What kind of evidence do you consider as sufficient?

[Jakob2]

And when I said your question doesn’t make any sense unless the hypothesis under test is precisely and completely identified you dismissed the idea.

my answer was:
What kind of evidence do you consider as sufficient?

Obviously i did not dismiss your idea, but tried to politely hint that the level of abstraction wasn´t sufficient. (of course in the spirit of an intelligent discussion )

Now you are questioning the hypothesis in my example and call me biased.

Mhm, i only pointed to the (quite obvious) fact that your answer exclusively addressed a negative result of the test; that is bias by definition, isn´t it?

Hence, I’m afraid there is no basis for any intelligent discussion here. Discussions started with an agenda are not my cup of tea.

I guess it illustrates why trying to control one´s own bias is so important, as there isn´t really evidence for an agenda.

 

[mmerrill99]

OT deserves imo an additional thread too

Given the fact that our (usual two channel stereophonic) reproduction is a very lossy version of reality that depends strongly on the experience and imagination ability (learned abilities) of the listeners i doubt that there can exist only one "right/correct" kind of reproduction that suits all humans.

I'm not so sure this is a correct statement based on the idea that auditory perception is actually a way of making sense of the incomplete set of signals in the waveform - the ill-posed problem. AFAIK, this is achieved by analysis/categorization & ultimately comparison to our internally stored rules/models for how the real world (our learning space)behaves models. We all have followed this same training of our auditory perception from infancy, hence The Harmon/Toole blind testing of speakers led to the same speaker preference - measurements seemed to show that smooth off-axis behavior was the determining factor.

So, although we are accepting of less than optimal sound & can live with it, even believing it is the best ever, when something which is more realistic (better matches our auditory expectations of real world sound) comes along it pushes aside our previous top-dog sound.

Edit: Sure I realize that the stereo playback is an illusion but it is still judged by a perception that bases it's analysis on the models of behavior of audio objects from the real world & as such better matching to this model provides a better illusion

 

[Pano]

Quite ironic given the audiophile hatred of tone controls.

Ha ha, true enough. But just imagine the mess if they did have tone controls!

 

[Jakob2]

<snip> We all have followed this same training of our auditory perception from infancy, hence The Harmon/Toole blind testing of speakers led to the same speaker preference - measurements seemed to show that smooth off-axis behavior was the determining factor.

"All humans" is meant to be understood as "to each and every human".
The "Harman results" are group results, individual response will most likely be (sometimes quite) different.

Interindividual difference can be surprisingly large due to physiological differences and experiences made.
That would not been a factor if we were able to reproduce original soundfields but as we can´t do that (beside binaural approaches) it is human´s response to the cues that trigger (or do not trigger) a certain kind of matching impression formed/created by our brain.

So, although we are accepting of less than optimal sound & can live with it, even believing it is the best ever, when something which is more realistic (better matches our auditory expectations of real world sound) comes along it pushes aside our previous top-dog sound.

Of course that can happen (as usually none of us really knows about the sound quality that can be realized with a certain content) but is there an inherent contradiction to my statement?

 

[mmerrill99]

"All humans" is meant to be understood as "to each and every human".
The "Harman results" are group results, individual response will most likely be (sometimes quite) different.

What I'm trying to say is that we all coalesce towards the same auditory perception just as in visual perception & the Harmon speaker results show that even with possible individual variability, all tested individuals scored the same speaker higher than the others. My take-away is that we are all in agreement with what represents the better sound/better illusion.

Interindividual difference can be surprisingly large due to physiological differences and experiences made.
That would not been a factor if we were able to reproduce original soundfields but as we can´t do that (beside binaural approaches) it is human´s response to the cues that trigger (or do not trigger) a certain kind of matching impression formed/created by our brain.

Yes, there are physiological differences but normal differences are compensated for - take the shape of the pinna & ear canals - all individual with individual resonances at the eardrums but we all seem to adjust to our own individual spectrum which doesn't upset our coalescing towards the same target realistic sound which leads me to believe that absolute frequency/amplitude is not the judgement criteria but many other things including the relative ratios within the soundfield of frequencies/amplitude.

It's a bit like how we seem to pick up language - we hear lots of examples of people using a collection of sounds in a particular way & eventually try to ape these sounds in the order we hear them over & over again. We get feedback from parents when we ape the correct sounds in the correct order & e have learned our first word & so on. It's a feedback mechanism using target examples which leads us all to the same words & language for whatever culture we are immersed in.

It's amazing what can be achieved with such a simple approach but huge amounts of examples.

So in the same way as we learn our language by constant exposure to words in our language, we learn & model our auditory processing system through being exposed to the sounds of the world. These are not culturally different & hence we all develop pretty universal internal models of what sounds natural

Of course that can happen (as usually none of us really knows about the sound quality that can be realized with a certain content) but is there an inherent contradiction to my statement?

Well, IMO based on the universal model that we all share regarding sound, I disagreed with your statement "i doubt that there can exist only one "right/correct" kind of reproduction "

As I said, because auditory perception is a best guess judgement, we can easily interfere with it by strongly held views of what we should be hearing & what the playback illusion should do, hence some people can get themselves into cul de sacs where they can convince themselves for years. But this is not proof that we don't share common internal auditory models as when people are not being led by strongly held beliefs I believe they recognize & coalesce towards the same better, more natural sound

 

[Markw4]

Harmon speaker results show that even with possible individual variability, all tested individuals scored the same speaker higher than the others. My take-away is that we are all in agreement with what represents the better sound/better illusion.

Harmon trains listeners for how they want them to rate speakers. If someone can't learn the Harmon way and pass their tests, they do not qualify to serve as listeners.

That is not the same as selecting listeners randomly from the population of all speaker buyers.

 

[mmerrill99]

Harmon trains listeners for how they want them to rate speakers. If someone can't learn the Harmon way and pass their tests, they do not qualify to serve as listeners.

That is not the same as selecting listeners randomly from the population of all speaker buyers.

Yea, I've heard that before but not sure how true it is - I think they did some tests on subjects without pre-training but yea, I know there are limitations to their tests. So it may be a bad example but my point still stands - I believe we all have roughly the same internalized model/rules for what defines realistic sound & if that is the way we judge the illusion from our playback systems then it means the systems that best match this model/rules will sound better, more natural.

 

[Jakob2]

What I'm trying to say is that we all coalesce towards the same auditory perception just as in visual perception & the Harmon speaker results show that even with possible individual variability, all tested individuals scored the same speaker higher than the others. My take-away is that we are all in agreement with what represents the better sound/better illusion.

I´m not sure that all "individuals score the same speaker higher than the others" as i haven´t seen individual responses from these tests. Were they published? If so i´d be very much interested in a citation.

Individual Responses from earlier experiments done by Toole /1/ showed that the variance was dependent on the specific loudspeakers. For example the Quad ESL 63 triggered a wide range of different jugdements, it got the lowest ratings overall from some listeners why others rated it among the best they ever heard.

On other loudspeakers the rating among the listeners was far more consistent.

<snip>So in the same way as we learn our language by constant exposure to words in our language, we learn & model our auditory processing system through being exposed to the sounds of the world. These are not culturally different & hence we all develop pretty universal internal models of what sounds natural

To that i agree for obvious reasons.

Well, IMO based on the universal model that we all share regarding sound, I disagreed with your statement "i doubt that there can exist only one "right/correct" kind of reproduction " .....

That seems to me as a "non sequitur" because we are talking about multidimensional evaluation of a reproduction that represents a multidimensional deviation from the original/natural sound that you´ve mentioned above.

......As I said, because auditory perception is a best guess judgement, we can easily interfere with it by strongly held views of what we should be hearing & what the playback illusion should do, hence some people can get themselves into cul de sacs where they can convince themselves for years. But this is not proof that we don't share common internal auditory models as when people are not being led by strongly held beliefs I believe they recognize & coalesce towards the same better, more natural sound

It could be that at a certain point overall convergence appears (although imo the "uncanny valley" problem will kick in), but today i can confirm that the widespread in judgement about big electrostats like the Quads still exists. Even when listeners agree quite closely in their ratings of other speakers they can rate the Quads totally differently from each other.

/1/ Floyd E.Toole. “Subjective Measurements of Loudspeaker Sound Quality and Listener Preferences,” J. Audio Eng. Soc., (1985) 33, pp. 2–31.

 

[gpauk]

I believe we all have roughly the same internalized model/rules for what defines realistic sound & if that is the way we judge the illusion from our playback systems then it means the systems that best match this model/rules will sound better, more natural.

But.. that's your belief, and it's just a hypothesis...

 

[mmerrill99]

"All humans" is meant to be understood as "to each and every human".
The "Harman results" are group results, individual response will most likely be (sometimes quite) different.

Interindividual difference can be surprisingly large due to physiological differences and experiences made.
That would not been a factor if we were able to reproduce original soundfields but as we can´t do that (beside binaural approaches) it is human´s response to the cues that trigger (or do not trigger) a certain kind of matching impression formed/created by our brain.


Of course that can happen (as usually none of us really knows about the sound quality that can be realized with a certain content) but is there an inherent contradiction to my statement?

I understand "all humans" to mean normal hearing humans or what is usually abbreviated in auditory research to NH, not hearing impaired people or HI

So my previous statement should read "What I'm trying to say is that we all (NH people) coalesce towards the same auditory perception & therefore there it follows that there should be "one same right/correct kind of reproduction" - we evaluate the reproduced sound in exactly the same way as we evaluate sounds from the natural world

 

[mmerrill99]

I´m not sure that all "individuals score the same speaker higher than the others" as i haven´t seen individual responses from these tests. Were they published? If so i´d be very much interested in a citation.

OK, I'll have to check this again

Individual Responses from earlier experiments done by Toole /1/ showed that the variance was dependent on the specific loudspeakers. For example the Quad ESL 63 triggered a wide range of different jugdements, it got the lowest ratings overall from some listeners why others rated it among the best they ever heard.

On other loudspeakers the rating among the listeners was far more consistent.

One issue with the Harmon tests was that the electronics used to drive the speakers were not particularly impressive so this becomes as much an amp/speaker combination test - it's actually the full playback chain that's under test. Harmon, like a lot of engineers ignore the electronics,considering them blameless - I don't subscribe to this. What the Quad results might be showing is a different window on what's upstream from it mixed with its own flaws?

To that i agree for obvious reasons.

That seems to me as a "non sequitur" because we are talking about multidimensional evaluation of a reproduction that represents a multidimensional deviation from the original/natural sound that you´ve mentioned above.

In the sense that we listen to our replay systems & analyze/judge the sound in the same way as we do sounds we encounter in the world, I don;t see it as multidimensional, if that's what you mean?

It could be that at a certain point overall convergence appears (although imo the "uncanny valley" problem will kick in), but today i can confirm that the widespread in judgement about big electrostats like the Quads still exists. Even when listeners agree quite closely in their ratings of other speakers they can rate the Quads totally differently from each other.

/1/ Floyd E.Toole. “Subjective Measurements of Loudspeaker Sound Quality and Listener Preferences,” J. Audio Eng. Soc., (1985) 33, pp. 2–31.

I think what might be confusing is that we can judge the same reproduction anomaly in different ways as usually there is not just one flaw so we may have different weightings between these flaws - maybe that's what you mean by "multidimensional evaluation" If that's what you meant then I agree but If we could just isolate to hearing one flaw at a time, I'm pretty sure we would all coalesce to the same correct sound.

 

[mmerrill99]

But.. that's your belief, and it's just a hypothesis...

Which part are yo referring to is my belief - that "we all have roughly the same internalized model/rules for what defines realistic sound" (talking about NH people here)?

Research into auditory perception is predicated on the basis that all our auditory systems work in the same way & also that there is an internal process/analysis that is the same for all. Their investigations/research is aimed at uncovering the models/rules that underlie auditory perception

Or "that is the way we judge the illusion from our playback systems then it means the systems that best match this model/rules will sound better, more natural."?

I can't see how we would use a different auditory mechanism to listen to our playback system?

 

[scott wurcer]

Or "that is the way we judge the illusion from our playback systems then it means the systems that best match this model/rules will sound better, more natural."?

Why? There is a large body of evidence that for reproduced music many people prefer very unnatural sound as their illusion. It's funny as this goes along your comments become more and more rigidly objectivist.

 

[mmerrill99]

Why? There is a large body of evidence that for reproduced music many people prefer very unnatural sound as their illusion.

Can you show the evidence?

It's funny as this goes along your comments become more and more rigidly objectivist.

I would categorise myself as a pragmatist - neither an objectivist nor subjectivist? I'm not blinded by the objectivist belief system!

 

[scott wurcer]

Can you show the evidence

I already mentioned Romy the Cat. Maybe you have not traveled in these circles, there is a large group of audiophiles that will have nothing but flea power SET amps and full range horns. Not to mention the folks that listen to 1920's to 1940's 78's on original equipment. I've listened to great music with many of them, you don't know how small your posse is.

Don't fool yourself all that stuff about complex changing waveforms, is delusional dogma (I only say dogma because it gets repeated ad nauseam) based on a basic lack of understanding of fundamentals.

I suggest you solicit a group of folks to explore your hypothesis and do the work and present results, the constant verbosity serves no purpose.

 

[Markw4]

... the constant verbosity serves no purpose.

+1

Suggest listen more, think longer, avoid excessive repetition. That, despite possible feelings of confidence that only you see reality as it truly is, and if others only knew what you know and were entirely rational then surely they would see things just as you do.

 

[rayma]

Why? There is a large body of evidence that for reproduced music many people prefer
very unnatural sound as their illusion.

I must say that this has also been my experience, at least for some, who have a strong
preference for highly colored sound. Sort of like those who heavily salt their food.
Who knows why they like it that way.

One year at CES I heard an very expensive French speaker system that had a pronounced
nasal quality. The French guys demoing it couldn't hear that, though. They thought it had
a "natural" sound, just like they sounded when talking.

 

[mmerrill99]

I already mentioned Romy the Cat. Maybe you have not traveled in these circles, there is a large group of audiophiles that will have nothing but flea power SET amps and full range horns. Not to mention the folks that listen to 1920's to 1940's 78's on original equipment. I've listened to great music with many of them, you don't know how small your posse is.

Don't fool yourself all that stuff about complex changing waveforms, is delusional dogma (I only say dogma because it gets repeated ad nauseam) based on a basic lack of understanding of fundamentals.

I suggest you solicit a group of folks to explore your hypothesis and do the work and present results, the constant verbosity serves no purpose.

I'm sorry that you think this is delusional dogma but you should ask your friend Mallinson (former ESS chief design engineer) about it as he has stated the same - I quote some things from your friend, Martin Mallinson

One
feature that allows the Sabre DAC to reach such high
performance is the modulator design: listening tests show that
experienced listeners can distinguish between a conventional
ΣΔ modulator and the extensions to the ΣΔ modulator as
implemented in our range of HyperStream modulators.
Experiments performed in our listening room suggest that it is
the absence of correlations in parameters that seem to
contribute to the listening experience. Specifically, two
potential correlations are important: the signal level and noise
level do not correlate – noise does not perceptibly vary with
amplitude or rate of change of the signal; and transient events
are rapidly extinguished in the integrators of the modulator
which suppresses anomalous noise potentially created as the
transient passes through the signal path.

You asked me elsewhere did I listen to & understand Mallinson's presentation - I sure did & quoted parts of it back to you.

One of the outstanding points was that ESS didn't discover issues by using their standard set of DAC tests, they listened to feedback from people within ESS who could discern audible issues & didn't dismiss them as imaginings "because our tests show the DAC is blameless" - they went in search of a measurement which would show the issue. According to Mallinson, when a quiet sound is between two music transients, the noise is modulated by this condition & this noise modulation is audible

Because ΣΔ modulators are very non-linear systems only
approximations of performance are possible in frequency
domain simulations. Time domain simulations are useful to
verify first-order stability (i.e. that the loop does not oscillate
in the typical condition) but beyond this, simulations must be
augmented with hardware tests. We have performed extensive
hardware testing of the HyperStream modulators over many
years (and many millions of shipped products) and
consequently have designed them to show additional features
that are not commonly tested. We believe that some of these
features of the modulator are the reasons why our users tell us
that the Sabre “sounds better” than similarly specified parts.

certain ΣΔ modulators when provided with a
rapidly changing input signal will exhibit non-linear noise
behavior as they process the transient. This is because all noise
shaping modulators are feedback systems and the usual design
process (supported by commonly available design tools)
operates to minimize in-band noise suppression while
maintaining stability. This noise-optimized stable loop
configuration will lead to an output that matches the input to
the required degree within the requested bandwidth as
expected. However this typical design process neglects the
dynamic response of each state variable: there are choices of
Q (and relative gain) that minimize noise but result in
relatively large lightly damped resonances of the internal state
variables. The consequence of this is that in a quiet passage of
music the state variables of the modulator are all operating
within a certain “state space” and the quantization noise
shaping is described by the noise characteristics in this
“volume” of the space.

 

[mmerrill99]

+1

Suggest listen more, think longer, avoid excessive repetition. That, despite possible feelings of confidence that only you see reality as it truly is, and if others only knew what you know and were entirely rational then surely they would see things just as you do.

Hey, I'm just stating my viewpoint with as much evidence as I can. The verbosity is my attempt at explaining what I believe/think. I'm willing to be wrong & have healthy debate about points - sorry if it comes across that I know more than others, I just state what I know & if this upsets anybody's world view there's nothing I can do about that.

Sorry if some understanding of auditory perception upsets some!!

 

[mmerrill99]

I already mentioned Romy the Cat. Maybe you have not traveled in these circles, there is a large group of audiophiles that will have nothing but flea power SET amps and full range horns.

Do SET amps & full range horns sound unnatural? I don't know haven't heard them.

Not to mention the folks that listen to 1920's to 1940's 78's on original equipment. I've listened to great music with many of them, you don't know how small your posse is.

As I said before there are 3 of us me kettle & pot but I suspect there are many more of us than you wish to admit

Don't fool yourself all that stuff about complex changing waveforms, is delusional dogma (I only say dogma because it gets repeated ad nauseam) based on a basic lack of understanding of fundamentals.

See Mallinson