khz in this
> kind of setup you have to guarantee that anything below doesn't change in
> the two situations. This is of paramount importance so even a measuring
> or scanning error this big can't be accepted.
They claim to have tested exactly that, and to have raised the cutoff for
the high frequency portion from 22hz to 26hz to prevent any audible
component in the audible range. I guess from your and Arnies comments it
would have been more convincing for them to have documented this better; my
guess is since the article was designed for a medical journal they felt it
was of lesser significance that the results attributable to the variables
under test.
> Anyway the subject is really interesting for me. If you have any
> other reference on this subject please let me know. The other reference
> to the inner intermodulation distortion of the ear itself reported few
> message ago is also really interesting.
>
I find much of the work being done on the physiological - psychological
links in brain development/function extremely fascinating.
Arny Krueger
> > I have detailed the method of analysis in another post. Briefly,
I
> > created two full-page images of the paper's graph's FRS and HCS
that
> > are the same size. When I superimpose these printed images and
hold
> > them up to the light, I find many areas of significant amplitude
> > difference in the 0-20 KHz range.
> I read your post, Arnie, and replied to it. The reason that they
look
> identical is that they are identical...they use the same audible
frequency
> file.
Harry, what's unclear about "I find many areas of significant
amplitude difference in the 0-20 KHz range."?
Harry, how does that relate to your response:
"The reason that they look identical is that they are
identical...they use the same audible frequency file."
I'm hoping that the THIRD time around you'll figure out that I'm
saying that they are different and NOT identical!
They should be identical but they AREN'T!
Arny Krueger
news:aqs9lj$732$1...@bourbaki.localdomain...
I can't imagine why. I've brought this point up again and again.
> Quote: "It is also important to show that the ENTIRE reproduction
system was free
> of nonlinear distortion, particularly at high frequencies."
> When in another discussion about 'flawed ABX tests' I suggested
that one of the
> problems with the published documentation proffered so far is that
the
> equipment itself has not been measured and quantified as being low
in non-linear
> distortion, or any other type of distortion for that matter, no one
seemed to
> agree!
I recall no such thing.
> This was just *one* of a multitude of issues that I raised. (and
was ignored)
Well, BEAR why don't you help me remember the specific post and
complete context?
> Which, in turn is just one factor in what I suggested would make up
a SOTA system -
> a basic requirement for a definitive test...
FWIW, the world seems to be full of equipment that has very low
nonlinear distortion up to at least 45 KHz. Regrettably there is also
a fair amount of equipment that is not.
Harry Lavo
news:aqrj75$rri$1...@bourbaki.localdomain...
> "Harry Lavo" <harry...@rcn.com> wrote in message
news:<aqp9f7$s2d$1...@bourbaki.localdomain>...
> >
> > After rereading the Oohashi et al article "Inaudible High-Frequency
Sounds
> > Affect Brain Activity: Hypersonic Effect " for a second time at my
leisure
> > last night (after posting the above link here on RAHE) I was struck how
in
> > one giant research leap, this effort (if supported by confirming
research)
> > can be reasonably by interpreted to support much of the subjectivist
points
> > of view expressed here on RAHE and challenges much of the dogma thrown
out
> > by the objectivist camp.
>
> This strikes me as a bit premature, even with your proviso. Let's just
> agree that it stirs the pot.
>
I'll accept that
> > If the purpose of that camp is to convert RAHE to
> > objectivist hegemony, than a large boulder has just been placed smack
dab in
> > the middle of the road.
>
> Um, a single, as-yet-unreplicated study has reached a remarkably
> different conclusion. If it's replicated, and if the effects are
> explained, then there's a lot of rewriting of science to be done.
> Those are still big ifs, especially since these effects seem to run
> counter to a lot of studies that HAVE been replicated.
>
Yes, I agree but I don't any have intetwined two disciplines and three
measurement devices quite like this one. I hope myself that it is either
supported or contra-indicated decisively in the next few years. But in
another post I explain that I doubt that will happen. I suspect the support
or erosion to be gradual and piecemeal instead.
Told you I was using layman terminology. The scientist did indicate quite
precisely, however, that both the ecc effects noted and the blood flow
concentration in the brain were effects and areas associated with pleasure
and emotion. In addition, subjects reported feeling more relaxed and that
the music sounded better...the scientists themselves pointed out the
congruence of these factorss. I used "pleasure centers' to try to convey
this in simplistic terms. Forgive me.
> > 2) Music without the overtones showed no difference from ambient
background
> > noise as far as brain activity was concerned.
>
> This would be a shocking result, but I suspect it's a misprint. Don't
> you mean that the "overtones," as you call them, were no different
> from ambient background in their physiological effect?
> >
Yes and no. The "overtones" as I call them were no different from ambient
background in their physiological effect when played (inaudibly) without
their musical base. BUT, the music played without the overtones ALSO showed
no difference from either ambient background or overtones in its
physiological effect. And this was radically different from the effect when
the music was played WITH its overtones. Look at the concentration graphs
shown in the accompaning data. This IS shocking. And it is why all the
obfuscation here about possible intermoducation effects makes not sense
except in the sense of perhaps adding back natural and anticipated
intermodulations as the result of the full frequency spectrum. The test
design, the equipment, and the presentation assured that there were no
artificial or equipment-created intermodulations.
> > 3) Inaudible overtones by themselves did not register...they only
registered
> > when combined with the audible music as a complete bundle. By
themselves
> > they registered only the same as ambient background noise.
>
> A surprising result, for which the researchers offered a few
> hypotheses but no real explanation. This suggests that a lot more work
> needs to be done.
Agree, one of the obvious areas for pursuit, as they noted. But not as
significant for music people as the absence of any significant physiological
elevation when music was played w/o its overtones.
> >
> > 4) It took awhile for the brain to react to the stimulus....the build up
> > took "tens of seconds" (20-40?) to register, and a like time to
unregister
> > following a switch back to overtoneless music, ambient noise, or
overtones
> > without music.
>
> This in the physiological testing.
Yep, but this physiological phenomenon is associated with emotional response
as the scientists point out. And how many times have you heard hear, or in
the reviews of the audio journals, that the evoked emotional response was
the true measure of the equipment. Makes for interesting speculation.
> >
> > Now with that behind us, let's look at the implications for the debates
> > raging here on RAHE. I will start with the most definitive and move on
to
> > the more speculative implications.
> >
> > 1) Do SACD and/or DVD-A really sound better (e.g.. more natural,
"real-er")
> > than CD or is that all in the imagination? The research clearly
indicates
> > (again, if eventually repeated and verified) that SACD and or DVD-A
*SHOULD*
> > sound better ("clearer","richer") and that CD's inherent frequency
> > limitation turns out to be a very bad thing, perception-of-music-wise.
>
> A possibility, though "clearly" is, again, a bit premature. Someone
> who wanted to convey the real import of this article, rather than use
> it to support his pre-existing point of view, would word this much
> more carefully.
> >
I don't think it is premature to say that if the research holds up (as I did
say) then these new formats SHOULD sound better. I didn't say that one
study by itself clearly established the phenomena or premise.
> > 2) Can SACD and/or DVD-A really "relax" us more than CD? Again, the
> > research pretty clearly states that this is a concomitant of wider
bandwidth
> > when reproducing music.
>
> Please note that neither SACD nor DVD-A were used in this research. So
> we really do not know what their effect would be.
I'll accept this; so noted. But the research did establish that
"relaxation" was a comcomitant of wider bandwidth. And since both SACD and
DVD-A have wider bandwidth than CD's, it is a logical deduction.
> >
> > 3) Is Quick-Switch ABX testing a la Arnie the best way to evaluate
music?
> > The CCIR protocols oft cited by Arnie as the "bible" for quick-switch
ABX
> > testing are directly challenged by Oohashi and the nine other authors.
They
> > flatly suggest that the research conducted in the early eighties and
given
> > as support for the CD 22khz cutoff were wrong
>
> "Flatly suggest" is an oxymoron. They note that their results differ
> from earlier ones. A stronger statement would be premature (which is
> why the researchers didn't make one).
>
>From the scientists summary conclusion "Explanation of the discrepancy
between the present and previous studies "
"......Muraoka et al. 1978; Plenge et al. 1979), which concluded that the
presence of sounds containing a frequency range above 15 kHz was not
recognized as making a difference in sound quality. The CCIR (1978), and the
current International Telecommunication Union-Radio communication sector
(ITU-R 1997), have recommended that sound samples used for the comparison of
sound quality should not last longer than 15-20 s (CCIR 1978; ITU-R 1997),
and that intervals between sound samples should be about 0.5-1 s (CCIR 1978)
because of short-term human memory limitations. Most of the previous
psychological experiments, including the studies by Muraoka et al. (1978)
and Plenge et al. (1979), were carried out using, essentially, the sound
presentation method recommended by the CCIR......
<separated for emphasis>
..... We also examined the psychological evaluation using the same material
and sound presentation system as was used for the present study, but
followed the presentation method recommended by the CCIR, and confirmed that
the results were in agreement with the studies by Muraoka et al. (1978) and
Plenge et al. (1979)"
In other words, using the same sound, equipment, and facilities...when the
comparison was reduced to snippets of sound and quick switching, there was
no difference. ..although their clearly was during the test. This is about
as direct a refutation as I can imagine.
> >(because both the
> > quick-switching and the short excerpts used "masked" the true reaction
of
> > the subjects by not allowing enough time for the brain to react to any
but
> > the audible portion of the signal..
>
> You're taking two separate findings here and trying to smash them
> together into one conclusion. Yes, there was a delay in brain
> recognition of the HFC. And yes, the subjects tended to prefer music
> with the HFC added. That doesn't mean there's a relationship between
> the two.
>
The author's clearly felt there was... the elevations all occured in ecc
measurement and blood flow areas of the brain associated with emotional
response and pleasure...totally congruent with the test subject's subjective
response.
> >Oohashi et al devote a summary section
> > specifically to this, which I've included as a footnote to this
post.(1).
> >
> > 4) Is hearing music the same thing as hearing anything else? The
evidence
> > presented hear pretty conclusively suggests not. Music has its own
"rules"
> > for the human physiology. One can properly infer, therefore, that what
may
> > be appropriate for discerning audible artifacts may not *necessarily* be
> > appropriate for discerning/evaluating music.
>
> This is baseless. The researchers used only one recording. How can you
> conclude anything about human reaction to different types of sounds
> from a test using only one recording?
> >
Okay, overdrawn on my part, perhaps, but certainly not baseless. As I
indicated, the links to subjectivist philosophy get more tenuous as we go
down the list.
> > 4) Is evaluating music in a relaxed sighted, state using "pleasure" as a
> > measure of correctness an subjectivist audiophile fantasy? This test
> > suggests that it may be a highly significant "indicator", especially of
the
> > system's ability to present the full overtone structure of music without
> > distortion, attenuation, or time-smear.
>
> Too bad the researchers didn't do the obvious: an ABX test using the
> longer musical passages, to see if subjects could identify a
> difference. Had they done so, this would have offered direct evidence
> that the short snippet approach is flawed. (And had they gotten a
> negative result, this would have called into question the results of
> their preference test.)
They got rating preferences in a dbt significant at the <1% level for
quality of the sound and relaxation when listening? That doesn't count?
> >
> > 5) Is their any support for the notion that evaluators should be able to
> > control switching? Not directly, but indirectly the fact that it takes
> > 20-40 seconds for a "pleasure state" to emerge and the old one to
disappear
> > suggests that users who know a piece, using their own music, on their
own
> > systems should be able to anticipate their response and evaluation of
the
> > upcoming segment of music, and compare the response and evaluation they
> > actually get (from the other piece of gear switched in) against this
fading
> > "memory". But this will only work if they control the switching and are
> > focused on the emotional and sound evaluation reaction of this familiar
> > piece of music. This is obviously a hypothesis to be pursued, it
certainly
> > cannot be dismissed out of hand as in the past by many objectivists,
based
> > on this research presented in this article.
>
> Interesting you should bring this up, given that this was a test in
> which the subjects did NOT control switching. (But that doesn't mean
> there's anything wrong with it!)
> >
This was a double-blind monadic-rating test...and the author's made a point
of saying they used the full three minute piece of music so there was no
switching.
> > 6) Can cables or amplifiers with similar electrical characteristics
"sound
> > different"? The scientists here hypothesized that the inaudible
overtones
> > of music may be sensed directly through the skull or other parts of the
body
> > (as in some other creatures). If this is true, then what happens in the
> > inaudible high frequency part of the spectrum that affects proper
> > reproduction of the sound (frequency deviation, time-smear, rfi
> > interference, etc.) could conceivable affect sound, not by fold-back
into
> > the audible part of the spectrum, but directly.
>
> Sure, if cables can produce these "hypersonic" effects. Can they?
> >
Don't know that they produce them, but most can pass them if produced
elsewhere in the system or by RFI infiltration.
> > There are probably many more implications possible arising from this
> > research; my guess it will become a *landmark* in the field of audio
> > reproduction and hearing physiology.
>
> And that's all it is at this point--a guess. (Nevertheless, a
> fascinating article. Thanks for bringing it to our attention.)
That's what I said, a guess :-)
>
> bob
>
Thanks for reading the test so carefully.
Harry Lavo
> "Harry Lavo" <harry...@rcn.com> wrote in message
news:<aqpf7...@enews4.newsguy.com>...
> > "Steve G" <ne...@stevethepsycho.co.uk> wrote in message
> > news:aqpc41$t2v$1...@bourbaki.localdomain...
> >
> > >snip reply to Arny<
> >
> > > However, notwithstanding this, the effects could be explained by
> > > nonlinearities in the ear. An auditory psychophysicist of my
> > > acquaintance writes:
> > >
> (...)
> > >
> > > These distortion products were first discovered back in the 1800s,
> > > BTW.
> > >
> >
> > Yep, and that's why the scientists went as far as they did tto prevent
any
> > interatction that was not "natural'...if you read their test and
> > reproduction setup carefully, you will see that.
>
> I think you might be misunderstanding what I say? I'm not saying that
> the effect uncovered is 'unnatural', but that it is probably a
> concomitant of the ear's nonlinear (amplification) behaviour. This
> nonlinear behaviour is well-established. It is - by definition -
> 'natural'!
>
> BTW, these ear-based distortion products are absolutely independent of
> any distortions in the sound-playing kit.
>
Then we are in agreement. My point now made clear in another post, was that
they took steps to eliminate the "non-natural" distortion.
> >
> > > Some further comments:
> > >
> (...)
> > >
> > > - Or more simply: It's not subliminal!
> >
> > The discrimination tests were in the physiological
measurements...clearly
> > there was a significant difference (at <1% error) that their *WAS* a
> > discrimination. But it was not aural, it was subliminal.
>
> No. The authors showed significant differences in hedonic (pleasure)
> *ratings*, as well as significant differences in cortical activity. As
> I say, it is axiomatic that significant differences in ratings
> absolutely *ensure* discriminative performance. Note that they did not
> explicitly test discriminative performance either.
>
I guess I've reached the limits of my statistical competence. I stand
corrected.
> Incidentally, when carrying out sensory threshold tests, it is very
> common for the subject to feel that they are behaving 'randomly'. This
> is expected, since the sensory signals are on the edge of
> perceptibility. If you asked those subjects if they were perfoming
> better than chance, there's every chance they'd say 'no'; however they
> do, and this performance is not 'subliminal'.
>
> Note also - p<.01 is not synonymous with '<1% error'! Minor point.
And I'm making mistakes even within the area of my statistical competence!
:-)
> >
> > Not necessarily restricted to music, but not generated by random noise,
> > either. Neither the ambient background noise nor reproduction of the
> > non-audible high-frequency overtones without an audible musical base
> > provided any discimination or differences whatsoever.
>
> They did not formally test discrimination performance, so we cannot
> make any strong statements here.
>
> Distortion products could (and can) be generated by (amusical)
> 'noise'. However, it is possible that these products are important in
> musical enjoyment, as the cited paper may suggest.
>
Nobody knows the mechanism at this point.
> > So it only happens
> > when the brain "makes sense" of something and expects that "high
frequency
> > content" to be there as it occurs in nature.
>
> Well, this is not true of distortion products - they happen
> continually. However, the perceptions they lead to may only have
> 'meaning' in some circumstances, perhaps.
>
That is my interpretation. The brain is an integrator...sound (or
distortion) only makes sense within a context. Without that context it is
just background.
> (...)
> ome facetious soul might say that the authors have a vested interest
> > > in finding (or emphasising the existence of) high-frequency effects,
> > > given that they have links to Philips! A facetious soul, but not me.
> > >
> >
> > Actually support and funding was listed by the authors:
>
> And actually, I was not only facetious, but wrong! I might've been
> thinking of the speaker system they had built by Pioneer! Never
> produce a hasty reply at the end of the day, eh...
>
> Steve.
>
Good advice I often find hard to follow......... :-)
Thomas A
Is it required to have ultrasonic sound to demonstrate this effect?
Ok, this is what I've read (summarized), and I do not understand this
fully and have not been able to track down any references. If two
tones of 18000 and 18500 Hz are produces simultaneously there is a
possibility to create IM dist with tones of 500 Hz if the sound
travels through a non-linear medium. Air is supposed to be linear, if
I understand it correctly. But walls and furniture create
non-linearites. (Also what happens to the linearites of air when there
is acoustical pressure from other frequencies?) If the microphones in
a recording studio have captured some but not all of this putiatve
distortion (because of damping panels etc), then this would need to be
re-created in the listening environment. If so (many ifs) then
ultrasonic frequencies of instruments *should* be produced to get the
natural sound. The third *if* is whether these ultrasonic frequencies
travels far in air enough to hit any non-linear medium and yield any
putative IM dist at audible levels.
Lastly, there is this discussion of SACD vs DVD-A. In SACD, I suspect
that most of the natural HF of musical instruments would be buried in
the ultransonic noise created by the SACD medium itself.
Thomas
Nousaine
snip....
>The test setup they used should have eliminated any electro-mechanical
>folddown into the audible frequency range. If their were "natural"
>intermodulations from including the inaudible fequency material, that is
>different. "Natural" is what we are looking for, I believe.
Actually that's not the case. They used a separate device to radiate the
hypersonic signal into the listening space.
Arny Krueger
news:aqslgo$bve$1...@bourbaki.localdomain...
Arny Krueger" <ar...@hotpop.com> wrote in message
news:aqove...@enews4.newsguy.com...
> Key to the success of this test would be a lack of changes to the
> "low pass" file other than removal of signals above 22-26 KHz.
Actually the file didn't change, but how it was reproduced changed
when the HFCs were added back in.
> With a simple graphics utility (I used Paintshop Pro 4) it is
> possible to get two full-page images that are the same size. I
> printed them out. When I superimpose these printed images and hold
> them up to the light, I find many areas of significant amplitude
> difference in the 0-20 KHz range.
> IOW I disagree with the following claim in the paper:
> "The spectrum of FRS was essentially the same as that of the source
> and contained both LFCs below and HFCs above 22 kHz."
"Denis Sbragion" <d.sbr...@infotecna.it> wrote in message
news:aqrglc$qo5$1...@bourbaki.localdomain...
> > Furthermore I took a look at the power spectrums measured
with the
> > supertweeter on and off (the FRS and HCS graphs in the first
picture) as
> > suggested by Arny Krueger, and there are clearly visible
differences well
> > below the 20 KHz limit. May be these are just measuring or
scanning
> > errors, but of course to test differences that are above 20 KHz
in this
> > kind of setup you have to guarantee that anything below doesn't
change in
> > the two situations. This is of paramount importance so even a
measuring
> > or scanning error this big can't be accepted.
> If they are different then they are natural, because the LFC
component of
> the test was exactly the same file, played through exactly the same
> electronics and loudspeakers.
Right, but as two independent analyses of the spectral plots provided
with the paper now show, the net effect was to alter the spectral
content of the LFC portion of the spectrum when the HFCs were added
back in.
BTW, I didn't come up with this idea in a vacuum, the possibility was
pointed out to me by Stanley Lipshitz some months ago.
Harry Lavo
> On Wed, 13 Nov 2002 00:38:31 GMT, "Harry Lavo" <harry...@rcn.com>
> wrote:
>
> >"Howard Ferstler" <fer...@attglobal.net> wrote in message
> >news:aqrg9...@enews2.newsguy.com...
> >> "Harry Lavo" <harry...@rcn.com> wrote in message
> >news:<aqqnis$f4c$1...@bourbaki.localdomain>...
> >>
> >> (snips)
>
> Harry is assuming that there were no audible intermodulation products
> present in the test, caused by the 'HF' playback chain.
>
> >> > I'm not assuming it...it was designed out of the test.
>
> No, it wasn't. As you say - read the test.
>
I have, others have. And it seems you are becoming a minority POV. Any
intermodulation in this test is "natural" as the result of the high
frequencies, not an artifact.
> >> > READ THE TEST.
> >> > Here, I'll make it easy for you.....
> >> >
> >> > http://jn.physiology.org/cgi/content/full/83/6/3548
> >>
> >> I looked it over. Here are some problems.
> >>
> >> First, how many recording microphones can respond high enough up (I
> >> mean high enough up with any degree of flatness and strength) for this
> >> to matter with typical program sources?
> >>
> >A B&K 2804 lab instrumentation mike, good out to about 50khz. Similar
mikes
> >are often used for orchestral recordings
>
> Er no, not 'often', in fact it's *extremely* rare to use such a
> microphone, as it has much higher self-noise than conventional
> large-capsule studio microphones.
>
The key here is the word "similar". B&K makes a commercial variation of
this mic, as you well know, that is used for orchestral recording. It does
have lower self noise and a to my understanding a rougher and somewhat less
extended high-end
> >> If the SACD and DVD-A recordings that seem to turn people on
> >> (including a number of reissues made with microphones that are
> >> anything but smooth and flat into the ultrasonic range) do not have
> >> signals from microphones and speakers that are flat and smooth into
> >> the ultrasonic range, then the findings of this paper (assuming they
> >> are correct) become rather meaningless.
> >>
> >
> >Not at all. New recordings made with recent high-rez equipment (listen
to
> >Allison Krauss, for example) sound absolutely spectacular, much better
than
> >the older stuff. It's just that the older stuff sounded better on SACD
> >thatn on CD.
>
> In your humble opinion, of course! Others report little or no
> difference - where it's even possible to find the same mix on both
> layers.....................
Little difference on the best, but still a difference.
>
> >BTW, My 4 track 30ips Otari MX-12 goes out to 35Hz. High enough for you?
>
> That's not even properly into the bass, never mind the treble! :-)
>
Ouch! :-)
> >> Going retro for a moment, how many LP cutting lathes can respond high
> >> enough up for these signals to be present on the LP recordings that so
> >> many subjectivists claim are superior to the compact disc for reasons
> >> that involve ultrasonic frequencies? Those kinds of signals do not
> >> exist on the LP record, and so even if this paper is on the mark it
> >> does not explain the so-called superior sound many individuals claim
> >> for the LP record. And how many lionized phonograph cartridges can
> >> respond high enough up (I mean high enough up with any degree of
> >> flatness and strength) for this to matter?
> >>
> >This has been covered her ad infinitum
>
> Quite so - and never answered by the vinyl fans..........
>
You don't think some moving coil cartridges with line-contact type stylii go
out to 40-50khz?
> >> This paper strikes me as a rather overdone hunt for a way to justify
> >> the new SACD and DVD-A technologies. When listening to typical musical
> >> sources for pleasure most enthusiasts would never, ever be able to
> >> link up with the kind of nuances that this paper outlines - assuming
> >> those nuances exist at all. That some individuals will latch on to it
> >> as a way to justify having spend big bucks on borderline fantasy
> >> technologies is no surprise.
> >
> >Well, if you think work started around 1989 was due to SACD, you may be
> >right, but I don't thing so.
>
> Interesting that in the intervening 13 years, there has been *zero*
> corroborating evidence from other sources. Especially interesting
> given the *huge* commercial pressure from promoters of SACD and DVD-A.
The article has been published for two years, Stewart. As you well know.
You want to join Arnie in the "guilt by innuendo" camp, shame on you. :-(
Mark Ovchain
from those paragons of mechanically generated raucousness, E, L, and P.
BEAR <bear...@netzero.net> wrote in message news:<aqs9lj$732$1...@bourbaki.localdomain>...
Around and around and around we go, and where we stop, nobody
knows!
> This was just *one* of a multitude of issues that I raised. (and was ignored)
>
> Which, in turn is just one factor in what I suggested would make up a SOTA
> system -
> a basic requirement for a definitive test...
Why then, oh ursine one, will you not tell us what kind of requirements
one would have to meet to have an "SOTA" system.
Would an inexpensive (in terms of high-end SOTA systems) system that was
flat, clean, and didn't even require equalization to function in its
carefully chosen room qualify for SOTA? Would a system that demonstrates
performance extremely close to the known limits of hearing suffice, even if
it did not use solid-unobtanium (thank you, Dick Pierce) cables and all
firebottle eletronics, with FET power-supply regulators and silver mains
cable?
Would one need a monoblock using an EIMAC 5667 in class A?
Would an EIMAC 8973 suffice?
Could we get by with an 833A?
Would the filament power have to be DC? How would we protect from
metal migration if it must be so?
How can we test and falsicate the hypothesis that a set of equipment
is SOTA? What are the standards? What experiment will demonstrate
the SOTA-ness of a setup?
Do you imagine that SOTA systems will do better with the ultrasonic
distortion problem mentioned above? Would it even be proper to
test an SOTA system to see if it has such a problem? How should
one test an SOTA system to see if it SOTA? What would you say about
a system consisting of solid-state amplification connected to a
set of electrostatic headphones? Must I use an 'oversampled' DAC
or an 'upsampled' DAC? Must I use only "analog" signal sources?
How must I arrange for those analog sources to repeat?
My correspondents tell me that this discussion has gone before, and
an answer was not forthcoming. Perhaps we shall see if you've
come up with one in the meanwhile.
In short, how do you, I, and the rest of the world reach a solid,
common, and scientifically supported conclusion that a given system
is "SOTA"?
Denis Sbragion
"Harry Lavo" <harry...@rcn.com> wrote in
news:aqu0v...@enews3.newsguy.com:
...
> audible component in the audible range. I guess from your and Arnies
> comments it would have been more convincing for them to have
> documented this better; my guess is since the article was designed for
> a medical journal they felt it was of lesser significance that the
> results attributable to the variables under test.
...
yep, it's of course one of the most important thing, and they have not
documented it enough. I'm looking forward to see if someone tries to
replicate these results, of course with greater care to all these important
details. Anyway, I think the use of PET and EEG is a great idea. No
subjective evaluation problems, provided of course that the subjects don't
know what they're listening to.
Regards,
Howard Ferstler
> I find the whole evolving area of understanding aspects of how the brain
> works fascinating. It have a simple example of it close at hand...my right
> ear has pretty severe hearing loss above 6k. When listening with both ears,
> I perceive the world as a binaural place with both ears conveying the same
> information. In musical terms, doesn't matter where the cymbols are , I
> hear them full frequency. Put a pillow over one ear, however, and I might
> as well be deaf on one side. The brain compensates.
It certainly does. No doubt, there are individuals out there who can
hear clearly to 18 kHz and higher who probably could not care less
about response nuances when listening to any kind of audio gear. They
would be just as happy with a Bose package system as with a serious
upscale system. On the other hand, there are probably guys who have
some degree of hearing loss, but who also have years of serious audio
experience, who would be able to quickly pinpoint audio artifacts that
went right by the guy who had sharper ears. The brain does indeed
compensate - to an extent.
However, we are not talking about listening vs hearing in this
situation. We are talking about some kind of mysterious ability to
hear ultrasonic frequencies with musical source material, even though
standard audiology tests might show an individual to have serious
hearing losses in the top two octaves below 20 kHz.
It looks to me as if some individuals might be using the paper as a
way to justify their golden-ear status, even though standard hearing
tests indicate that they are not really all that blessed with sharp
hearing.
At nearly 60 years of age, I am anything but a golden ear (although,
unlike you I am able to hear flat with both ears to well past 6 kHz),
and I realize that any hi-fi perceptive abilities I have involve my
experience and not some kind of mystical ability to detect ultrasonic
frequencies.
Howard Ferstler
Stewart Pinkerton
wrote:
>"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
>news:aqt3kt$h1h$1...@bourbaki.localdomain...
>> On Wed, 13 Nov 2002 00:38:31 GMT, "Harry Lavo" <harry...@rcn.com>
>> wrote:
>>
>> >"Howard Ferstler" <fer...@attglobal.net> wrote in message
>> >news:aqrg9...@enews2.newsguy.com...
>> >> "Harry Lavo" <harry...@rcn.com> wrote in message
>> >news:<aqqnis$f4c$1...@bourbaki.localdomain>...
>> >>
>> >> (snips)
>>
>> Harry is assuming that there were no audible intermodulation products
>> present in the test, caused by the 'HF' playback chain.
>>
>> >> > I'm not assuming it...it was designed out of the test.
>>
>> No, it wasn't. As you say - read the test.
>>
>
>I have, others have. And it seems you are becoming a minority POV. Any
>intermodulation in this test is "natural" as the result of the high
>frequencies, not an artifact.
No Harry, you are once again working on 'edited higlights'. The HF
replay system has *not* been shown to have no IM artifacts, indeed the
'FRS' plot is *significantly* different from the 'LF' plot in the 'LF'
region, which is *highly* suspicious.
>> >> First, how many recording microphones can respond high enough up (I
>> >> mean high enough up with any degree of flatness and strength) for this
>> >> to matter with typical program sources?
>> >>
>> >A B&K 2804 lab instrumentation mike, good out to about 50khz. Similar mikes
>> >are often used for orchestral recordings
>>
>> Er no, not 'often', in fact it's *extremely* rare to use such a
>> microphone, as it has much higher self-noise than conventional
>> large-capsule studio microphones.
>>
>The key here is the word "similar". B&K makes a commercial variation of
>this mic, as you well know, that is used for orchestral recording. It does
>have lower self noise and a to my understanding a rougher and somewhat less
>extended high-end
Sorry Harry, but you are *at best* being disingenuous here. The mic
with lower self-noise has a lower bandwidth. I canna change the laws
o' physics, cap'n!
That means that they are *not* similar in the parameter of interest,
which is bandwidth.
>You don't think some moving coil cartridges with line-contact type stylii go
>out to 40-50khz?
Sure I do - but where are the *albums* with response above 15kHz?
>> >> This paper strikes me as a rather overdone hunt for a way to justify
>> >> the new SACD and DVD-A technologies. When listening to typical musical
>> >> sources for pleasure most enthusiasts would never, ever be able to
>> >> link up with the kind of nuances that this paper outlines - assuming
>> >> those nuances exist at all. That some individuals will latch on to it
>> >> as a way to justify having spend big bucks on borderline fantasy
>> >> technologies is no surprise.
>> >
>> >Well, if you think work started around 1989 was due to SACD, you may be
>> >right, but I don't thing so.
>>
>> Interesting that in the intervening 13 years, there has been *zero*
>> corroborating evidence from other sources. Especially interesting
>> given the *huge* commercial pressure from promoters of SACD and DVD-A.
>
>The article has been published for two years, Stewart. As you well know.
>You want to join Arnie in the "guilt by innuendo" camp, shame on you. :-(
Harry, that paper was presented to an AES convention in 1991. It
failed peer review for publication in the AES journal, and the authors
took *nine years* to finally find acceptance in a journal totally
unrelated to audio. You know this, you are choosing to ignore it. Who
is being dishonest?
BEAR
> "BEAR" <bear...@netzero.net> wrote in message
> news:aqs9lj$732$1...@bourbaki.localdomain...
> > Arny Krueger wrote:
> >
> > > "Denis Sbragion" <d.sbr...@infotecna.it> wrote in message
> > > news:aqord3$li4$1...@bourbaki.localdomain
> > > > Hello Arny,
> > > >
> > > > "Arny Krueger" <ar...@hotpop.com> wrote in news:aqn9cs$ru$1
> > > > @bourbaki.localdomain:
> > > ....
>
<Snipp>
>
>
> I recall no such thing.
>
> > This was just *one* of a multitude of issues that I raised. (and
> was ignored)
>
> Well, BEAR why don't you help me remember the specific post and
> complete context?
You may recall a few months back a long and detailed exposition on "SOTA"
equipment? No? Perhaps you have posted too many times, and can't recall many
as they all blur together?
>
>
> > Which, in turn is just one factor in what I suggested would make up
> a SOTA system -
> > a basic requirement for a definitive test...
>
> FWIW, the world seems to be full of equipment that has very low
> nonlinear distortion up to at least 45 KHz. Regrettably there is also
> a fair amount of equipment that is not.
Really? Do tell. What do you specifically define "low nonlinear distortion" as
being?
What speakers do you feel meet this spec up to 45kHz.??
Harry Lavo
So you don't think this would happen in nature? Pray, what mechanism is
creating this "unatural intermodulation effect"...given their test setup.
>
> BTW, I didn't come up with this idea in a vacuum, the possibility was
> pointed out to me by Stanley Lipshitz some months ago.
>
Ah, Dr. Lipshitz of "debunk SACD" fame, I presume. Nice to have authority
figures on your side, right Arny?
Arny Krueger
It obviously happened in nature. However it happened, its a gross
error in the experiment.
>Pray, what mechanism is creating this "unatural intermodulation
effect"...given their test setup.
Harry who are you quoting here? Yourself? I never said that!
It doesn't matter these differences are being created. They represent
an error in the experiment. Key to any experiment is controlling all
relevant variables. Obviously frequency response in the LFC range
slipped out of their control.
> > BTW, I didn't come up with this idea in a vacuum, the possibility
was
> > pointed out to me by Stanley Lipshitz some months ago.
> Ah, Dr. Lipshitz of "debunk SACD" fame, I presume. Nice to have
authority
> figures on your side, right Arny?
Well it is true that I appear to have the JAES editorial review board
on my side in this issue...
Arny Krueger
> > FWIW, the world seems to be full of equipment that has very low
> > nonlinear distortion up to at least 45 KHz. Regrettably there is
also
> > a fair amount of equipment that is not.
> Really? Do tell. What do you specifically define "low nonlinear
distortion" as
> being?
Under 0.1% IM seems low enough.
> What speakers do you feel meet this spec up to 45kHz.??
In speakers, maintaining low nonlinear distortion at high frequencies
is pretty easy since most distortion mechanisms are stroke-related.
At high frequencies, Xmax requirements are small.
I've had my eye on the Vifa TX-25TG30-04
http://www.partsexpress.com/pdf/264-555.pdf
My measurements show that Vifa is if anything understating its
capabilities, at least in terms of frequency response.
Howard Ferstler
>
> "Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
> news:aqt3kt$h1h$1...@bourbaki.localdomain...
> > Howard wrote:
> > >> Going retro for a moment, how many LP cutting lathes can respond high
> > >> enough up for these signals to be present on the LP recordings that so
> > >> many subjectivists claim are superior to the compact disc for reasons
> > >> that involve ultrasonic frequencies? Those kinds of signals do not
> > >> exist on the LP record, and so even if this paper is on the mark it
> > >> does not explain the so-called superior sound many individuals claim
> > >> for the LP record. And how many lionized phonograph cartridges can
> > >> respond high enough up (I mean high enough up with any degree of
> > >> flatness and strength) for this to matter?
> > >This has been covered her ad infinitum
> > Quite so - and never answered by the vinyl fans..........
> You don't think some moving coil cartridges with line-contact type stylii go
> out to 40-50khz?
Out to 40-50 kHz flat and smooth? Is there a way to validate
this? Also, just kind of record wear are we talking about at
40-50 kHz? How many plays can an LP endure before those
signals are either erased or fully screwed up?
Howard Ferstler
Harry Lavo
The LFC files are identical. Any intermodulation effects are from the
SEPARATE addition (through separate sound file, amplifier chain, and
speakers) of the high frequency content. Thus any difference is due to the
natural intermodulation of the sound...as it occurs in nature. That is not
a test flaw...it is part of the experiment.
Keep in mind Arny that the purpose of the test is to determine the
importance of the presence of inaudible signals (extended high frequency
response, if you will) not to determine if those frequencies can't interact
in nature with those lower down. The elevated physiological functions seem
to indicate it is important and positive.
Harry Lavo
> On Wed, 13 Nov 2002 18:48:45 GMT, "Harry Lavo" <harry...@rcn.com>
> wrote:
>
> >"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
> >news:aqt3kt$h1h$1...@bourbaki.localdomain...
> >> On Wed, 13 Nov 2002 00:38:31 GMT, "Harry Lavo" <harry...@rcn.com>
> >> wrote:
> >>
> >> >"Howard Ferstler" <fer...@attglobal.net> wrote in message
> >> >news:aqrg9...@enews2.newsguy.com...
> >> >> "Harry Lavo" <harry...@rcn.com> wrote in message
> >> >news:<aqqnis$f4c$1...@bourbaki.localdomain>...
> >> >>
> >> >> (snips)
> >>
> >> Harry is assuming that there were no audible intermodulation products
> >> present in the test, caused by the 'HF' playback chain.
> >>
> >> >> > I'm not assuming it...it was designed out of the test.
> >>
> >> No, it wasn't. As you say - read the test.
> >>
> >
> >I have, others have. And it seems you are becoming a minority POV. Any
> >intermodulation in this test is "natural" as the result of the high
> >frequencies, not an artifact.
>
> No Harry, you are once again working on 'edited higlights'. The HF
> replay system has *not* been shown to have no IM artifacts, indeed the
> 'FRS' plot is *significantly* different from the 'LF' plot in the 'LF'
> region, which is *highly* suspicious.
>
No the LCF portion was exactly the same. When the high frequency portion
was added via a separate, synced system, it may have slightly changed the
lower as would happen in nature. And that is what Arnie is cavilling about.
But it is irrelevant (except as an explanation of what might have been
heard) to the fundamental conclusion of the study -- that inclusion of
inaudible higher frequencies was signifcantly important to
perception/enjoyment of music.
> >> >> First, how many recording microphones can respond high enough up (I
> >> >> mean high enough up with any degree of flatness and strength) for
this
> >> >> to matter with typical program sources?
> >> >>
> >> >A B&K 2804 lab instrumentation mike, good out to about 50khz. Similar
mikes
> >> >are often used for orchestral recordings
> >>
> >> Er no, not 'often', in fact it's *extremely* rare to use such a
> >> microphone, as it has much higher self-noise than conventional
> >> large-capsule studio microphones.
> >>
> >The key here is the word "similar". B&K makes a commercial variation of
> >this mic, as you well know, that is used for orchestral recording. It
does
> >have lower self noise and a to my understanding a rougher and somewhat
less
> >extended high-end
>
> Sorry Harry, but you are *at best* being disingenuous here. The mic
> with lower self-noise has a lower bandwidth. I canna change the laws
> o' physics, cap'n!
>
You do understand that "somewhat less extended high-end" translates
approximately to "lower bandwidth" do you not? Cmon, Pinkie, even Lords
know that!.
> That means that they are *not* similar in the parameter of interest,
> which is bandwidth.
>
As I acknowledged, but since their is no way of knowing how important each
octave of inaudible sound is, I think we can assume the lower portion
(perhaps 20khz-35khz ?? where most of them are in much music is probably
most relevant. So perhaps extension to 35khz, for example, accounts for
95% of any possible impact. I assume some future research will deal with
this. Until then, we can only guess how much is enough.
> >You don't think some moving coil cartridges with line-contact type stylii
go
> >out to 40-50khz?
>
> Sure I do - but where are the *albums* with response above 15kHz?
>
There are plenty...and even on the ones that don't extend much beyond that,
pehaps the gradual physical attenuation of an extended recording has a
different effect than an abruptly cut-off electrical "wall". At least the
tested "wall" showed *no* positive emotional impact.
> >> >> This paper strikes me as a rather overdone hunt for a way to justify
> >> >> the new SACD and DVD-A technologies. When listening to typical
musical
> >> >> sources for pleasure most enthusiasts would never, ever be able to
> >> >> link up with the kind of nuances that this paper outlines - assuming
> >> >> those nuances exist at all. That some individuals will latch on to
it
> >> >> as a way to justify having spend big bucks on borderline fantasy
> >> >> technologies is no surprise.
> >> >
> >> >Well, if you think work started around 1989 was due to SACD, you may
be
> >> >right, but I don't thing so.
> >>
> >> Interesting that in the intervening 13 years, there has been *zero*
> >> corroborating evidence from other sources. Especially interesting
> >> given the *huge* commercial pressure from promoters of SACD and DVD-A.
> >
> >The article has been published for two years, Stewart. As you well know.
> >You want to join Arnie in the "guilt by innuendo" camp, shame on you.
:-(
>
> Harry, that paper was presented to an AES convention in 1991. It
> failed peer review for publication in the AES journal, and the authors
> took *nine years* to finally find acceptance in a journal totally
> unrelated to audio. You know this, you are choosing to ignore it. Who
> is being dishonest?
You know for a fact that their prior paper *failed* peer review? Even Arny
didn't claim to know. All you know is that it was not published for
whatever reason, as I pointed out elsewhere.
Moreover, it was a different paper with more than twice as many
contributors, additional tests, and more sophisticated analysis. And it is
geared at a more sophisticated audience and one with no axe to grind. Your
really think the biggies behind the AES in 1991 wanted a paper that said
that CD was fundamentally flawed?
Harry Lavo
> "Harry Lavo" <harry...@rcn.com> wrote in message
news:<aqslgo$bve$1...@bourbaki.localdomain>...
>
> > I find the whole evolving area of understanding aspects of how the brain
> > works fascinating. It have a simple example of it close at hand...my
right
> > ear has pretty severe hearing loss above 6k. When listening with both
ears,
> > I perceive the world as a binaural place with both ears conveying the
same
> > hear them full frequency. Put a pillow over one ear, however, and I
might
> > as well be deaf on one side. The brain compensates.
>
> It certainly does. No doubt, there are individuals out there who can
> hear clearly to 18 kHz and higher who probably could not care less
> about response nuances when listening to any kind of audio gear. They
> would be just as happy with a Bose package system as with a serious
> upscale system. On the other hand, there are probably guys who have
> some degree of hearing loss, but who also have years of serious audio
> experience, who would be able to quickly pinpoint audio artifacts that
> went right by the guy who had sharper ears. The brain does indeed
> compensate - to an extent.
>
> However, we are not talking about listening vs hearing in this
> situation. We are talking about some kind of mysterious ability to
> hear ultrasonic frequencies with musical source material, even though
> standard audiology tests might show an individual to have serious
> hearing losses in the top two octaves below 20 kHz.
>
The studies subjects fell between the ages of 18-43, with most of them
between 19 and 34. I think we can assume they had pretty good to excellent
hearing. Why are you imputing that they had hearing loss? As for the rest
of us, what we can hear of the inaudible higher frequencies will depend on
how they are transmitted, which is unknow. If it turns out to be some
mechanism other than the ears traditional function, then we may or may not
be able to hear it with age, and or the brain may/may not be able to "fake
it" from only fragmentary data. If the brain can reconstruct cymbals for
me, I expect it may be able to do the latter as well.
> It looks to me as if some individuals might be using the paper as a
> way to justify their golden-ear status, even though standard hearing
> tests indicate that they are not really all that blessed with sharp
> hearing.
>
Which individuals are those?
I perceive this as an inuendo aimed at me. I have never claimed to have a
golden ear. i have claimed to have been raised with the highest of
commercial "fi" since I was eight years of age(1), to have sung my way
through grade school and high school, to have been a lifelong attendee of
live concerts of jazz, folk, and orchestral music, and to have been trained
at the Institute for Audio Reseach under John Woram and to have spent a half
dozen years doing professional recording. So I think I have a pretty good
reference for judging most music reproduction and the ins and outs of the
production process, from musicians and their instruments, to the halls or
clubs they perform in, to the microphones and media and techniques used to
record them to the systems used to reproduce them in a home environment from
one, two,four, and five channels. Morover my first wife and my two best
friends were professional musicians who made their living at it. We did a
lot of listening, both comparative, and for pleasure.
> At nearly 60 years of age, I am anything but a golden ear (although,
> unlike you I am able to hear flat with both ears to well past 6 kHz),
> and I realize that any hi-fi perceptive abilities I have involve my
> experience and not some kind of mystical ability to detect ultrasonic
> frequencies.
>
Good for you. That's all I claim as well. But perhaps we perceive those
ultrasonic frequencies despite ourselves. Mystery for now. Perhaps highly
explainable in the future.
(1) My dad built his own equipment, was a jazz drummer, and built a seven
store chain of audio salons and electronics parts stores in the late
forties/early fifties. Doctors and Lawyers and others wealthy enough to
afford the gear drove hundreds of miles to buy from him, and he and I and
others drove hundreds of miles to install systems in their homes. He and
Emerson Cook were friends and I grew up with tape recorders, Newcomb Triodes
and JBL corner horns, Rek-0-Kut and Presto turntables with Norelco
cartridges, and audiophile records...and tapes...some professionally
recorded by my father (his studio had direct to disk lathes and we used and
were distributors for Magnecord in the Northeastern United States.....yes in
the late 40's). The company name was Alvo Electronics...and its 1954 and 55
catalogs are a treasured reference for me to this day. I just picked up a
Newcomb amp on ebay and in my old age plan to put together a reasonable
facsimile of that first system. I built my first system in high school,
took it to college, upgraded it there, and converted to stereo in 1963 using
both tape and records.
Bob Marcus
>
> Lastly, there is this discussion of SACD vs DVD-A. In SACD, I suspect
> that most of the natural HF of musical instruments would be buried in
> the ultransonic noise created by the SACD medium itself.
>
system (a la SACD, with a lower sampling rate). I don't know how this
system deals with noise, but if it's similar to SACD, it uses noise
shaping, which moves the noise into the "hypersonic" band. If that's
the case, then a fair bit of the HFC in this test is noise, not
"overtones," as Harry calls them, with all kinds of implications for
the results.
Seems to me it would have been better if they'd done this experiment
using PCM with a high sampling rate instead.
boob
Bob Marcus
It's not a refutation. It's a different result using a different
method. One of those methods is wrong, but we don't know which one
yet.
Also, we don't know which of the several parameters that differed
between the tests (length of sample, frequency of switching,
preference vs. identification) is responsible for the difference in
result, if in fact this test holds up.
bob
Harry Lavo
Why do you continue to be so obtuse? Their experiment wasn't specifically
to see if inaudible high frquencies could/could not be heard aurally...it
was to determine if they somehow made a difference in the appreciateion of
music. They proved concludsively that they did. They admit they don't know
the mechanism as yet.
> >Pray, what mechanism is creating this "unatural intermodulation
> effect"...given their test setup.
>
> Harry who are you quoting here? Yourself? I never said that!
>
It isn't a quote. it is putting a name to it. The only way this would be a
problem for the experiment would be if it was created by the equipment...it
wasn't. So if you are cavilling about intermodulation, the only legitimate
cavilling would be about equipment-derived effects - thus "unatural
intermodulation effect".
> It doesn't matter these differences are being created. They represent
> an error in the experiment. Key to any experiment is controlling all
> relevant variables. Obviously frequency response in the LFC range
> slipped out of their control.
>
They weren't trying to control it except not to have the equipment add it.
> > > BTW, I didn't come up with this idea in a vacuum, the possibility
> was
> > > pointed out to me by Stanley Lipshitz some months ago.
>
> > Ah, Dr. Lipshitz of "debunk SACD" fame, I presume. Nice to have
> authority
> > figures on your side, right Arny?
>
> Well it is true that I appear to have the JAES editorial review board
> on my side in this issue...
>
You've polled them, Arny? You've tape recorded them? Perhaps you'd like to
give us their discussion and vote?
Harry Lavo
Did I say flat and smooth? (hint: no) Do we know how flat and smooth they
have to be? (second hint: no)
Take a look at old IAR issues, where the frequency responses were published.
And more recent cartridges have gotten better.
As to how many plays, line-contact stylii and more sophisticated resonance
control have greatly reduced wear, even at extremely high frequencies.
Cheers,
Harry
Steve G
> "Steve G" <ne...@stevethepsycho.co.uk> wrote in message
> news:aqro98$toc$1...@bourbaki.localdomain...
> > "Harry Lavo" <harry...@rcn.com> wrote in message
> news:<aqpf7...@enews4.newsguy.com>...
> > > "Steve G" <ne...@stevethepsycho.co.uk> wrote in message
> > > news:aqpc41$t2v$1...@bourbaki.localdomain...
> > >
> >
> > any distortions in the sound-playing kit.
> >
>
> Then we are in agreement. My point now made clear in another post, was that
> they took steps to eliminate the "non-natural" distortion.
They did, although if Fig 1 is an accurate representation of what was
heard, they have added more problems for interpretation since the HCS
component of the HCS+LCS appears to be different to the HCS component
when played alone.
(...)
> >
> > Note also - p<.01 is not synonymous with '<1% error'! Minor point.
>
> And I'm making mistakes even within the area of my statistical competence!
> :-)
Well, if by '1% error' you mean 'p<.01 says that a difference of the
magnitude found in the experiment - or greater - would occur 1% of the
time by chance alone, given no actual performance differences' ... you
would be correct. Typing '1% error' take less effort...!
(...)
> >
> > Distortion products could (and can) be generated by (amusical)
> > 'noise'. However, it is possible that these products are important in
> > musical enjoyment, as the cited paper may suggest.
> >
>
> Nobody knows the mechanism at this point.
If it's ear-based dist prods, then we know the 'mechanism' by which
the HF sounds might influence sounds in the audible range. However -
yes - we don't know when the mechanism is 'important'.
Incidentally, even if the effects found are explicable in terms of IM
effects in the *playback* system, the paper remains interesting - both
in the effects found upon brain function by small modulations in the
audible signal, and in which frequencies contribute to these effects.
Steve.
Steve G
(...)
>
> Is it required to have ultrasonic sound to demonstrate this effect?
Not for ear-based distortion products.
> Ok, this is what I've read (summarized), and I do not understand this
> fully and have not been able to track down any references. If two
> tones of 18000 and 18500 Hz are produces simultaneously there is a
> possibility to create IM dist with tones of 500 Hz if the sound
> travels through a non-linear medium. (...)
These are not the dist prods of which I speak! The chaps I'm speaking
of occur in your ear, i.e., you hear tones which did not actually
exist in the external sound field! These dist prods are completely
independent of any other IM distortion, such as may be found in the
playback equipment.
Other contributors in this thread are discussing the type of IM dist
you mention above.
Steve.
KikeG
> >news:aqt3kt$h1h$1...@bourbaki.localdomain...
> replay system has *not* been shown to have no IM artifacts, indeed the
> 'FRS' plot is *significantly* different from the 'LF' plot in the 'LF'
> region, which is *highly* suspicious.
But if the HF replay system had IM artifacts that fell into the
audible range, they would have been heard when playing just the HFC
signal. And according to the paper, when just the HFC signal was
played, it couldn't be tell apart from just ambient noise, so it was
inaudible. So there were no IM products due to the HF replay system
that feel into the audible range.
Also, the HFC and the LFC were played through totally separate signal
and transduction paths, so the only possibility of interaction between
both signals is on the air or the ear. I'm I missing something?
BEAR
<snipped>
Whomever you are "Mark O"...
>
>
> Why then, oh ursine one, will you not tell us what kind of requirements
> one would have to meet to have an "SOTA" system.
Again?
Where have you been? Not reading rahe?
Let's use simple LOGIC.
What does "SOTA" stand for? Oh? You know: State-Of-The-Art.
Note it is not State Of The Science, btw... small detail.
But, let's apply that difficult to fathom logic. Ok. First we have to do a
survey of
extant equipment. For the benefit of simplicity for now we can limit ourselves
to
just commercially available equipment. Then we run the survey through a spread
sheet and see which equipment appears to have the best specifications. From this
shorter list we then actually TEST said equipment to see if it meets the
published
specs. THEN we run it through some more vigorous tests (since the published
specs
don't usually include all sorts of things like CCIR type IM tests, or tests with
difficult
loads, etc...). THEN we check to see if some of the more "highly regarded"
pieces from
subjectivist sources are in the list, if not we add them in too. Then we test
some more.
Now we cull this through our spread sheet and come up with a handful of
candidates.
Iterate a bit more, adding any other more creative tests we can come up with...
and
THEN listen. Listen blind, half blind, half drunk, sleeping, awake, late, early,
in
different combinations... analyze, DBT, ABX, ABS, whatever... (...it is your
SOTA list.)
From this we then come up with a short list that then represents what is the SOTA at the
time of the test, given the equipment tested. Clearly this is a substantial
undertaking,
since each and every piece of hardware type and category needs to be tested and
analyzed.
In otherwords, SOTA is self defining as the *best available* today.
>
>
> Would an inexpensive (in terms of high-end SOTA systems) system that was
> flat, clean, and didn't even require equalization to function in its
> carefully chosen room qualify for SOTA?
What do those words mean? Nice words, but not terribly easy to define or
quantify. How can anyone replicate "flat, clean and carefully chosen in its
room??"
In essence this is meaningless and totally subjective.
Expense, btw, is a red herring - I don't care one way or the other what
something costs
with respect to this discussion. Inexpensive is just fine.
> Would a system that demonstrates
> performance extremely close to the known limits of hearing suffice, even if
> it did not use solid-unobtanium (thank you, Dick Pierce) cables and all
> firebottle eletronics, with FET power-supply regulators and silver mains
> cable?
That would depend upon your definition of the "known limits of hearing." And, of
course
how you managed to *parameterize* and quantify that so that it could be
*replicated*
and tested at another time and place.
And, afaik, Dick did not coin the term "unobtainium."
>
>
> Would one need a monoblock using an EIMAC 5667 in class A?
> Would an EIMAC 8973 suffice?
> Could we get by with an 833A?
Are you kidding? Do you really think I don't know what transmitting tubes are?
Have you ever used any in real life? I have.
>
>
> Would the filament power have to be DC? How would we protect from
> metal migration if it must be so?
Are you using a Directly or Indirectly heated cathode, mon ami?
>
>
> How can we test and falsicate the hypothesis that a set of equipment
> is SOTA? What are the standards? What experiment will demonstrate
> the SOTA-ness of a setup?
The set-up is another issue. For now we can start with defining parameters
which the *hardware* has to meet. So far no one has stepped forward to
even suggest these details. Why don't you? You seem to know everything.
What *experiment*?? Gee... I dunno.
What do you want to experiment with?
The whole point of the SOTA discussion *was* that ur not likely to hear certain
things, if the gear is not up to the task. One way to be certain that you've got
a good
*shot* at being able to reproduce *everything* that *can* be reproduced today is
to have the *best* equipment and set up possible. And that is, ummm, by
definition
what SOTA *is*, yes?
Seems incredibly straightforward and simple to me. Why is this difficult to
follow?
>
>
> Do you imagine that SOTA systems will do better with the ultrasonic
> distortion problem mentioned above?
Huh??
Are you looking to chum the waters? Do we have source that reaches into
the ultrasonic? If so, then the SOTA must by definition INCLUDE that issue.
> Would it even be proper to
> test an SOTA system to see if it has such a problem?
Better question, wise guy, is: has any prior ABX or DBT tested the system or
any PART of the system for any sort or type of non-linear distortion? IF so,
please
cite.
> How should
> one test an SOTA system to see if it SOTA?
Now, that is a good question. Since it is still very unclear that by testing all
the
gear in the world and then picking the "best" from those tests, and hooking it
up
that the thing will sound worth a damn. Will it? What is your thought on that?
> What would you say about
> a system consisting of solid-state amplification connected to a
> set of electrostatic headphones?
Interesting... how do you know that the headphones when connected to
your head provide a flat response? How do you measure said ESL headphones
for non-linear distortion? What about the "difference" caused by the lack of
interaural path?
> Must I use an 'oversampled' DAC
> or an 'upsampled' DAC? Must I use only "analog" signal sources?
> How must I arrange for those analog sources to repeat?
Gee... questions, questions... how about answering the question about ABX
and non-linear distortion? That would be useful.
So far, you've steadfastly avoided that.
>
>
> My correspondents tell me that this discussion has gone before, and
> an answer was not forthcoming. Perhaps we shall see if you've
> come up with one in the meanwhile.
So, you are involved in a deliberate effort to coordinate with others in order
to
control and manipulate this forum and this discussion? I see. Hmmm...
>
>
> In short, how do you, I, and the rest of the world reach a solid,
> common, and scientifically supported conclusion that a given system
> is "SOTA"?
Start by seeing what is the best you can find in extant commercial gear, and
come up with that list based upon the *best* efforts that you can make to *test*
said gear in a meaningful way. I have said this so many times it is getting
boring
and stupid.
How to put said gear into a room and make the ROOM sound good, or SOTA is
another
topic for another day, please. Actually PLEASE start your own thread on this
subject
and espouse your point of view, since you seem to be so incredibly knowledgeable
about
these things - perhaps you can help us all to do a better job in our own rooms.
Yes?
I shall be anxiously awaiting your exposition on this topic.
Regards,
Arny Krueger
news:ar0k9...@enews2.newsguy.com>> Lastly, there is this discussion of SACD vs. DVD-A. In SACD, I
>> suspect that most of the natural HF of musical instruments would
>> itself.
>>
> This is not a trivial point. The recording system used was a one-
Right, the details on this part of the experiment were sketchy, at
best.
>I don't know
> how this system deals with noise, but if it's similar to SACD, it
> uses noise shaping, which moves the noise into the "hypersonic"
> band. If that's the case, then a fair bit of the HFC in this test
> is noise, not "overtones," as Harry calls them, with all kinds of
> implications for the results.
Again, what we don't know isn't very comforting.
> Seems to me it would have been better if they'd done this
> experiment using PCM with a high sampling rate instead.
Agreed. But PCM can (and probably should) use noise-shaping, too.
Arny Krueger
> "Arny Krueger" <ar...@hotpop.com> wrote in message
> news:aqu11...@enews3.newsguy.com...
>> "Harry Lavo" <harry...@rcn.com> wrote in message
>> news:aqs7ih$5lr$1...@bourbaki.localdomain...
>>
>>>> I have detailed the method of analysis in another post.
>>>> Briefly, I created two full-page images of the paper's graph's
>>>> FRS and HCS that are the same size. When I superimpose these
>>>> printed images and hold them up to the light, I find many
>>>> areas of significant amplitude difference in the 0-20 KHz
>>>> range.
>>
>>> I read your post, Arnie, and replied to it. The reason that
>>> they look identical is that they are identical...they use the
>>> same audible frequency file.
>>
>> Harry, what's unclear about "I find many areas of significant
>> amplitude difference in the 0-20 KHz range."?
>>
>> Harry, how does that relate to your response:
>>
>> "The reason that they look identical is that they are
>> identical...they use the same audible frequency file."
>>
>> I'm hoping that the THIRD time around you'll figure out that I'm
>> saying that they are different and NOT identical!
>> They should be identical but they AREN'T!
> The LFC files are identical.
Well, they started out that way...
> Any intermodulation effects are
> from the SEPARATE addition (through separate sound file,
> amplifier chain, and speakers) of the high frequency content.
Who says the problem could only be intermodulation effects?
> Thus any difference is due to the natural intermodulation of the
> sound...as it occurs in nature. That is not a test flaw...it is
> part of the experiment.
"Natural intermodulation"? Can you cite a refereed paper in a
professional journal explaining this?
> Keep in mind Arny that the purpose of the test is to determine the
> importance of the presence of inaudible signals (extended high
> frequency response, if you will) not to determine if those
> frequencies can't interact in nature with those lower down.
We know that answer already. At anything like reasonable listening
levels, intermodulation like that doesn't happen.
> The elevated physiological functions seem to indicate it is
important
> and positive.
Looks to me like there are some kind of artifacts of their test rig
that show up in spectrograms. Could be linear, could be nonlinear. If
the authors provided a complete set of test results, we could
probably tell. But they didn't, so all we know that people were
hearing different stuff below 20 KHz.
OK, so hearing different stuff below 20 KHz makes people respond
differently. Where's the news in that?
Arny Krueger
>> Harry, that paper was presented to an AES convention in 1991. It
>> failed peer review for publication in the AES journal, and the
>> authors took *nine years* to finally find acceptance in a
>> journal totally unrelated to audio. You know this, you are
>> choosing to ignore it. Who is being dishonest?
> You know for a fact that their prior paper *failed* peer review?
We know for a fact that it didn't pass peer review until the ca. 2000
publication.
> Even Arny didn't claim to know. All you know is that it was not
> published for whatever reason, as I pointed out elsewhere.
It was obviously a paper that people were interested in. I think its
safe to say that they fluffed and puffed it until they found someone,
someplace to publish it in a peer-reviewed journal.
> Moreover, it was a different paper with more than twice as many
> contributors, additional tests, and more sophisticated analysis.
As I showed, the basic experiment was the same, some of the authors
were the same, and the words in the later paper were often
paraphrases of the earlier paper. The paraphrasing was so extensive
that had some of the same authors been listed, it would have been
plagiarism.
Looking at the paper again, it looks to me like it may have been
handed off to a second team when the first lost team lost interest in
it, or went on to bigger and better things.
> And it is geared at a more sophisticated audience and one with no
> axe to grind.
That's both the good news and the bad news, right? It's kinda strange
when a paper cites lots of articles that were cited in the JAES, but
could never muster the required approvals so it could be published in
the JAES. Given that the paper was initially given to an AES
conference, the natural place for its publication would have been the
JAES, right?
>Your really think the biggies behind the AES in
> 1991 wanted a paper that said that CD was fundamentally flawed?
I think that such a paper would be very commercially rewarding for
the whole audio industry. Since the AES leadership is drawn from the
audio industry, they would have been beneficiaries as well.
I don't think that proper credit has been extended to the AES for
sticking to their principles when so much that could profit them was
at stake.
The paper also contradicts itself in a way. It claims that these
brain wave changes are why people don't like music that has been
low-pass filtered, but it doesn't explain why people can't bring
these likes and dislikes to the point of conscious expression. Seems
like some kind of long term blind test would provide similar results.
Harry Lavo
Even should you be correct, there is the little matter that CD-type
frequency response doesn't move people, physiologically verified, versus
more full range reproduction. That not significant enough for you?
Harry Lavo
> "Harry Lavo" <harry...@rcn.com> wrote in message
> news:ar0jq...@enews2.newsguy.com
>
> >> Harry, that paper was presented to an AES convention in 1991. It
> >> failed peer review for publication in the AES journal, and the
> >> authors took *nine years* to finally find acceptance in a
> >> journal totally unrelated to audio. You know this, you are
> >> choosing to ignore it. Who is being dishonest?
>
> > You know for a fact that their prior paper *failed* peer review?
>
> We know for a fact that it didn't pass peer review until the ca. 2000
> publication.
>
No, Arny, the early paper was not submitted to the Journal of
Neuropsychology. The later and expanded and reworked study was submitted to
the Journal of neuropsychology. Neither you nor I know when or why. You
keep trying to imply the study was submitted and rejected and it is pure
speculation.
> > Even Arny didn't claim to know. All you know is that it was not
> > published for whatever reason, as I pointed out elsewhere.
>
> It was obviously a paper that people were interested in. I think its
> safe to say that they fluffed and puffed it until they found someone,
> someplace to publish it in a peer-reviewed journal.
>
I don't think it is safe to say that at all. You have no idea what
transpired. I'll now say what I wanted to say above but held back....you
are engaged in intellectual dishonesty and slander to protect your beliefs.
it doesn't become you.
> > Moreover, it was a different paper with more than twice as many
> > contributors, additional tests, and more sophisticated analysis.
>
> As I showed, the basic experiment was the same, some of the authors
> were the same, and the words in the later paper were often
> paraphrases of the earlier paper. The paraphrasing was so extensive
> that had some of the same authors been listed, it would have been
> plagiarism.
>
That still doesn't make it the same article. And you ignore the fact that
more and different supporting tests were run, using different measures, and
more and different statistical analysis was done of the results.
> Looking at the paper again, it looks to me like it may have been
> handed off to a second team when the first lost team lost interest in
> it, or went on to bigger and better things.
>
Again, Arny.....totally out of line. You are beginning to outdo Howard in
speculating what others think and what their motivations are.
> > And it is geared at a more sophisticated audience and one with no
> > axe to grind.
>
> That's both the good news and the bad news, right? It's kinda strange
> when a paper cites lots of articles that were cited in the JAES, but
> could never muster the required approvals so it could be published in
> the JAES. Given that the paper was initially given to an AES
> conference, the natural place for its publication would have been the
> JAES, right?
>
Depends on whether it was written as a general interest paper for the JAES.
As I said before, it was done by medical scientists and it may simply may
not have been an appropriate article for the journal. Since you say the
language is similar, then you must realize that it is load with medical
terminology and is not geared towards audio engineers or audiophiles. I've
point that out before...you choose to ignore it and cast your inuendos. Now
we see the real scientist in you at work, right Arny?
> >Your really think the biggies behind the AES in
> > 1991 wanted a paper that said that CD was fundamentally flawed?
>
> I think that such a paper would be very commercially rewarding for
> the whole audio industry. Since the AES leadership is drawn from the
> audio industry, they would have been beneficiaries as well.
>
> I don't think that proper credit has been extended to the AES for
> sticking to their principles when so much that could profit them was
> at stake.
>
I'd like to think so, but I've never been part of any industry organization
that didn't have its biases and politics. Perhaps you are right, though, and
politics may not have been a factor. I am pretty sure the writing style
would have been, however.
> The paper also contradicts itself in a way. It claims that these
> brain wave changes are why people don't like music that has been
> low-pass filtered, but it doesn't explain why people can't bring
> these likes and dislikes to the point of conscious expression. Seems
> like some kind of long term blind test would provide similar results.
Now perhaps you should reflect upon that. Then you might begin to grasp why
many of us believe there is a major difference in what is heard when
listening relaxed, and what is heard when intellectually focused on hearing
differences.
These folks didn't know they were being tested or what was being tested.
They weren't asked to compare anything. But nonetheless, on a monadic
basis, they rated the variable of music reproduction higher in the test
using inaudible frequencies.
An audio dealer (a very good one but he will be unnamed here, since he did
not give me permission for attribution) and I discussed this once with
regard to the tube vs transistor amplifier issue. We both agreed the ideal
test would be to have identical black boxes as amplifiers placed in the
system and then switched every three months. The subject would be asked to
keep simple weekly diaries indicating how much time he had spent that week
listening, and how much or how little enjoyment he derived from that
listening (perhaps there might be a more sophisticated way of getting at
this). At the end of the year, with order reversed for every other subject,
one would have a checkerboard which could be used to statistically test
whether the periods of tube use gave significantly higher levels of
listening and/or enjoyment when listening. We were both betting on tubes,
of course, and the test is obviously almost impossible to execute (I love
tubes but I could never figure out how to mask a tube amp as a solid state
amp without giving itself away). As well as impossibly "grand" for any but
the best-funded organizations to undertake. But it would finally answer
that question once and for all.
The test presented in this paper is an equally rigorous and sophisticated
test.
Stewart Pinkerton
>"Arny Krueger" <ar...@hotpop.com> wrote in message
>news:aqu11...@enews3.newsguy.com...
>> I'm hoping that the THIRD time around you'll figure out that I'm
>> saying that they are different and NOT identical!
>>
>> They should be identical but they AREN'T!
>
>The LFC files are identical. Any intermodulation effects are from the
>SEPARATE addition (through separate sound file, amplifier chain, and
>speakers) of the high frequency content. Thus any difference is due to the
>natural intermodulation of the sound...as it occurs in nature. That is not
>a test flaw...it is part of the experiment.
Harry, that is just *bad* science. The replay system is absolutely
*not* 'natural', and who's to say what caused the obvious differences
in the FRS and LFC spectra? What we need is *confirming* data from
another source. The experiment was done eleven years ago, there is
*huge* commercial pressure for corroboration from the SACD and DVD-A
lobbies, and what evidence do we have from other experiments? NONE.
Stewart Pinkerton
>> >"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
>> >news:aqt3kt$h1h$1...@bourbaki.localdomain...
>> The HF
>> replay system has *not* been shown to have no IM artifacts, indeed the
>> 'FRS' plot is *significantly* different from the 'LF' plot in the 'LF'
>> region, which is *highly* suspicious.
>
>But if the HF replay system had IM artifacts that fell into the
>audible range, they would have been heard when playing just the HFC
>signal.
Possibly, or possibly they needed the reinforcement of the LF signal
to pass the threshold of audibility.
The important point is that, despite the enormous commercial pressures
to validate this experiment, there has been *no* confirmation from
other experiments.
>And according to the paper, when just the HFC signal was
>played, it couldn't be tell apart from just ambient noise, so it was
>inaudible. So there were no IM products due to the HF replay system
>that feel into the audible range.
See above.
>Also, the HFC and the LFC were played through totally separate signal
>and transduction paths, so the only possibility of interaction between
>both signals is on the air or the ear. I'm I missing something?
You're missing the fact that there has been no replication of these
results, for starters. This ain't rocket science, and Sony are
*desperate* to prove that we need 100kHz bandwidth, so where is the
flood of corroborating results that you would expect?
Stewart Pinkerton
>"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
>news:aqullh$7pn$1...@bourbaki.localdomain...
>> On Wed, 13 Nov 2002 18:48:45 GMT, "Harry Lavo" <harry...@rcn.com>
>> wrote:
>> >Any
>> >intermodulation in this test is "natural" as the result of the high
>> >frequencies, not an artifact.
>>
>> No Harry, you are once again working on 'edited higlights'. The HF
>> replay system has *not* been shown to have no IM artifacts, indeed the
>> 'FRS' plot is *significantly* different from the 'LF' plot in the 'LF'
>> region, which is *highly* suspicious.
>>
>No the LCF portion was exactly the same. When the high frequency portion
>was added via a separate, synced system, it may have slightly changed the
>lower as would happen in nature.
Harry, the addition was *not* 'natural', it came via a replay system.
Where is the *corroboration* which would verify this experiment?
>since their is no way of knowing how important each
>octave of inaudible sound is, I think we can assume the lower portion
>(perhaps 20khz-35khz ?? where most of them are in much music is probably
>most relevant.
Harry, if there's no way of knowing, then you cannot assume
*anything*. Sheesh, that's what science is *about*. Now, since that
experiment failed peer review by the JAES editorial committee eleven
yeers ago, and only recently squeezed into a totally unrelated medical
journal with no audio expertise, where is the *corroboration* which is
a *vital* part of real science? Remember cold fusion?
> So perhaps extension to 35khz, for example, accounts for
>95% of any possible impact. I assume some future research will deal with
>this. Until then, we can only guess how much is enough.
Quite so, hence perhaps extension above 20kHz counts for f**k all. We
need *evidence*, not mere speculation.
>> >You don't think some moving coil cartridges with line-contact type stylii go
>> >out to 40-50khz?
>>
>> Sure I do - but where are the *albums* with response above 15kHz?
>>
>There are plenty...
Name two...................
>and even on the ones that don't extend much beyond that,
>pehaps the gradual physical attenuation of an extended recording has a
>different effect than an abruptly cut-off electrical "wall". At least the
>tested "wall" showed *no* positive emotional impact.
You are extrapolating wildly here. Especially since you just claimed
that a 'brick-wall' filter has no psychoacoustic effect...........
I'm a sniper - I get no satisfaction from a barrel full of fish.
>> Harry, that paper was presented to an AES convention in 1991. It
>> failed peer review for publication in the AES journal, and the authors
>> took *nine years* to finally find acceptance in a journal totally
>> unrelated to audio. You know this, you are choosing to ignore it. Who
>> is being dishonest?
>
>You know for a fact that their prior paper *failed* peer review? Even Arny
>didn't claim to know. All you know is that it was not published for
>whatever reason, as I pointed out elsewhere.
Right, so a paper presented at an AES convention was not published in
the AES Journal for some other reason than failing peer review. Do you
see black helicopeters above your house? :-)
>Moreover, it was a different paper with more than twice as many
>contributors, additional tests, and more sophisticated analysis.
And the same basic experiment.
> And it is
>geared at a more sophisticated audience and one with no axe to grind.
The AES has no axe to grind, either. Especially since a positive
result would prove the need for a wider bandwidth system which could
sell the same music all over again on a newer medium. More work for
AES members.........................
> Your
>really think the biggies behind the AES in 1991 wanted a paper that said
>that CD was fundamentally flawed?
Those helicopters are getting lower...........................
chung
>
> Even should you be correct, there is the little matter that CD-type
> frequency response doesn't move people, physiologically verified, versus
> more full range reproduction. That not significant enough for you?
>
One of the most moving musical experiences for me is listening to
Horowitz's Moscow concert. The Rachmaninoff Preludes, the Scribian
Etudes, and the Liszt Traumereii move me to tears. Many of my friends,
audiophiles or not, share the same feelings. I have listened to the CD,
Laserdisc and the DVD versions, and they are all equally powerful. I
cannot imagine the vinyl version moving me any more.
Chung.
Thomas A
> "Thomas A" <Thomas_...@hotmail.com> wrote in message
> news:aqrfu...@enews2.newsguy.com...> news:<aqpce...@enews4.newsguy.com>...
> > > "Thomas A" <Thomas_...@hotmail.com> wrote in message
> > > news:aqovr...@enews4.newsguy.com...> news:<aqnb94$1l2$1...@bourbaki.localdomain>...> > > > > news:aqn9cs$ru$1...@bourbaki.localdomain...> > > > > > news:aqn5u3$uou$1...@bourbaki.localdomain
> > > > > >
> > > > > > > http://www.prosoundweb.com/chat_psw/transcripts/rupert.php
> > > > > >
> > > > > > Well, here's the quote:
> > > > > >
> > > > > > "We discovered that there was a 3 dB peak at 54kHz Geoff's golden
> > > > > > ears had perceived that there was a difference."
> > > > > >
> > > > > > What is not known is the order and damping of this resonance.
> > > > > > Depending on the actual parameters, there could be significant
> > > > > > response variations below 20 KHz.
> > > > > >
> > > > >
> > > > > True, it is possible. But you are ignoring all the clinical
> evidence
> > > > > included in the links.
> > > >
> > > > I've previously read that article and also their former article
> > > > regarding the "hypersonic" effect. I have several questions to this
> > > > article. First, have anyone else repeated their experiments and if so
> > > > do they get the same results? Second, is the high amount of HF unique
> > > > to this kind of music/instruments? Would there be an effect if they
> > > > made the same experiment with orchestral music? If so, why did they
> > > > not show this with ordinary music instruments? That would have
> > > > strengthen their conclusions. I've read that it is possible to hear
> > > > the effects of HF just above the hearing limit since it can modulate
> > > > the audible lower freqeuncies, IF the HF is high enough in level.
> > > > Don't know if that is published though.
> > > >
> > > > Thomas
> > >
> > > Reread the article. They built the reproduction system carefully to
> prevent
> > > any artifacts, and raised the cutoff frequency for the high frequency
> filter
> > > to specifically prevent audible artifacts. They also controlled for
> this
> > > variable in the test design itself.
> >
> > I do believe that they tried that. However, peer reviewed articles
> > need not imply that they are "true". Many articles with so-called
> > astonishing results have been published and later shown to be caused
> > by flaws in the test setup or the intepretations. Since I do both
> > write and review articles in the field of Microbiology, I know that
> > this happens. So to be solid, the experiments need to be repeated by
> > an independent lab to be meaningful. The lack of articles showing a
> > negative result in this area may be due to that it is difficult to
> > publish negative results; however, that does not imply that this
> > article can be challenged since it is now published. But as far I as
> > know, this is the only group that has found this result.
> >
>
> Absolutely, you won't find an argument from me here. From my very first
> post i said (if verified by future testing) now, didn't I. But the test was
> published two years ago. So to duplicate it would require dedication,
> money, time and for what end. To debunk the findings..only Arnie has the
> motivation to do that. I don't think it will happen. What will happen are
> other studies that push the boundries further...if the test results hold up,
> or will run across contra-indicating facts as a result and these will be
> published in due course.
>
> > If the intepretations are correct, another group should do it with a
> > different method, perhaps making DBTs in combination with mismatch
> > negativity response measurements of the brain. This method measures a
> > special signal in the brain which respond to changes in sound.
>
> I'm not sure I understand what you mean here, but I suspect the same caveats
> apply that I expressed above.
>
> At this point it is just one finding and therefore a very tentative "fact".
> But it is a fact that has enormous implications.....and certainly it should
> create a newfound sense and caution here in this forum about what is "known"
> about hearing and music.
>
> Haven't seen evidence that it has started to have that effect, however. :-)
I just wonder whether this finding is based on something new or not.
If two tones quite close in frequency combine they may lead to beat
waves or Tartini tones. So I wonder then if inaudible high frequency
sound may combine with audible sound to create effects that can be
perceived.
http://www.societymusictheory.org/www-talk/smt/2002may.smt/0118.html
http://www.societymusictheory.org/www-talk/smt/2002may.smt/0101.html
Thomas
Thomas A
> Thomas_...@hotmail.com (Thomas A) wrote in message news:<aqovr...@enews4.newsguy.com>...
> (...)
> > I've previously read that article and also their former article
> > regarding the "hypersonic" effect. I have several questions to this
> > article. First, have anyone else repeated their experiments and if so
> > do they get the same results?
>
> yet. The closest approximation appears to be Fujioka et al (2002), who
> studied evoked cortical magnetic fields (another measure of cortical
> function) to single-frequency tones of up to 40kHz. No cortical
> activity was demonstrated with ultrasounds. The authors suggest that
> their delivery system involved only air-conducted sounds (i.e., no
> bone-conducted sound), although I don't see how they guaranteed that.
> Despite these results, they did not specifically disagree with Oohashi
> et al.
>
> > Second, is the high amount of HF unique to this kind of music/instruments?
>
>
> > Would there be an effect if they
> > made the same experiment with orchestral music?
>
>
> > If so, why did they not show this with ordinary music instruments? That would
> > have strengthen their conclusions.
>
> although pursuing the role of ear-based distortion products would;
> indeed, they may explain the effect entirely, and would be consistent
> with the findings of Fujioka et al, too.
>
> > I've read that it is possible to hear
> > the effects of HF just above the hearing limit since it can modulate
> > the audible lower freqeuncies, IF the HF is high enough in level.
> > Don't know if that is published though.
>
> going back to the late 1800s, though I'm not sure if any are available
> online, without a subscription.
>
> If I knew my tweeter could perform >20kHz, I would play around myself,
> to see if I could obtain clear audible dist prods from one (or more)
> inaudible primaries. However, if the tweeter doesn't have signficant
> response way oop there, it'd be all a bit futile...!
>
> Steve.
> Fujioaka et al (2002). Neuroscience 112, 367-381.
You may be interested to read this (I recently found this):
http://www.societymusictheory.org/www-talk/smt/2002may.smt/0102.html
Basically it describes Tartini tones (when two tones combine to create
a lower tone) which have been known since the 1800s, but also contains
references quite recent about the auditory system that enable us to
percieve this.
IF the japanese study is true, I can only speculate. This is really
not my field, but it is possible that high frequency signals combine
with lower frequency to create Tartini tones that may be scored as
"audibly different".
Thomas
Arny Krueger
>> OK, so hearing different stuff below 20 KHz makes people respond
>> differently. Where's the news in that?
> Even should you be correct, there is the little matter that CD-
> type frequency response doesn't move people, physiologically
> verified, versus more full range reproduction.
I think I missed something here, like the scientific paper that
supports this claim.
>That not significant enough for you?
What's significant to me is all the scientific papers that say that
response much beyond 15 KHz is sonically irrelevant. Shall I cite
them?
chung
>
> Even should you be correct, there is the little matter that CD-type
> frequency response doesn't move people, physiologically verified, versus
> more full range reproduction. That not significant enough for you?
>
CD's move me just as much as, if not more than, vinyl. I have been
listening to vinyl since the '60's. It is the performance that moves people.
I would also say that CD's give a more full range reproduction, not
vinyl. Both in terms of S/N and frequency response.
Chung.
Thomas A
I was unclear in my presentation. I do not really mean of the IM dist
in the equipment or speakers, but of the IM dist created by non-linear
media other than that. That could be non-linearities in air,
discontinuites creating reflections e.g. walls, furniture and the
non-linearities of the inner ear as well. All which are components of
the listening environment.
Thomas
Richard D Pierce
>
> Even should you be correct, there is the little matter that CD-type
> frequency response doesn't move people, physiologically verified, versus
> more full range reproduction. That not significant enough for you?
>
I'm sorry, Harry, but this statement is simply ill-informed and
samcks of arrogance as well. You state it as if it was an
established, undisputed fact.
There are PLENTY of people for whom this statement is simply
false. One of the most satisfying recordings I have, to merely
pick ONE example from many, is the Saorgin performance of the
Buxtehude organ works. Another is Gabe Weiner's recording of the
Bach St. John Passion. There are many such examples.
--
| Dick Pierce |
| Professional Audio Development |
| 1-781/826-4953 Voice and FAX |
| DPi...@world.std.com |
Dave Collins
pat...@dircon.co.uk (Stewart Pinkerton) wrote:
> You're missing the fact that there has been no replication of these
> results, for starters. This ain't rocket science, and Sony are
> *desperate* to prove that we need 100kHz bandwidth, so where is the
> flood of corroborating results that you would expect?
A minor point, but the SACD system is generally "only" 50kHz. There is
a switchable filter on some players that allows for 100kHz, but they
strongly recommend you leave it at 50k.
What if it turns out there really is a "high" generated by the
out-of-band energy? Will the government step in to regulate it?
Personally I'd go with the intermodulation theory. If there was >20k
hearing in adult humans, wouldn't it have been discovered by now? That
would be a major discovery, cover of the ASA journal, at the least...
DC
Nousaine
This is somewhat off-topic but I've heard Mead Killion of Etymotic Research use
the term Tartini tones to describe certain sounds produced in the ear canal and
radiated out (can be heard outside the ear by others) under certain conditions.
Howard Ferstler
>
> "Howard Ferstler" <fer...@attglobal.net> wrote in message
> news:aqvp9l$nv9$1...@bourbaki.localdomain...
> > Harry Lavo wrote:
> > >
> > > "Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
> > > news:aqt3kt$h1h$1...@bourbaki.localdomain...
> >
> > > > Howard wrote:
> > > > >> Going retro for a moment, how many LP cutting lathes can respond
> high
> > > > >> enough up for these signals to be present on the LP recordings that
> so
> > > > >> many subjectivists claim are superior to the compact disc for
> reasons
> > > > >> that involve ultrasonic frequencies? Those kinds of signals do not
> > > > >> exist on the LP record, and so even if this paper is on the mark it
> > > > >> does not explain the so-called superior sound many individuals
> claim
> > > > >> for the LP record. And how many lionized phonograph cartridges can
> > > > >> respond high enough up (I mean high enough up with any degree of
> > > > >> flatness and strength) for this to matter?
> >
> > > > >This has been covered here ad infinitum
> > > > Quite so - and never answered by the vinyl fans..........
> > > You don't think some moving coil cartridges with line-contact type
> stylii go
> > > out to 40-50khz?
> > Out to 40-50 kHz flat and smooth? Is there a way to validate
> > this? Also, just kind of record wear are we talking about at
> > 40-50 kHz? How many plays can an LP endure before those
> > signals are either erased or fully screwed up?
> >
> > Howard Ferstler
> Did I say flat and smooth? (hint: no) Do we know how flat and smooth they
> have to be? (second hint: no)
Well, I take issue with the implication that flat and smooth
are not issues that matter. If as you say the frequencies
well above 20 kHz are important, then it stands to reason
that a driver or stylus that is reproducing them should be
doing the job in a flat, smooth, and undistorted manner. Any
other way would not make sense.
> Take a look at old IAR issues, where the frequency responses were published.
> And more recent cartridges have gotten better.
But do typical disc program sources have such frequency
extension? Has anyone ever tried to see if this is the case?
> As to how many plays, line-contact stylii and more sophisticated resonance
> control have greatly reduced wear, even at extremely high frequencies.
Has anyone, other than someone who is bent upon proving the
superiority of the LP record, ever tried to see if what you
say is true? At the frequencies you note, there has to be
huge pressures being applied to the disc material. What's
more, at those frequencies the disc material itself probably
has irregularities that are as etched into the disc surface
as profoundly as the lathe-cut material.
And Stewart is correct: the vinyl fans really never have
adequately answered questions about the behavior of the
medium out to 40-50 kHz. Most LP records are lucky to get
out to 15 kHz. The guys who cut them do not bother to go
much higher, at least at levels that are meaningful, because
of technical limitations.
Howard Ferstler