CD Vs. Vinyl?

[Guest]

Archived from groups: rec.audio.high-end (More info?)

chung wrote:
> Michael Mossey wrote:
> > Chung wrote:
> >> Michael Mossey wrote:
> >>
> >> >
> >> > I was a student at Caltech 1987 to 1991.
> >>
> >> Let me get this straight. You went to Caltech, and you believe
that
> >> interconnects may sound different?
> >
> > I don't believe that interconnects sound different---I'm just open
to
> > it as a possibility, and I see problems in the tests that have
> > attempted to rule out that possibility.
>
> So you don't believe it, but you are open to it as a possibilty. What

> evidence have you encountered that leads to this openness? From
reviews?

Primarily from four blind trials, the very first I ever did, in which I
felt sure I knew what cable I was hearing, felt that the differences
were easily audible, and was right four times. That made an
impression.

(But winning a big prize the first time you step in a casino makes a
big impression too.)

On another occasion, I did eight trials and was wrong three times.


>
> BTW, you could be a very busy man, if you are open to possibilities
that
> you don't believe in, and you want to test them. Remember the old
> saying: if you have too open a mind, your brain may fall out.

Yeah, yeah. And if you have too closed a mind, your brain will get
claustrophobic.

>
> >
> >> Don't you think that if there were
> >> audible differences, they will be accompanied by easily
meaasureable
> >> differences? So why not make measurements, instead of long DBT's?
Do
> > you
> >> have any theory as to why they may sound different?
> >
> > If we knew what to measure, yes. Anyway, don't they all measure
> > differently? Isn't the question whether the difference is within
the
> > threshold of hearing?
>
> Well, for starters, frequency response, distortion, and noise added.
We
> know that we can only detect about a 0.3 dB difference in level (but
> most likely only in a quick switching test, I might add) at 1KHz.

According to the psychoacoustics textbook I'm reading, you are
right---about discrimination on test tones such as noise or pure tones.
However, the books also say that based on the information that the
auditory nerve can carry, the theoretical DL (discriminatory limen)
could be more like 0.1 dB, it's the way the higher centers process the
signal that reduces the DL.

Discriminating test tones is a task very unlike listening for musical
enjoyment. So I'm going to be carefully examined what conclusions can
be drawn from psychoacoustical testing and whether it is valid to
extrapolate them to listening to the vastly more complex musical
signals. See below..

> There
> is really no reason why a mechanically sound cable can introduce
> distortion or noise, but those impairments can easily be measured
down
> to the -100 dB level. Any reason why you think there may be
impairments
> from cables that can be heard but not measured?
>
> If two cables measure differently, like one is 1 inch longer, you
think
> they may sound different? Why do you think that we might not know
what
> to measure, when it comes to cables?
>
> >
> >>
> >> Just out of curiosity, what did you major in at Caltech?
> >
> > "Engineering and Applied Science." It gave me a good background in
> > linear time-invariant models of systems.
>
> Well, one of my professors started a linear systems course by stating

> that the whole blackboard is the set of all systems. Then he placed a

> chalk dot on the board, and said that the dot represented the set of
all
> linear systems. It may be worthy to note that it is almost the
> definition of reductionism to consider linear time-invariant systems.

> Not that there is anything wrong with being reductionistic.

Precisely, it is reductionist. Notice that the ear is not linear
(taken as a whole; parts of it are).

Fourier showed that a complex signal can be decomposed into sine waves.
A LTI system can be modelled entirely by its frequency response.

Psychoacoustical experiments are done sometimes with sine waves. There
is a difficulty in extending the results to complex signals. That
would be easy with a LTI system, but the ear isn't one.



>
> It is really unusual to hear a graduate from Caltech talk about
> interconnects possibly sounding different, yet have no theory for why

> they may sound different, or believe that they might sound different.
Or
> hear him say that there are digital artifacts, but presents no
evidence.

You mean the indoctrination didn't stick?

-Mike

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

Michael Mossey <michaelmossey@yahoo.com> wrote:
> nabob33@hotmail.com wrote:
> > Michael Mossey wrote:
> >
> > > Don't get confused between subjective accuracy and objective
> > accuracy.
> >
> > You're the one who's confused, or trying to confuse. There is no such
> > thing as "subjective accuracy," as you define it. There may be a
> > "subjective sense of similarity," but to appropriate a technical term
> > like accuracy to elevate this concept is to muddy the waters, not
> > clarify them.

> Let me just explain the danger as I see it... I'm referring to the
> danger of placing objective measurements above conscious experience.

Rather than spend time on that, how about you consider perception
of audible difference in light of this undisputable fact:

The conclusions about the physical world drawn from 'conscious experience' are
often inaccurate if they rely on only sense perception, belief, and memory.

Optical illusions are simple examples.
The 'phantom switch' -- where nothing is changed, but the listener
thinks it has -- is a slightly more complex one. The fallibility
of 'eyewitness testimony' is well-known to members of the scientific,
legal and law enforcement communities.


> The whole point of audio is to duplicate a conscious experience. And
> some people are concerned with simply creating the most enjoyable
> conscious experience possible, while I'm concerned mostly with
> replicating the conscous experience of live music. So "accuracy" or
> "similarity" or whatever you want to call it--this concept should apply
> first to similarity of conscious experience.

Accuracy of memory would be a problem here.

> It is certainly easier to demonstrate that two signals measure
> similarly than it is to demonstrate that two conscious experiences are
> the same. But this should not place measurements above conscious
> experience. Because .... (continued below)




> >
> > > It SOUNDS more accurate;
> >
> > But it IS less accurate. Therefore, this seeming "accuracy" is likely
> > some form of illusion.

> because it is too easy to apply the label "illusion" to conscious
> experiences that seem to contradict the measurements, without actually
> having demonstrated anything about consciousness.

By this criterion, complete fantasy becomes as legitimate a 'report' of
reality as a model derived from careful observation. Does that matter
to you?


> Example: intermodulation distortion interferes with the conscious
> experience of music more than harmonic distortion. So an amplifier
> that has a tiny amount of intermodulation distortion and NO harmonic
> distortion, is less accurate than an amplier that has a lot of harmonic
> distortion and no IM distortion. The amplifier with less distortion in
> terms of numbers is arguably LESS accurate. This shows the danger of
> separating the conscious effect of distortion from the measurement of
> distortion.

It actually simply shows the danger of misinterpreting measurements.
Now, how has it been determined that IM distortion interferes with
the conscious experience of music more than harmonic distortion?

> Or, as another possibility, a form of distortion could actually
> contribute to replicating a conscious experience. In that case, it
> simply would not be true to claim that the perceived similarity is an
> illusion. If a PET scan could prove the similarity of conscious
> experience, we would have a measureable and objective way to make more
> accurate recordings through the addition of distortion. As I see it,
> our limited ability to measure conscious experience is what prevents us
> from considering this possibitilty-- not that this possibility doesn't
> exist.

As I see it, you are striving mightily to find an excuse not to believe
a more straightforward explanation -- that some 'differences' are imaginary.

--

-S
It's not my business to do intelligent work. -- D. Rumsfeld, testifying
before the House Armed Services Committee

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

Michael Mossey <michaelmossey@yahoo.com> wrote:
> Chung wrote:
>> Michael Mossey wrote:
>>
>> >
>> > I was a student at Caltech 1987 to 1991.
>>
>> Let me get this straight. You went to Caltech, and you believe that
>> interconnects may sound different?

> I don't believe that interconnects sound different---I'm just open to
> it as a possibility, and I see problems in the tests that have
> attempted to rule out that possibility.

You need to bring some hard experimental evidence to the table that shows
these problems with the test you speak of. Addressing conceptual problems
alone don't cut it. Chances are this thread will deterioriate soon and
eventually be terminated, like all the others like it that have occured
previously if you don't.

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

On 4 Apr 2005 23:51:31 GMT, "Michael Mossey" <michaelmossey@yahoo.com>
wrote:

>Stewart Pinkerton wrote:

>Playback of a stored digital signal
>is a different system than a live digital feed.

This is a mere assertion, with no basis in fact.

> Therefore it might
>behave differently. I still think that we need to search for evidence
>that it behaves differently. Within a model of that system there might
>be no way that it could behave differently. But that model is not
>reality. It is "reductionist" to draw conclusions about reality from
>the basis of a model. Not necessarily invalid, but still reductionist.
> At the moment I'm simply being careful about what conclusions can be
>drawn from a live digital feed test.

The system does not differ, whether the source material is a live feed
or stored data. Hence, it does *not* behave differently. Please read
up on digital audio before making such assertions.



>> Your hypothesis is not tenable until you offer some *evidence* to
>> support it. You have been asked several times exactly what 'digital
>> artifacts' you are referencing, but you have never answered this
>> simple question.

>I don't know what digital artifacts. Could be something in a mechanism
>already modelled, such as jitter, or it could be something new. I
>could be wrong that digital has artifacts--perhaps I like analog better
>for its distortion.

That is a *much* more likely scenario.

> In the words of Skeptic Magazine, "Maybe yes,
>maybe no."

Skeptic Magazine is not reality............

If you want a better guide to reality than Skeptic Magazine, try
Occam's Razor.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

On 4 Apr 2005 23:53:56 GMT, "Michael Mossey" <michaelmossey@yahoo.com>
wrote:

>There's really no need to choose between "analog is accurate"/"analog
>is euphonic". The answer might depend on the individual. I think it
>is reductionist to answer this question in one way that applies to all
>individuals.

You have a love of the term 'reductionism', but you seem to use it as
a convenient excuse for anything with which you disagree. The physical
soundfield is reality, and it either is, or is not, an accurate
representation of the soundfield at a particular listening position at
the original live performance (obviously, you can't use 'accuracy'
when referring to a studio recording).

There *is* a need to choose between "analogue is accurate" and
"analogue is euphonic", since this refers to the physical soundfield,
not to an individual impression. You and I will certainly get a
different impression of a musical performance, but this applies to a
live performance as well as a recording.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

On 4 Apr 2005 23:53:01 GMT, Michael <newsoffthewire@comcast.net>
wrote:

>Stewart Pinkerton wrote:
> > That is simply not true.
>
>We're running around in circles, Stew. Just think back when you were a
>kid, remember the products, remember the popularity, and you'll realize
>that the world of stereo is not dyeing.

You make another wrong assumption. When I was a kid, there *was* no
stereo......................

And while you can certainly dye stero products any colour you like,
their availabilty is unquestionably decreasing. Your attempt to
include multi-channel witrhin the compass of 'stereo, is sheer
desperation, since that is the nub of the matter. In the rarefied
world of the serious audiophile, call it 'high end' if you will,
two-channel is moribund.

> > Options are not sales.
>
>Very true, which is why we do not call sales “options”, and vice-versa.
> Options dictate the level of market interest. I could delve into the
>economics of this, but this is neither the time nor the place.

Indeed it's not, since sales are what matter, and you are ducking the
issue by pretending otherwise.

> > That doesn't mean there are any sales. Would *you* buy a $10,000
> > amplifier without a demo?
>
>See, friend, your diluting the point. The original point was in regards
>to stereo only stores and not places where stereo equipment can be heard.

Not at all, I'm trying to drag you *back* to the point - sales of
high-end 2-channel systems are decreasing. It is a dying market.

>It's a very common practice for one to audition the equipment at a home
>theater store, go online, get the best price for the equipment, and
>order it online.

While that's certainly a standard slimeball practice, only the very
dense would not realise that in short order, there will not *be* any
brick and mortar stores at which to audition these items. Hey, wait a
minute, another one just closed.................

>I still contend and can prove that stereo sales are not slowing down.

So offer some proof.

>These findings discount the sale of multichannel equipment. For an
>outsider, he would ask how they discern between stereo equipment and
>multichannel equipment, since at times, the division bell is blurred.

Well duh, multichannel has more than two channels?
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

On 4 Apr 2005 23:58:00 GMT, "Michael Mossey" <michaelmossey@yahoo.com>
wrote:

>The whole point of audio is to duplicate a conscious experience.

Some would say that it is to reproduce the original sound. You are
attempting to overcomplicate what it a simple goal - if difficult to
realise.

You accuse others of 'reductionism', I accuse you of sophistry.

>Example: intermodulation distortion interferes with the conscious
>experience of music more than harmonic distortion. So an amplifier
>that has a tiny amount of intermodulation distortion and NO harmonic
>distortion, is less accurate than an amplier that has a lot of harmonic
>distortion and no IM distortion. The amplifier with less distortion in
>terms of numbers is arguably LESS accurate. This shows the danger of
>separating the conscious effect of distortion from the measurement of
>distortion.

You claim to have graduated from Caltech, majoring in Engineering, and
you believe that such amplifiers are possible? Hmmmm

Besides, all you argument above shows, is that THD is not the best
indicator of audible distortion. We knew that already.

>Now, if I want to assert that digital is less accurate than analog, of
>course I eventually have to provide evidence that a measureable form of
>distortion exists. Digital has measureable distortion

Does it? What distortion is that?

>--the question is
>whether it is audible. I know it's not a question for you, but it is
>still an open question for me.

I fear that your life will be full of 'open questions' if you persist
in ignoring evidence. Life is short, questions are many.

--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

chung wrote:
> Michael Mossey wrote:
> > Chung wrote:
> >> Michael Mossey wrote:
> >>
> >> >
> >> > I was a student at Caltech 1987 to 1991.
> >>
> >> Let me get this straight. You went to Caltech, and you believe
that
> >> interconnects may sound different?
> >
> > I don't believe that interconnects sound different---I'm just open
to
> > it as a possibility, and I see problems in the tests that have
> > attempted to rule out that possibility.
>
> So you don't believe it, but you are open to it as a possibilty. What

> evidence have you encountered that leads to this openness? From
reviews?
>
> BTW, you could be a very busy man, if you are open to possibilities
that
> you don't believe in, and you want to test them. Remember the old
> saying: if you have too open a mind, your brain may fall out.
>
> >
> >> Don't you think that if there were
> >> audible differences, they will be accompanied by easily
meaasureable
> >> differences? So why not make measurements, instead of long DBT's?
Do
> > you
> >> have any theory as to why they may sound different?
> >
> > If we knew what to measure, yes. Anyway, don't they all measure
> > differently? Isn't the question whether the difference is within
the
> > threshold of hearing?
>
> Well, for starters, frequency response, distortion, and noise added.
We
> know that we can only detect about a 0.3 dB difference in level (but
> most likely only in a quick switching test, I might add) at 1KHz.
There
> is really no reason why a mechanically sound cable can introduce
> distortion or noise, but those impairments can easily be measured
down
> to the -100 dB level. Any reason why you think there may be
impairments
> from cables that can be heard but not measured?
>
> If two cables measure differently, like one is 1 inch longer, you
think
> they may sound different? Why do you think that we might not know
what
> to measure, when it comes to cables?

I just read in Moore that it is a pretty good assumption that Weber's
law holds not just for the ear as a whole, but for each "unit" (hair
cell/neuron) within the ear. That's a pretty good argument that
experiments to detect the DL (discriminatory limen) on pure tones will
be relevant to the transfer function on audio equipment.. i.e., if a
cable is flat to better than one DL then it provides no audible
distortion.. at least in its transfer function.

Non-linearities in the cable... you say they can be measured down to
-100 dB.

Note that measuring the transfer function of a LTI system completely
characterizes it. So does measuring the impulse response. This is
because the response to a complex signal can be summed from the
components of that signal, in a linear system.

However, this is not true of a non-linear system. So what would it
mean to *characterize* the nonlinearity of a non-linear system? In
what way could we make a few measurements and satisfy ourselves that we
have knowledge of that system? Satisfy ourselves that we can predict
that system's response to any signal?

Want to take a shot at this question?


>
> >
> >>
> >> Just out of curiosity, what did you major in at Caltech?
> >
> > "Engineering and Applied Science." It gave me a good background in
> > linear time-invariant models of systems.
>
> Well, one of my professors started a linear systems course by stating

> that the whole blackboard is the set of all systems. Then he placed a

> chalk dot on the board, and said that the dot represented the set of
all
> linear systems. It may be worthy to note that it is almost the
> definition of reductionism to consider linear time-invariant systems.

> Not that there is anything wrong with being reductionistic.
>
> It is really unusual to hear a graduate from Caltech talk about
> interconnects possibly sounding different, yet have no theory for why

> they may sound different, or believe that they might sound different.
Or
> hear him say that there are digital artifacts, but presents no
evidence.

By the way, science proceeds from the hypothesis stage to the
experiment stage. I'm in the hypothesis stage. Do you want to stamp
out competing hypotheses before they can be tested? Isn't that
unscientific and more like a turf-war?

-Mike

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

"chung" <chunglau@covad.net> wrote in message
news:d2v8d805o1@news4.newsguy.com...
> Harry F Lavo wrote:
> > "Chung" <chunglau@covad.net> wrote in message
> > news:d2p2o801nt9@news1.newsguy.com...
> >> Harry F Lavo wrote:
> >> > "Chung" <chunglau@covad.net> wrote in message
> >> > news:d2mfh30263k@news4.newsguy.com...
> >> >> Harry F Lavo wrote:
> >> >> > "Stewart Pinkerton" <patent3@dircon.co.uk> wrote in message
> >> >> >
> >> >> >
> >> >> > The industry is firmly committed to high res as sounding better.
> > They
> >> > just
> >> >> > don't know what its commercial future is. Hard to find a sound
> > engineer
> >> >> > anywhere who will support your position. Hard to find any pro
gear
> > that
> >> > is
> >> >> > not upgrading to 24/96 (in or out) or greater. Would Benchmark
have
> >> > built
> >> >> > in 24/96 downconversion if it didn't count in monitoring quality?
> > And
> >> > don't
> >> >> > tell me that it is only for mixing...that may hold for the inputs,
> > but
> >> > not
> >> >> > for the outputs. The ordinary pro audio world is moving to 24/96
> >> > surround
> >> >> > as a needed output, whether as straight DVD-A, or as downmixed
Dolby
> >> > Digital
> >> >> > / DTS. With the stereo tracks maintained at this or a higher
24/192
> >> > rate.
> >> >> > And a substantial portion of the classical pro audio world has
moved
> > to
> >> > DSD
> >> >> > for its superior naturalness as a mastering technology using
simple
> >> > micing.
> >> >> >
> >> >>
> >> >> Given the line of reasoning you have demonstrated above, can we
finally
> >> >> conclude that digital audio (including the redbook CD standard)
sounds
> >> >> better than vinyl? After all, the industry has been firmly committed
to
> >> >> digital audio since, oh, the early '80's.
> >> >>
> >> >> It's so nice to have closure...
> >> >
> >> > Only by your stretch of logic.......
> >>
> >> Somehow, your own logic, applied to vinyl vs digital, sounds rather
> >> illogical to you, I guess...
> >>
> >
> > The only thing that I've ever claimed for vinyl is that in my own
informal
> > testing among younger people raised on cd's, they come away prefering
vinyl
> > when presented with commercial material from identical masters in the
two
> > media. This done using my equipment played through my system. And I
> > presented it as antecdotal, not 'evidence' in support of the fact that
many
> > people, audiophiles and non-audiophiles alike (when exposed to very good
> > vinyl playback) seem to find it more compelling (descirbed as "more
> > realistic") than CD. I've further stated that for me this hasn't been
an
> > issue since about 1990 when CD playback equipment finally got good
enough to
> > allow me to enjoy the unmistakable benefits of the silver disk because
the
> > perceived difference if any from vinyl had gotten narrow enough that
> > convenience won out. But in close comparative listening, I still
usually
> > prefer the vinyl for sound quality alone. This as I have pointed out is
> > also consistent with my preference for jazz, chamber, and
> > singer-songwriters...none of which makes large dynamic demands at the
> > frequency extremes. So where in all this do you see logical flaws?
>
> Well, when I presented your line of reasoning (basically you were saying
> that since pro-audio has gone to 24/96 that must mean hi-rez sounds
> better) and deduced from that line that digital audio must sound better
> than vinyl, since pro-audio has gone digital since 1980's, you said that
> it was a stretch of logic on my part. That's why I said that your own
> logic somehow did not sound logical *to you*, when it was used to
> compare vinyl vs CD or digital audio.
>

No, it means high-rez has gotten close and the convenience, editing, and
mixing advantages (and cost) are much on digital's side. Moreover, we are
talking pro input here, for a long time 24/48 and for a substantial time
24/96. Not the final consumer output of 44.1/16 CD.

>
> >
> >> >
> >> > You've had closure...or at least a closed mind to vinyl...for as long
as
> >> > you've been here.
> >>
> >
> >> I have some vinyl recordings that are well recorded, and I have digital
> >> recordings that are poorly recorded. What I am objecting to, if you
have
> >> noticed, are pseudo-technical claims of the imaginary short-comings of
> >> digital so as to prove the superiority of vinyl.
> >
> > Would you care to quote me? I don't recall making those claims for a
long,
> > long time..if ever.
>
> Well, in this instance, in this very thread, people were saying that
> there were digital artifacts, yet presented no evidence. Just off the
> top of my head, I remember you were quoting someone saying that PCM
> sounded bad, and IIRC, you said it was one of the best articles on PCM?
> Perhaps I am objecting to those views you readily accept and welcome?
>

What other people has said is of no relevance to what I may have said.

> Here's that post of yours:
>
>
http/groups-beta.google.com/group/rec.audio.high-end/msg/8db30fee70542793?dmode=source
>

An article simply refering you folks to an article by Ed Meitner? In praise
of DSD? What has that to do with my views on vinyl?

> And here is another example of you actually making claims of the
> imaginary short-comings of digital:
>
>
http/groups-beta.google.com/group/rec.audio.high-end/msg/47a2dc9821fe05ce?dmode=source
>
> where you claimed:
>
> "(my issues) with CD's tendency to have less depth and a 'flatter' bass
> and lower midrange dimensionality."
>
> "(SACD) has the absence of digital artifacts, same as vinyl and
> pre-recorded tape".
>
> I am sure there are a lot more examples.
>

There aren't...as you must know if you reviewed the post. I was descibing
specifically some of the subjective sonic problems that I and many other
(and many professional recording people) recognize as problems with the CD
playback standard...indeed many here in the objectivist camp acknowledge the
filtering issue with regard to CD playback. I went on to discuss analog
tape, DSD, and PCM, again from a listening standpoint and concluded:

"I'm not sure what your point is here. My issues have always largely been
with the high-end of the CD standard. And secondarily with CD's tendency to
have less depth and a 'flatter' bass and lower midrange dimensionality.
SACD solves these problems wonderfully. So does DVD-A. Both assuming
equipment that allows the advantage to come through, of course."

Where in this do you find me anti-digital or obsessively pro-vinyl? I
didn't even discuss vinyl.

> Also, interesting that when I said I object to something, you
> immediately assume that I was objecting to what you claim, as in here
> and now.
>

???? I don't understand what you are saying.

> >
> >>
> >> >
> >> > Measured by recreating the emotional impact and sense of "realism" of
> > the
> >> > original performance, vinyl still has much to recommend it, given
> > careful
> >> > selection of the components reproducing it.
> >>
> >
> >> You get it all wrong. The creation of emotional impact and realism is
in
> >> the hands of the performers and the producers, and is certainly not
> >> better done via the LP technology. It fact, LP technology is
> >> demonstrably less accurate, by all measures of accuracy.
> >
> > I don't get it wrong at all. The medium that lets whatever is there in
a
> > great performance get through most akin to that same performance heard
live
> > is the superior medium, from a musical standpoint regardless of the
> > technical merits of same. Thus I often prefer vinyl to cd in this
regard,
> > and usually prefer SACD to DVD-A, and either to CD. Based on my
emperical
> > observation that for me, these media offer more examples of this "akin
to
> > live" experience.
>
> Of course, there is the minor problem that digital redbook audio so
> often proves to be much more transparent (like the Lipshitz test found),
> and vinyl is demonstrably inaccurate compared to CD. The point is that
> the realism and emotion should be set by the performers and the
> producers, and the medium has to be as faithful as possible.

As an engineer, I can understand your viewing it that way. But if the goal
is to promote as close as possible the reproduction/emotional response of a
live performance through a long chain of electronics and mechnical devices,
then the most technically accurate in conventional measurements may not be
the one that succeeds the most in "getting out of the way" to allow the
performance to assert itself. As Michael points out, we aren't necessarily
even yet aware that what we are measuring are the most important things,
sonically.

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

Michael Mossey wrote:

> We could dismiss this experience, but consider this thought
experiment:
>
> I am blindfolded and kept in a room. Once per day someone walks in
the
> room and speaks to me. There are a total of ten people and every day
> one of them is chosen at random to enter the room and speak to me. I
> am asked to give a name to each voice. At the end of one year, I am
> tested to see if I can consistently name and differentiate the ten
> people.
>
> I'm willing to bet that I could do it, and that it would be easy.
> Notice that I would have no chance to put one voice up against the
> other in time. This says something about audio memory. I will be
> investigating this and looking to see if it bears any relation to the
> difference between non-comparitive and comparitive listening.

It doesn't, because the difference between human voices is so far above
threshold, and because human voices are sounds that we have learned to
recognize patterns in (unlike, say, partial loudness differences). This
is also why we have little trouble recognizing familiar voices over the
telephone, or even cellphones.

BTW, the same thing is true of music. We have learned to recognize
patterns in music (and some of that recognition--though probably not
much of it--may be hard-wired), but those differences are far above
threshold, so the experience is of no relevance to questions of
audio-level differences.

bob

 

[chung]

Archived from groups: rec.audio.high-end (More info?)

Michael Mossey wrote:
> chung wrote:
>> Michael Mossey wrote:
>> > Chung wrote:
>> >> Michael Mossey wrote:
>> >>
>> >> >
>> >> > I was a student at Caltech 1987 to 1991.
>> >>
>> >> Let me get this straight. You went to Caltech, and you believe
> that
>> >> interconnects may sound different?
>> >
>> > I don't believe that interconnects sound different---I'm just open
> to
>> > it as a possibility, and I see problems in the tests that have
>> > attempted to rule out that possibility.
>>
>> So you don't believe it, but you are open to it as a possibilty. What
>
>> evidence have you encountered that leads to this openness? From
> reviews?
>>
>> BTW, you could be a very busy man, if you are open to possibilities
> that
>> you don't believe in, and you want to test them. Remember the old
>> saying: if you have too open a mind, your brain may fall out.
>>
>> >
>> >> Don't you think that if there were
>> >> audible differences, they will be accompanied by easily
> meaasureable
>> >> differences? So why not make measurements, instead of long DBT's?
> Do
>> > you
>> >> have any theory as to why they may sound different?
>> >
>> > If we knew what to measure, yes. Anyway, don't they all measure
>> > differently? Isn't the question whether the difference is within
> the
>> > threshold of hearing?
>>
>> Well, for starters, frequency response, distortion, and noise added.
> We
>> know that we can only detect about a 0.3 dB difference in level (but
>> most likely only in a quick switching test, I might add) at 1KHz.
> There
>> is really no reason why a mechanically sound cable can introduce
>> distortion or noise, but those impairments can easily be measured
> down
>> to the -100 dB level. Any reason why you think there may be
> impairments
>> from cables that can be heard but not measured?
>>
>> If two cables measure differently, like one is 1 inch longer, you
> think
>> they may sound different? Why do you think that we might not know
> what
>> to measure, when it comes to cables?
>
> I just read in Moore that it is a pretty good assumption that Weber's
> law holds not just for the ear as a whole, but for each "unit" (hair
> cell/neuron) within the ear. That's a pretty good argument that
> experiments to detect the DL (discriminatory limen) on pure tones will
> be relevant to the transfer function on audio equipment.. i.e., if a
> cable is flat to better than one DL then it provides no audible
> distortion.. at least in its transfer function.
>
> Non-linearities in the cable... you say they can be measured down to
> -100 dB.

Actually more like -130 dB or even better...
>
> Note that measuring the transfer function of a LTI system completely
> characterizes it. So does measuring the impulse response. This is
> because the response to a complex signal can be summed from the
> components of that signal, in a linear system.
>
> However, this is not true of a non-linear system. So what would it
> mean to *characterize* the nonlinearity of a non-linear system? In
> what way could we make a few measurements and satisfy ourselves that we
> have knowledge of that system? Satisfy ourselves that we can predict
> that system's response to any signal?
>
> Want to take a shot at this question?

An interconnect cable is actually one of the most linear components that
you will encounter, assuming you are not using it to arc-weld. To prove
its linearity, simply run a distortion test. Next question?

Or are you now assuming that the cable may be time-variant? If so,
please present evidence.

>
>
>>
>> >
>> >>
>> >> Just out of curiosity, what did you major in at Caltech?
>> >
>> > "Engineering and Applied Science." It gave me a good background in
>> > linear time-invariant models of systems.
>>
>> Well, one of my professors started a linear systems course by stating
>
>> that the whole blackboard is the set of all systems. Then he placed a
>
>> chalk dot on the board, and said that the dot represented the set of
> all
>> linear systems. It may be worthy to note that it is almost the
>> definition of reductionism to consider linear time-invariant systems.
>
>> Not that there is anything wrong with being reductionistic.
>>
>> It is really unusual to hear a graduate from Caltech talk about
>> interconnects possibly sounding different, yet have no theory for why
>
>> they may sound different, or believe that they might sound different.
> Or
>> hear him say that there are digital artifacts, but presents no
> evidence.
>
> By the way, science proceeds from the hypothesis stage to the
> experiment stage. I'm in the hypothesis stage. Do you want to stamp
> out competing hypotheses before they can be tested? Isn't that
> unscientific and more like a turf-war?

It does not mean that you have to ignore progress already made. And some
hypotheses are simply without virtue to the point that it is a waste of
time to test them. For instance, if someone says elephants can fly, are
you going to test that hypothesis? Doesn't it make more sense for the
one proposing such a hypothesis to supply evidence? No, someone simply
saying that he has seen elephants flying is not enough evidence. And it
makes no sense to argue that whether elephants fly or not is a function
of one's consciousness...

 

[chung]

Archived from groups: rec.audio.high-end (More info?)

Michael Mossey wrote:
snip
>
> Example: intermodulation distortion interferes with the conscious
> experience of music more than harmonic distortion. So an amplifier
> that has a tiny amount of intermodulation distortion and NO harmonic
> distortion, is less accurate than an amplier that has a lot of harmonic
> distortion and no IM distortion. The amplifier with less distortion in
> terms of numbers is arguably LESS accurate. This shows the danger of
> separating the conscious effect of distortion from the measurement of
> distortion.

The nonlinearlity that generates IM will always generate harmonic
distortion, so your example is a poor one.

In the case of two interconnects, all distortion products are -100dB or
less, frequency response is < 0.1 dB in the audio range, and measured
added noise is -100dB or less. Shouldn't that be enough to convince you
that they sound the same?


>
> Or, as another possibility, a form of distortion could actually
> contribute to replicating a conscious experience. In that case, it
> simply would not be true to claim that the perceived similarity is an
> illusion.

So why would an ABX type test not show positive results?

> If a PET scan could prove the similarity of conscious
> experience, we would have a measureable and objective way to make more
> accurate recordings through the addition of distortion. As I see it,
> our limited ability to measure conscious experience is what prevents us
> from considering this possibitilty-- not that this possibility doesn't
> exist.
>
> Now, if I want to assert that digital is less accurate than analog, of
> course I eventually have to provide evidence that a measureable form of
> distortion exists. Digital has measureable distortion--the question is
> whether it is audible. I know it's not a question for you, but it is
> still an open question for me.

If it is audible, why would an ABX type not show it?

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

uraniumcommittee@yahoo.com wrote:
> Theporkygeorge@aol.com wrote:
> > uraniumcommittee@yahoo.com wrote:
> > > > The preference for tubes is hardly based on the idea that
"older
> is
> > > > better." There may indeed be a tiny band of people that believe
> > this
> > > > but that hardly acounts for the majority of people that prefer
> tube
> > > > amplification.
> > > >
> > >
> > >
> > > What evidence do you have that the majority prefer tube
> > amplification?
> >
> >
> > Where did I ever say that the majority of people prefer tube
> > amplification?
>
> Here:
>
> "There may indeed be a tiny band of people that believe
> this but that hardly acounts for the majority of people that prefer
> tube
> amplification."


I suggest you read it more carefully. I was saying that *amoung those
that prefer tube amps* the majority do not prefer them because they
believe older is better.



> >
> >
> > > I certainly don't, and I have good ears.
> >
> >
> > I certainly do in my system and I have very good hearing.
> >
> >
> > Tubes distort in rather
> > > unpleasant ways.
> >
> >
> > I bit broad but if you find it unpleasant then fine. I find tubes
to
> > sound more like real music for the most part.
>
> Nope.


Wrong. Your claim was broad and I do find tube amps to sound more like
live music by and large.



Added colorations throughout the mid-range especially.




Sorry but that is yet another hasty generalizartion.



Scott Wheeler

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

On 6 Apr 2005 00:50:57 GMT, "Michael Mossey" <michaelmossey@yahoo.com>
wrote:

>By the way, science proceeds from the hypothesis stage to the
>experiment stage. I'm in the hypothesis stage. Do you want to stamp
>out competing hypotheses before they can be tested? Isn't that
>unscientific and more like a turf-war?

Actually, science proceeds from the *observation* stage to the
hypothesis stage to the experiment stage. Can you report any reliable
and repeatable observation which would lead to your hypothesis? Simply
making something up is not science.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

Michael Mossey <michaelmossey@yahoo.com> wrote:
> chung wrote:
> > Michael Mossey wrote:
> > > Chung wrote:
> > >> Michael Mossey wrote:
> > >>
> > >> >
> > >> > I was a student at Caltech 1987 to 1991.
> > >>
> > >> Let me get this straight. You went to Caltech, and you believe
> that
> > >> interconnects may sound different?
> > >
> > > I don't believe that interconnects sound different---I'm just open
> to
> > > it as a possibility, and I see problems in the tests that have
> > > attempted to rule out that possibility.
> >
> > So you don't believe it, but you are open to it as a possibilty. What

> > evidence have you encountered that leads to this openness? From
> reviews?

> Primarily from four blind trials, the very first I ever did, in which I
> felt sure I knew what cable I was hearing, felt that the differences
> were easily audible, and was right four times. That made an
> impression.

> (But winning a big prize the first time you step in a casino makes a
> big impression too.)

> On another occasion, I did eight trials and was wrong three times.


> >
> > BTW, you could be a very busy man, if you are open to possibilities
> that
> > you don't believe in, and you want to test them. Remember the old
> > saying: if you have too open a mind, your brain may fall out.

> Yeah, yeah. And if you have too closed a mind, your brain will get
> claustrophobic.

> >
> > >
> > >> Don't you think that if there were
> > >> audible differences, they will be accompanied by easily
> meaasureable
> > >> differences? So why not make measurements, instead of long DBT's?
> Do
> > > you
> > >> have any theory as to why they may sound different?
> > >
> > > If we knew what to measure, yes. Anyway, don't they all measure
> > > differently? Isn't the question whether the difference is within
> the
> > > threshold of hearing?
> >
> > Well, for starters, frequency response, distortion, and noise added.
> We
> > know that we can only detect about a 0.3 dB difference in level (but
> > most likely only in a quick switching test, I might add) at 1KHz.

> According to the psychoacoustics textbook I'm reading, you are
> right---about discrimination on test tones such as noise or pure tones.
> However, the books also say that based on the information that the
> auditory nerve can carry, the theoretical DL (discriminatory limen)
> could be more like 0.1 dB, it's the way the higher centers process the
> signal that reduces the DL.

> Discriminating test tones is a task very unlike listening for musical
> enjoyment.

True. *They are often more sensitive tests* of difference perception.
Pink noise and test tones commonly allow the testee to discern levels of
difference *smaller* than those achieved with music.

> So I'm going to be carefully examined what conclusions can
> be drawn from psychoacoustical testing and whether it is valid to
> extrapolate them to listening to the vastly more complex musical
> signals. See below..

The conclusion to be drawn from work thus far where the two sorts
of signals have been compared, is that test tones and pink noise
are allow for *more sensitive* discrimination of difference than
music does.

Now I will play Kreskin and predict your response:
"They tested for the wrong differences."



--

-S
It's not my business to do intelligent work. -- D. Rumsfeld, testifying
before the House Armed Services Committee

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

Steven Sullivan wrote:
> Michael Mossey <michaelmossey@yahoo.com> wrote:
> > nabob33@hotmail.com wrote:
> > > Michael Mossey wrote:
> > >
> > > > Don't get confused between subjective accuracy and objective
> > > accuracy.
> > >
> > > You're the one who's confused, or trying to confuse. There is no
such
> > > thing as "subjective accuracy," as you define it. There may be a
> > > "subjective sense of similarity," but to appropriate a technical
term
> > > like accuracy to elevate this concept is to muddy the waters, not
> > > clarify them.
>
> > Let me just explain the danger as I see it... I'm referring to the
> > danger of placing objective measurements above conscious
experience.
>
> Rather than spend time on that, how about you consider perception
> of audible difference in light of this undisputable fact:
>
> The conclusions about the physical world drawn from 'conscious
experience' are
> often inaccurate if they rely on only sense perception, belief, and
memory.

Steven, the conclusions of psychoacoustical research and blind tests
are "drawn from conscious experience." The subject has to be conscious
of something in order to report it.


In "Memory and Attention", Norman: "We would
like to peel back the different levels one by one, starting with an
understanding of the sensory organs, moving through sensation and
perception and finally ending at decision making and thinking.
... One problem is that no process can be analyzed in isolation. We
can, for example, analyze sensations only through the responses
made by our experimental subjecs, and these responses must be the
result of the whole structure of their psychological processes, from
sensation through decision making."

>
> Optical illusions are simple examples.
> The 'phantom switch' -- where nothing is changed, but the listener
> thinks it has -- is a slightly more complex one. The fallibility
> of 'eyewitness testimony' is well-known to members of the scientific,
> legal and law enforcement communities.

Illusions are possible, of course. What you think you hear doesn't
necessarily match what is coming in your ears. But that applies
equally to blind listening as to sighted listening.

>
>
> > The whole point of audio is to duplicate a conscious experience.
And
> > some people are concerned with simply creating the most enjoyable
> > conscious experience possible, while I'm concerned mostly with
> > replicating the conscous experience of live music. So "accuracy"
or
> > "similarity" or whatever you want to call it--this concept should
apply
> > first to similarity of conscious experience.
>
> Accuracy of memory would be a problem here.

Let's do a thought experiment. Do you think you could tell,
blindfolded, the difference between the sound of a friend speaking to
you, and the playback of your friend's voice over a stereo system?

If you think that you can, you are successfully judging the similarity
to a memory of an aural experience.

>
> > It is certainly easier to demonstrate that two signals measure
> > similarly than it is to demonstrate that two conscious experiences
are
> > the same. But this should not place measurements above conscious
> > experience. Because .... (continued below)
>
>
>
>
> > >
> > > > It SOUNDS more accurate;
> > >
> > > But it IS less accurate. Therefore, this seeming "accuracy" is
likely
> > > some form of illusion.
>
> > because it is too easy to apply the label "illusion" to conscious
> > experiences that seem to contradict the measurements, without
actually
> > having demonstrated anything about consciousness.
>
> By this criterion, complete fantasy becomes as legitimate a 'report'
of
> reality as a model derived from careful observation. Does that
matter
> to you?

I don't follow you. I believe in formulating models and testing them
scientifically. I'm saying that consciousness ought to be one of the
things we model and test. We can't separate it from the audio
experience.

>
>
> > Example: intermodulation distortion interferes with the conscious
> > experience of music more than harmonic distortion. So an amplifier
> > that has a tiny amount of intermodulation distortion and NO
harmonic
> > distortion, is less accurate than an amplier that has a lot of
harmonic
> > distortion and no IM distortion. The amplifier with less
distortion in
> > terms of numbers is arguably LESS accurate. This shows the danger
of
> > separating the conscious effect of distortion from the measurement
of
> > distortion.
>
> It actually simply shows the danger of misinterpreting measurements.

Right.. measurements must be interpreted in light of what we know about
the ear/brain.

> Now, how has it been determined that IM distortion interferes with
> the conscious experience of music more than harmonic distortion?

Probably from blind listening experiences. I just know it is a common
and undisputed statement, even in the psychoacoustics textbook I have.

>
> > Or, as another possibility, a form of distortion could actually
> > contribute to replicating a conscious experience. In that case, it
> > simply would not be true to claim that the perceived similarity is
an
> > illusion. If a PET scan could prove the similarity of conscious
> > experience, we would have a measureable and objective way to make
more
> > accurate recordings through the addition of distortion. As I see
it,
> > our limited ability to measure conscious experience is what
prevents us
> > from considering this possibitilty-- not that this possibility
doesn't
> > exist.
>
> As I see it, you are striving mightily to find an excuse not to
believe
> a more straightforward explanation -- that some 'differences' are
imaginary.

I do believe that some differences are imaginary. Kind of takes the
wind out of the sails of that point, huh?

-Mike

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

Stewart Pinkerton wrote:
> On 4 Apr 2005 23:58:00 GMT, "Michael Mossey"
<michaelmossey@yahoo.com>
> wrote:
>
> >The whole point of audio is to duplicate a conscious experience.
>
> Some would say that it is to reproduce the original sound. You are
> attempting to overcomplicate what it a simple goal - if difficult to
> realise.

Here's an example where the goal is reproduce a sound. At JPL we
bombard spacecraft with very loud noise to simulate the environment
inside a launching spacecraft. We are trying to reproduce the sound of
launch. Now, it doesn't matter whether anyone hears that sound or what
they think of it.

It is the conscious experience of music that inspires us to try to
reproduce that conscious experience. I don't see how you can disagree
with that statement.

Sure, it is very useful during design to work with an audio system
isolated from the human experience of it. But you had better check how
it sounds before you're done, or else you are essentially playing
tennis without a net. See below.

>
> You accuse others of 'reductionism', I accuse you of sophistry.

It is clear that you use simple reasoning and simple models. I think
that doesn't cut it, if you wish to know something about audio.

>
> >Example: intermodulation distortion interferes with the conscious
> >experience of music more than harmonic distortion. So an amplifier
> >that has a tiny amount of intermodulation distortion and NO harmonic
> >distortion, is less accurate than an amplier that has a lot of
harmonic
> >distortion and no IM distortion. The amplifier with less distortion
in
> >terms of numbers is arguably LESS accurate. This shows the danger
of
> >separating the conscious effect of distortion from the measurement
of
> >distortion.
>
> You claim to have graduated from Caltech, majoring in Engineering,
and
> you believe that such amplifiers are possible? Hmmmm

Don't you know what a thought experiment is? Whether those amplifiers
can be built is not relevant to the point of the thought experiment.

>
> Besides, all you argument above shows, is that THD is not the best
> indicator of audible distortion. We knew that already.

And *how* did we learn that? By listening! What my argument shows is
that you can't isolate audio engineering from perception.

Or do you propose to say that a bunch of literally deaf-since-birth
engineers who never let anyone listen to their designs somehow got the
idea that THD is not the best indicator of audible distortion?

>
> >Now, if I want to assert that digital is less accurate than analog,
of
> >course I eventually have to provide evidence that a measureable form
of
> >distortion exists. Digital has measureable distortion
>
> Does it? What distortion is that?

Any difference that a very sensitive instrument reports between the
input and output of a ADC/DAC.

-Mike

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

Stewart Pinkerton wrote:
> On 4 Apr 2005 23:53:56 GMT, "Michael Mossey"
<michaelmossey@yahoo.com>
> wrote:
>
> >There's really no need to choose between "analog is
accurate"/"analog
> >is euphonic". The answer might depend on the individual. I think
it
> >is reductionist to answer this question in one way that applies to
all
> >individuals.
>
> You have a love of the term 'reductionism',
> but you seem to use it as
> a convenient excuse for anything with which you disagree.

Ad hominen and strawman.

> The physical
> soundfield is reality, and it either is, or is not, an accurate
> representation of the soundfield at a particular listening position
at
> the original live performance (obviously, you can't use 'accuracy'
> when referring to a studio recording).

With two channel audio, I think it is pretty undisputed that the
soundfield is NOT an accurate representation.. i.e. most anyone could
tell whether they were hearing live music or a reproduction, not to
mention that two channels can't reproduce the wavefront properly from
all directions.

So the question becomes, "Out of two inaccurate soundfields which is
more accurate?" This question must be answered by *listening.* I read
the Harman white paper by Floyd Toole describing their experiments with
loudspeakers and listening panels. Guess what? They had people
*listen* and rate preference, and then used the result of listening to
formulate a model of speaker preference (or accuracy, the concepts
overlap somewhat here). They didn't try to formulate the model before
anyone listened.

>
> There *is* a need to choose between "analogue is accurate" and
> "analogue is euphonic", since this refers to the physical soundfield,
> not to an individual impression.

It looks like this statement is based on your idea that you can
separate engineering and listening, but I'm not sure what you saying.
It is clear that we have different reactions to analog: it sounds like
"air" and "warmth" to you, while it sounds like realism to me.

-Mike

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

Stewart Pinkerton wrote:
> On 4 Apr 2005 23:51:31 GMT, "Michael Mossey"
<michaelmossey@yahoo.com>
> wrote:
>
> >Stewart Pinkerton wrote:
>
> >Playback of a stored digital signal
> >is a different system than a live digital feed.
>
> This is a mere assertion, with no basis in fact.

Look at the systems end-to-end. One system has storage media in it,
the other doesn't. Hence they are different systems.

Now, I'm not claiming that storage of digital data degrades it. I'm
not claiming any specific method of degradation. I'm just pointing out
that changing the system has the potential of changing the behavior in
a way we don't expect.

>
> > Therefore it might
> >behave differently. I still think that we need to search for
evidence
> >that it behaves differently. Within a model of that system there
might
> >be no way that it could behave differently. But that model is not
> >reality. It is "reductionist" to draw conclusions about reality
from
> >the basis of a model. Not necessarily invalid, but still
reductionist.
> > At the moment I'm simply being careful about what conclusions can
be
> >drawn from a live digital feed test.
>
> The system does not differ, whether the source material is a live
feed
> or stored data. Hence, it does *not* behave differently. Please read
> up on digital audio before making such assertions.
>
>
>
> >> Your hypothesis is not tenable until you offer some *evidence* to
> >> support it. You have been asked several times exactly what
'digital
> >> artifacts' you are referencing, but you have never answered this
> >> simple question.
>
> >I don't know what digital artifacts. Could be something in a
mechanism
> >already modelled, such as jitter, or it could be something new. I
> >could be wrong that digital has artifacts--perhaps I like analog
better
> >for its distortion.
>
> That is a *much* more likely scenario.
>
> > In the words of Skeptic Magazine, "Maybe yes,
> >maybe no."
>
> Skeptic Magazine is not reality............
>
> If you want a better guide to reality than Skeptic Magazine, try
> Occam's Razor.

It's my experience that the ability to say "Maybe yes, maybe no" is an
important part of critical thinking and an important part of moving
toward the truth.

Skeptic Magazine just published an article on audio, by Ethan Winer,
that you would probably love. It agrees with you in every respect.

-Mike

 

[Guest]

Archived from groups: rec.audio.high-end (More info?)

"Michael Mossey" <michaelmossey@yahoo.com> wrote in message
news:d2skes0677@news4.newsguy.com...

> You could also argue that I'm attacking anything I disagree with by
> claiming it is uncertain, such as a criminal defense attorney would do
> to create a "reasonable doubt." I don't think I am, because I think
> that it is the fact that *consciousness* is involved in audio research
> that makes reductionism dangerous. (I think reductionism is a
> wonderful way to make a better digital network.) And because I am
> willing to change my mind when I eventually fail to find the evidence
> to support my hypothesis. When searching for evidence, I will weight
> particularly highly tests that try to account for the mechanism of
> forming conscious musical impressions.

How do you decide when you have "failed to find" evidence, or
anything for that matter? It seems rather open-ended.

- Gary Rosen