Big differences between 44.1 and 96Khz. Why?
philicorda
audible difference between recording at 44.1 and 96Khz. When I have
experimented with average converters, both sound fairly similar.
But... the other day a friend brought in a Sony PCM-D50 and claimed there
was a huge difference at 96K. We tried a few bind tests using the
internal mics on voice and piano, and sure enough the 96Khz sounded
obviously 'better'. Clearer, more present etc.
The listening setup was nothing special, a Teac A-X3000 amplifier and
Tannoy Mercury MX2 speakers in an average living room.
I'm not totally convinced that the extra bandwidth should make so much
difference, so what else might be going on, and how can I test for it?
Peter Larsen
Antialias filtering and/or aliasing is my unskilled guess, alias-tones cause
sonic havoc by not having any harmonic relationship with the audio.
Kind regards
Peter Larsen
William Sommerwerck
anti-aliasing filter.
Find a microphone with a relatively limited bandwidth -- say, one that cuts
off abruptly above 12kHz. There "should" be less audible difference between
the sampling rates.
Similarly, a wideband condenser mic -- say, one going to 25kHz or higher --
"should" show greater audible differences.
Scott Dorsey
>I've always been of the opinion that there should not be an obvious
>audible difference between recording at 44.1 and 96Khz. When I have
>experimented with average converters, both sound fairly similar.
>
>But... the other day a friend brought in a Sony PCM-D50 and claimed there
>was a huge difference at 96K. We tried a few bind tests using the
>internal mics on voice and piano, and sure enough the 96Khz sounded
>obviously 'better'. Clearer, more present etc.
Watch out! Almost always when I hear things sound "better" in that way,
more careful listening tests show that it actually sounds worse.
>The listening setup was nothing special, a Teac A-X3000 amplifier and
>Tannoy Mercury MX2 speakers in an average living room.
>
>I'm not totally convinced that the extra bandwidth should make so much
>difference, so what else might be going on, and how can I test for it?
All kinds of stuff can be going on, from clocks that are stable enough to
use at 44.1 but not 96 to resampling issues, bum anti-aliasing filters
(a major problem in 1985 but these days hardly ever an issue thanks to
sigma-delta converters) to digital noise leaking into the analogue signal
and causing beat products.
Tell me what model the front end A/D chips are and I'll see which of these
are likely suspects.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Mark
>
> But... the other day a friend brought in a Sony PCM-D50 and claimed there
> was a huge difference at 96K. We tried a few bind tests using the
> internal mics on voice and piano, and sure enough the 96Khz sounded
> obviously 'better'. Clearer, more present etc.
>
...
I think a measurement is call for here..
Mark
Moshe
I've heard differences before.
Would I be able to consistently pick them out in a double blind
test?
Doubtful.
cedricl
> I've always been of the opinion that there should not be an obvious
> audible difference between recording at 44.1 and 96Khz. When I have
> experimented with average converters, both sound fairly similar.
>
> But... the other day a friend brought in a Sony PCM-D50 and claimed there
> was a huge difference at 96K.
I've found a greater difference in sound quality occurs when going
from 16 bit to 24 bit regardless of sampling rate.
Moshe
Me too.
William Sommerwerck
> Doubtful.
Since when is double-blind testing the standard for making such evaluations?
Scott Dorsey
>On Thu, 13 May 2010 19:44:33 -0700 (PDT), Mark wrote:
>>> But... the other day a friend brought in a Sony PCM-D50 and claimed there
>>> was a huge difference at 96K. We tried a few bind tests using the
>>> internal mics on voice and piano, and sure enough the 96Khz sounded
>>> obviously 'better'. Clearer, more present etc.
>>
Which one, though?
>
>I've heard differences before.
>Would I be able to consistently pick them out in a double blind
>test?
>Doubtful.
You might. Try a Panasonic SV-3700 and listen to the difference between
44.1 ksamp/sec and 48 ksamp/sec operation.... they use the same anti-aliasing
filters for both and there is substantial aliasing at 44.1. Mind you, it
sounds dreadful at 48, but it sounds even worse at 44.1. The effect is not
subtle.
Digital conversion is pretty much a solved problem, but not all implementations
are good. Some of the earlier ones were just godawful.
Arny Krueger
news:Dc%Gn.67091$x54....@newsfe10.ams2
> I've always been of the opinion that there should not be
> an obvious audible difference between recording at 44.1
> and 96Khz. When I have experimented with average
> converters, both sound fairly similar.
> But... the other day a friend brought in a Sony PCM-D50
> and claimed there was a huge difference at 96K. We tried
> a few bind tests using the internal mics on voice and
> piano, and sure enough the 96Khz sounded obviously
> 'better'. Clearer, more present etc.
I'll bet money that your tests were neither time-synched nor level-matched.
People have been testing this for about a decade, using really carefully run
tests.
There's an extant JAES paper:
http://drewdaniels.com/audible.pdf
Discussed further here:
http://www.aes.org/journal/online/comment/?ID=14195&type=report
Audibility of a CD-Standard A/DA/A Loop Inserted into High-Resolution Audio
Playback
Authors:Meyer, E. Brad; Moran, David R.
Affiliation:Boston Audio Society, Lincoln, MA, USA
JAES Volume 55 Issue 9 pp. 775-779; September 2007
[Engineering Report] Claims both published and anecdotal are regularly made
for audibly superior sound quality for two-channel audio encoded with longer
word lengths and/or at higher sampling rates than the 16-bit/44.1-kHz CD
standard. The authors report on a series of double-blind tests comparing the
analog output of high-resolution players playing high-resolution recordings
with the same signal passed through a 16-bit/44.1-kHz "bottleneck." The
tests were conducted for over a year using different systems and a variety
of subjects. The systems included expensive professional monitors and one
high-end system with electrostatic loudspeakers and expensive components and
cables. The subjects included professional recording engineers, students in
a university recording program, and dedicated audiophiles. The test results
show that the CD-quality A/D/A loop was undetectable at normal-to-loud
listening vels, by any of the subjects, on any of the playback systems. The
noise of the CD-quality loop was audible only at very elevated levels.
There is more information about the listening tests here:
http://www.bostonaudiosociety.org/explanation.htm
The above tests were hardly unqiue. People have been failing to show an
audible difference between 44.1 and 96 KHz sampling for several decades.
The question you have to answer for yourself is whether the results you
obtained represented some dramatic insight or correction that suddenly
corrected the many dozens of past failures, or whether your own tests were
somehow less exact then the many that went before it and provided a
different result.
Arny Krueger
message news:hsjcdf$l1o$1...@news.eternal-september.org
Since no later than 1983:
http://www.aes.org/e-lib/browse.cfm?elib=3839
William Sommerwerck
news:66qdncIl0oiHx3DW...@giganews.com...
All this is is a paper about how to correctly set up and evaluate ABX
testing. There is no AES standard for equipment evaluation (see above), ABX
or otherwise.
This is like saying that because someone publishes a paper about evidence
for the existence of Atlantis, that Atlantis's existence is, ipso facto,
accepted by archaeologists.
Arny Krueger
> "Arny Krueger" <ar...@hotpop.com> wrote in message
> news:66qdncIl0oiHx3DW...@giganews.com...
>> "William Sommerwerck" <grizzle...@comcast.net> wrote
>> in message news:hsjcdf$l1o$1...@news.eternal-september.org
>
>>>> I've heard differences before. Would I be able to
>>>> consistently pick them out in a double-blind test?
>>>> Doubtful.
>
>>> Since when is double-blind testing the standard for
>>> making such evaluations?
>
>> Since no later than 1983:
>> http://www.aes.org/e-lib/browse.cfm?elib=3839
> All this is is a paper about how to correctly set up and
> evaluate ABX testing.
If you actually read the citation deeply enough to perceive its title, you
will find that it is a paper about how to set up and run "High-Resolution
Subjective Testing". If you are interested in doing low resolution
subjective testing (e.g., sighted evaluations) then you can persist in not
following up on the link I provided.
> There is no AES standard for
> equipment evaluation (see above), ABX or otherwise.
The non-existence of a formal organizational standard does not disprove or
in any way detract from the existence of a de facto standard.
There is also the EBU official document BS 1116:
http://www.tech.plymouth.ac.uk/spmc/pdf/audio/ITU/BS1116.pdf
"To conduct subjective assessments in the case of systems generating small
impairments, it is necessary to select an
appropriate method. The �double-blind triple-stimulus with hidden reference�
method has been found to be especially
sensitive, stable and to permit accurate detection of small impairments.
Therefore, it should be used for this kind of test."
There *are* AES standards and recommendations for doing listening tests.
So, your basic statement is just plain wrong.
William Sommerwerck
>>>>> consistently pick them out in a double-blind test?
>>>>> Doubtful.
>>>> Since when is double-blind testing the standard for
> >>> making such evaluations?
>>> Since no later than 1983:
>>> http://www.aes.org/e-lib/browse.cfm?elib=3839
>> All this is is a paper about how to correctly set up and
>> evaluate ABX testing.
> If you actually read the citation deeply enough to perceive its title, you
> will find that it is a paper about how to set up and run "High-Resolution
> Subjective Testing".
David Clark's OPINION about what comprises such testing.
> If you are interested in doing low resolution subjective testing (e.g.,
> sighted evaluations) then you can persist in not following up on the
> link I provided.
Double-blind testing is a form of subjective evaluation.
>> There is no AES standard for
>> equipment evaluation (see above), ABX or otherwise.
> The non-existence of a formal organizational standard does not disprove or
> in any way detract from the existence of a de facto standard.
> There is also the EBU official document BS 1116:
> http://www.tech.plymouth.ac.uk/spmc/pdf/audio/ITU/BS1116.pdf
> There *are* AES standards and recommendations for doing listening tests.
> So, your basic statement is just plain wrong.
This is a recommendation, not a requirement.
Frank Stearns
- snips -
>The above tests were hardly unqiue. People have been failing to show an
>audible difference between 44.1 and 96 KHz sampling for several decades.
What about retaining higher sampling rates *all the way through* production (mix,
comp, eq, etc), versus going all the way through at 44.1?
We do roughly half and half here. 44.1/24 can sound really good, but 88.2 or 96 has
something more to offer (even after the final "step down" to 44.1/16 for the CD).
The difference is enough to overcome some of the logistical hassles associated with
doing higher rates.
I'd love to keep it simple and just run everything at 44.1/24 -- but for the
higher-end classical projects, we almost always use higher rates. You can usually
hear differences more readily with that kind of instrumentation/vocals than with
something full of intentional distortion (guitar FX, for example), untrained voices
on SM58s, and other "effects" from the rock stage.
Now, can the average consumer hear a diff on average speakers in an average room?
Probably not. But in a good room on a good system, it's often not all that subtle.
Does it matter? Matters to me; matters to some of my classical clients.
Is 44.1/16 theoretically possible of complete transparancy within the limits of
human perception? Certainly that's likely, though from the get-go one is close to
the edge.
As a practical matter, though, filter issues, arithmetic issues, clocks, DAC/ADC
linearities (those that remain), and how well all those things are implemented in
the _real world_, can show their issues at the different rates. IMO.
Frank
Mobile Audio
--
.
George Orwell
Since Gordon passed away!
Il mittente di questo messaggio|The sender address of this
non corrisponde ad un utente |message is not related to a real
reale ma all'indirizzo fittizio|person but to a fake address of an
di un sistema anonimizzatore |anonymous system
Per maggiori informazioni |For more info
https://www.mixmaster.it
Moshe
I never said it was.
Moshe
> Moshe <goldee._l...@gmail.com> wrote:
>>On Thu, 13 May 2010 19:44:33 -0700 (PDT), Mark wrote:
>>>> But... the other day a friend brought in a Sony PCM-D50 and claimed there
>>>> was a huge difference at 96K. We tried a few bind tests using the
>>>> internal mics on voice and piano, and sure enough the 96Khz sounded
>>>> obviously 'better'. Clearer, more present etc.
>>>
>>> I think a measurement is call for here..
>
> Which one, though?
>>
>>I've heard differences before.
>>Would I be able to consistently pick them out in a double blind
>>test?
>>Doubtful.
>
> You might. Try a Panasonic SV-3700 and listen to the difference between
> 44.1 ksamp/sec and 48 ksamp/sec operation.... they use the same anti-aliasing
> filters for both and there is substantial aliasing at 44.1. Mind you, it
> sounds dreadful at 48, but it sounds even worse at 44.1. The effect is not
> subtle.
Interesting. I was using a Tascam DA-30 at the time and to *me*
44.1 sounded better than 48k. I never knew why though.
I also had a DA-20 which I thought sounded terrible at any
sampling rate.
I think it was the error correction that wasn't working correctly.
Odd thing I remember was I had a Sony consumer DAT deck, albeit
from the ES series, and it sounded better than either of the
Tascams at the time.
> Digital conversion is pretty much a solved problem, but not all implementations
> are good. Some of the earlier ones were just godawful.
> --scott
Agreed.
Scott Dorsey
>On 14 May 2010 09:39:39 -0400, Scott Dorsey wrote:
>
>> Moshe <goldee._l...@gmail.com> wrote:
>>>On Thu, 13 May 2010 19:44:33 -0700 (PDT), Mark wrote:
>>>>> But... the other day a friend brought in a Sony PCM-D50 and claimed there
>>>>> was a huge difference at 96K. We tried a few bind tests using the
>>>>> internal mics on voice and piano, and sure enough the 96Khz sounded
>>>>> obviously 'better'. Clearer, more present etc.
>>>>
>>>> I think a measurement is call for here..
>>
>> Which one, though?
>>>
>>>I've heard differences before.
>>>Would I be able to consistently pick them out in a double blind
>>>test?
>>>Doubtful.
>>
>> You might. Try a Panasonic SV-3700 and listen to the difference between
>> 44.1 ksamp/sec and 48 ksamp/sec operation.... they use the same anti-aliasing
>> filters for both and there is substantial aliasing at 44.1. Mind you, it
>> sounds dreadful at 48, but it sounds even worse at 44.1. The effect is not
>> subtle.
>
>Interesting. I was using a Tascam DA-30 at the time and to *me*
>44.1 sounded better than 48k. I never knew why though.
The DA-30 was much better in that regard, although you should know that there
were at least three different electronics revisions of the DA-30 that were
all somewhat different.
>I also had a DA-20 which I thought sounded terrible at any
>sampling rate.
>I think it was the error correction that wasn't working correctly.
I find the converters in the DA-20 aren't all that offensive, although I
can pretty clearly tell the difference between the internal converters and
the external Prism.
>Odd thing I remember was I had a Sony consumer DAT deck, albeit
>from the ES series, and it sounded better than either of the
>Tascams at the time.
And that's basically how things work. You pays your money and you takes
your chance.
William Sommerwerck
>>> consistently pick them out in a double-blind test?
>>> Doubtful.
>> Since when is double-blind testing the standard for
>> making such evaluations?
> Since Gordon passed away!
I hope that's a light-hearted joke -- especially coming from someone who
refuses to use his real name -- as Gordon was a good friend whom I miss very
much.
It is arguable that, more than any other single person, J Gordon Holt is
responsible for the overall high quality of modern audio components. He was
the first person to publish a magazine that said, in no uncertain terms,
that there is a difference between good and bad sound reproduction (the
"good" being that which comes closest to live sound), and that these
differences are not usually measurable. He helped create a market for
better-quality equipment, which in turn made it possible for designers to
improve their products, as there was a greater chance of selling them.
Of course, I can't prove cause and effect. Engineers are always trying to
improve their designs, and at least some of these improvements would have
occurred without "The Stereophile"'s influence. But when you purchase
modestly priced equipment that delivers very high performance, it's not just
because it was made in China -- it's partly because Gordon was informing
readers and kicking the industry.
Gordon and I didn't agree on everything, but he was one of the few genuinely
thoughtful, intelligent people I've known. He also had a great sense of
humor, something that most people in the audio industry are sorely lacking.
William Sommerwerck
>>> consistently pick them out in a double-blind test?
>>> Doubtful.
>> Since when is double-blind testing the standard
>> for making such evaluations?
> I never said it was.
Then why did you suggest D-BT as the testing protocol?
Moshe
I didn't suggest it as the testing protocol.
I only commented on *my* experiences.
Moshe
> And that's basically how things work. You pays your money and you takes
> your chance.
> --scott
Ain't life sweet !!
Haha!
Laurence Payne
<grizzle...@comcast.net> wrote:
>It is arguable that, more than any other single person, J Gordon Holt is
>responsible for the overall high quality of modern audio components. He was
>the first person to publish a magazine that said, in no uncertain terms,
>that there is a difference between good and bad sound reproduction (the
>"good" being that which comes closest to live sound), and that these
>differences are not usually measurable. He helped create a market for
>better-quality equipment, which in turn made it possible for designers to
>improve their products, as there was a greater chance of selling them.
Which could be rewritten as: "He created the audiophool market".
Neither statement prove anything much.
William Sommerwerck
>> is responsible for the overall high quality of modern audio components.
>> He was the first person to publish a magazine that said, in no uncertain
>> terms, that there is a difference between good and bad sound reproduction
>> (the "good" being that which comes closest to live sound), and that these
>> differences are not usually measurable. He helped create a market for
>> better-quality equipment, which in turn made it possible for designers to
>> improve their products, as there was a greater chance of selling them.
> Which could be rewritten as: "He created the audiophool market".
> Neither statement prove anything much.
I was going to say something about that (because "Stereophile" /was/
indirectly responsible for a lot of the idiocy that's sullied the audiophile
industry), bur decided not to because I knew someone else would.
Gordon was a fairly sober reviewer. He did not recommended expensive cables,
or hockey pucks for your amplifier, or things of that sort. He once did an
article about the likelihood of whether many of these esoteric (to put it
kindly) beliefs were plausible. He came down largely on the side of "not
plausible".
Now, if you think that anyone who sits down carefully and tries to make a
thoughtful choice of audio equipment is an "audiophool" -- then I have a
bone to pick.
Sean Conolly
news:FomdnazJwehd8nDW...@posted.palinacquisition...
> "Arny Krueger" <ar...@hotpop.com> writes:
>
> - snips -
>
>>The above tests were hardly unqiue. People have been failing to show an
>>audible difference between 44.1 and 96 KHz sampling for several decades.
>
> What about retaining higher sampling rates *all the way through*
> production (mix,
> comp, eq, etc), versus going all the way through at 44.1?
>
> We do roughly half and half here. 44.1/24 can sound really good, but 88.2
> or 96 has
> something more to offer (even after the final "step down" to 44.1/16 for
> the CD).
> The difference is enough to overcome some of the logistical hassles
> associated with
> doing higher rates.
>
> I'd love to keep it simple and just run everything at 44.1/24 -- but for
> the
> higher-end classical projects, we almost always use higher rates. You can
> usually
> hear differences more readily with that kind of instrumentation/vocals
> than with
> something full of intentional distortion (guitar FX, for example),
> untrained voices
> on SM58s, and other "effects" from the rock stage.
Now this is the point that I'd like to see in a double-blind test: given a
final output of 44.1/16 CD format, can listeners constistantly hear an
improvement if the audio was processed at a higher sampling rate up to the
final conversion? My intuition is that under tightly controlled conditions
you wouldn't see a significant correlation to the the listeners perception.
I think the only likely reason for there to be any improvement is if the
processing software itself simply works better at the higher sample rate.
Sean
Scott Dorsey
>Now this is the point that I'd like to see in a double-blind test: given a
>final output of 44.1/16 CD format, can listeners constistantly hear an
>improvement if the audio was processed at a higher sampling rate up to the
>final conversion? My intuition is that under tightly controlled conditions
>you wouldn't see a significant correlation to the the listeners perception.
No. And vice-versa... if you have an output at 96 ksamp/sec, and you do
some internal work at 44.1, listeners can't consistently hear anything.
There was a study showing this in the JAES recently.
>I think the only likely reason for there to be any improvement is if the
>processing software itself simply works better at the higher sample rate.
The only case where I can think of that being the case would be for click
and pop reduction for disc transcription, where the ultrasonic content can
make detecting the exact beginning and end of a pop more accurate.
My personal suspicion is that all of the audible differences between high
sample rates have to do with conversion artifacts and extended bandwidth
producing beat notes in the audible band. But by the same token,
if going to a different rate alters the artifacts in a good way, why not
just go with it?
Mark
Again this is a "tastes great..less filling" kind of argument.
The ear/brain vs the machine.
I am amazed at what the ear/brain can do. It is amazing to me that the
ear/brain can discern not simply the meaning, but also the EMOTIONS in
spoken voice. No machine can come close to doing that. Music
looks like a jumbled mess on a spectrum analyzer, yet the ear/brain
has no trouble at all following the individual voices and instruments.
There are many examples where the ear/brain excels.
On the other hand I am also amazed that an Audio Precision analyzer
can quantify harmonics and intermod at -120dBc or detect a frequency
response error or 0.25 dB. No ear/brain can do that.
Each tool has its place.
The task at hand is to identify the right tool for the right job.
Mark
Arny Krueger
>>>>>> I've heard differences before. Would I be able to
>>>>>> consistently pick them out in a double-blind test?
>>>>>> Doubtful.
>
>>>>> Since when is double-blind testing the standard for
>>>>> making such evaluations?
>
>>>> Since no later than 1983:
>>>> http://www.aes.org/e-lib/browse.cfm?elib=3839
>
>>> All this is is a paper about how to correctly set up and
>>> evaluate ABX testing.
>
>> If you actually read the citation deeply enough to
>> perceive its title, you will find that it is a paper
>> about how to set up and run "High-Resolution Subjective
>> Testing".
> David Clark's OPINION about what comprises such testing.
It's a refereed paper, which means that it had the editorial board of the
AES standing behind it.
>> If you are interested in doing low resolution subjective
>> testing (e.g., sighted evaluations) then you can persist
>> in not following up on the > link I provided.
> Double-blind testing is a form of subjective evaluation.
A simple truism, taken from the title of the paper.
>>> There is no AES standard for
>>> equipment evaluation (see above), ABX or otherwise.
>
>> The non-existence of a formal organizational standard
>> does not disprove or in any way detract from the
>> existence of a de facto standard.
>> There is also the EBU official document BS 1116:
>> http://www.tech.plymouth.ac.uk/spmc/pdf/audio/ITU/BS1116.pdf
>
>> There *are* AES standards and recommendations for doing
>> listening tests. So, your basic statement is just plain
>> wrong.
> This is a recommendation, not a requirement.
You're clutching at straws. In the academic and professional worlds, DBTs
are the gold standard for subjective tests.
Arny Krueger
message
news:FomdnazJwehd8nDW...@posted.palinacquisition
>
> - snips -
>
>> The above tests were hardly unqiue. People have been
>> failing to show an audible difference between 44.1 and
>> 96 KHz sampling for several decades.
>
> What about retaining higher sampling rates *all the way
> through* production (mix, comp, eq, etc), versus going
> all the way through at 44.1?
Once an audio signal has been converted to digital, that signal picks up
zero noise and distortion unless they are intentionally added, or the data
is resampled. Therefore it doesn't matter whether the lower sample rate is
applied at the beginning, the end or someplace in-between.
> We do roughly half and half here. 44.1/24 can sound
> really good, but 88.2 or 96 has something more to offer
> (even after the final "step down" to 44.1/16 for the CD).
> The difference is enough to overcome some of the
> logistical hassles associated with doing higher rates.
If large amounts of nonlinear distoriton are added to the signal while it is
in the digital domain, higher sample rates can have an advantage, because
there are some aliasing-type artifacts that are picked up when nonlinear
distortion is added in the digital domain. This is one reason why there can
be audiable advantages to doing synthesis of sounds with very high sample
rates.
> I'd love to keep it simple and just run everything at
> 44.1/24 -- but for the higher-end classical projects, we
> almost always use higher rates. You can usually hear
> differences more readily with that kind of
> instrumentation/vocals than with something full of
> intentional distortion (guitar FX, for example),
> untrained voices on SM58s, and other "effects" from the
> rock stage.
> Now, can the average consumer hear a diff on average
> speakers in an average room? Probably not. But in a good
> room on a good system, it's often not all that subtle.
A zillion people have said that, and as a rule they can't do enough
well-designed listening tests to support their claims.
> Does it matter? Matters to me; matters to some of my
> classical clients.
How can we be sure that it just isn't in their heads?
> Is 44.1/16 theoretically possible of complete
> transparancy within the limits of human perception?
When tested well, it always seems to work out that way.
Sean Conolly
news:hsk8t6$d3t$1...@panix2.panix.com...
Or even if it just alters the client's expectation of the sound. If the
client thinks it sounds better, that's good enough for me.
Sean
William Sommerwerck
> It's a refereed paper, which means that it had the editorial board
> of the AES standing behind it.
Oh, I am so impressed. The deaf leading the deaf.
>> This is a recommendation, not a requirement.
> You're clutching at straws. In the academic and professional
> worlds, DBTs are the gold standard for subjective tests.
And what sort of objective, useful information do they provide?
Moshe
Well for one thing, when done properly, they can be a very good
tool for exposing people who claim they can hear differences in
sound quality when they wrap marbles around their line cords or
magic marker their CD's etc.
William Sommerwerck
>>> worlds, DBTs are the gold standard for subjective tests.
>> And what sort of objective, useful information do they provide?
> Well for one thing, when done properly, they can be a very good
> tool for exposing people who claim they can hear differences in
> sound quality when they wrap marbles around their line cords or
> magic marker their CDs, etc.
DBTs reveal what is and is not audible under the conditions of D-B testing.
This may or may not have any relation with what one hears when one sits down
to listen.
Michael Dobony
>> I've heard differences before. Would I be able to
>> consistently pick them out in a double-blind test?
>> Doubtful.
>
> Since when is double-blind testing the standard for making such evaluations?
Spectrometer and reference mic. Even then the results would be difficult to
measure without a sophisticated and dedicated evaluation program.
However, double-blind testing is a valid scientific measurement instrument.
Then again, music is in the ear of the listener. One may prefer 44.1 while
another prefers 96. One really needs to compare to the original, live
source.
Don Pearce
There is nothing in the double blind protocol that prevents you from
sitting down to listen. Indeed you can conduct your listening in
whatever fashion you feel will maximise your chances of hearing a
difference. All the double blind protocol does is prevent either you
or the test conductor from knowing which option you are listening to.
d
Don Pearce
>
>You might. Try a Panasonic SV-3700 and listen to the difference between
>44.1 ksamp/sec and 48 ksamp/sec operation.... they use the same anti-aliasing
>filters for both and there is substantial aliasing at 44.1. Mind you, it
>sounds dreadful at 48, but it sounds even worse at 44.1. The effect is not
>subtle.
But in this case what you are hearing is not the difference between
44.1 and 96kHz, but the difference between a good and a poor
implementation. To be useful, the test should use exemplary
implementations of both.
d
Sylvan Morein, DDS
Dc%Gn.67091$x54....@newsfe10.ams2
> I'm not totally convinced that the extra bandwidth should make so much
> difference, so what else might be going on, and how can I test for it?
Of course it's better; it's more than twice as better.
You had you're ears checked, moron?
http://robertmorein.blogspot.com/
"I don't really have a replacement career, it's a very gnawing thing."
Robert Morein
Dresher, PA
(310) 237-6511
(215) 646-4894
Neil Gould
>>>> You're clutching at straws. In the academic and professional
>>>> worlds, DBTs are the gold standard for subjective tests.
>
>>> And what sort of objective, useful information do they provide?
>
>> Well for one thing, when done properly, they can be a very good
>> tool for exposing people who claim they can hear differences in
>> sound quality when they wrap marbles around their line cords or
>> magic marker their CDs, etc.
>
> DBTs reveal what is and is not audible under the conditions of D-B
> testing.
>
conditions of the testing. DBTs eliminate some sources of error inherent to
SBTs and other less stringent methodologies.
> This may or may not have any relation with what one hears
> when one sits down to listen.
>
Unless you are referring to someone not involved in the DBT, your conclusion
is incorrect. If you *are* referring to someone not involved in the DBT,
what is your point?
--
best,
Neil
--- news://freenews.netfront.net/ - complaints: ne...@netfront.net ---
Laurence Payne
<grizzle...@comcast.net> wrote:
>DBTs reveal what is and is not audible under the conditions of D-B testing.
>This may or may not have any relation with what one hears when one sits down
>to listen.
Let's try a thought experiment. Set up a system at home, listen for
as long as you like. Weeks even. Include a box with a switch that,
when operated, may or not swap in an alternative component. Give
someone else a key to your house. When you were out, maybe he
visited. Maybe he turned the switch. Maybe it did something.
If your log of "something has changed - even my wife noticed" failed
to correlate, how would you attack the methodology of the experiment?
:-)
Arny Krueger
>>> David Clark's OPINION about what comprises such testing.
>> It's a refereed paper, which means that it had the
>> editorial board of the AES standing behind it.
>
> Oh, I am so impressed. The deaf leading the deaf.
That's a cheap shot. You don't even know who was on the AES editorial
board.
>>> This is a recommendation, not a requirement.
>
>> You're clutching at straws. In the academic and
>> professional worlds, DBTs are the gold standard for
>> subjective tests.
>
> And what sort of objective, useful information do they
> provide?
Blind tests provide exactly the same information as sighted tests, minus the
listener bias.
Arny Krueger
DBTs are just like sitting down and listening, except that listener bias has
been removed as a strong influence.
Arny Krueger
news:tbmu29maqcvm.1tsc0nglhbf82$.d...@40tude.net
> On Fri, 14 May 2010 15:49:34 -0700, William Sommerwerck
> wrote:
>
>>>> David Clark's OPINION about what comprises such
>>>> testing.
>>
>>> It's a refereed paper, which means that it had the
>>> editorial board of the AES standing behind it.
>>
>> Oh, I am so impressed. The deaf leading the deaf.
>>
>>
>>>> This is a recommendation, not a requirement.
>>
>>> You're clutching at straws. In the academic and
>>> professional worlds, DBTs are the gold standard for
>>> subjective tests.
>>
>> And what sort of objective, useful information do they
>> provide?
That same kind of information as any other listening evaluation, only the
listener's prejudices are held in abeyance.
> Well for one thing, when done properly, they can be a
> very good tool for exposing people who claim they can
> hear differences in sound quality when they wrap marbles
> around their line cords or magic marker their CD's etc.
Not only that, but when you are comparing a product with a good reputation
to a product with no reputation at all, you can easily leave the reputations
behind and just listen to the music.
William Sommerwerck
>> D-B testing. This may or may not have any relation with what
>> one hears when one sits down to listen.
> Let's try a thought experiment. Set up a system at home, listen for
> as long as you like. Weeks even. Include a box with a switch that,
> when operated, may or not swap in an alternative component. Give
> someone else a key to your house. When you were out, maybe he
> visited. Maybe he turned the switch. Maybe it did something.
> If your log of "something has changed - even my wife noticed" failed
> to correlate, how would you attack the methodology of the experiment?
I wouldn't. In fact, you made my case for me. Perfectly. Thank you.
D-BT is not "science". It is a testing protocol which might or might not be
useful.
William Sommerwerck
>> conditions of D-B testing. This may or may not have any
>> relation with what one hears when one sits down to listen.
> DBTs are just like sitting down and listening, except that
> listener bias has been removed as a strong influence.
DBT is nothing of the sort. It is a different listening experience.
Next thing, youlll be telling me there's no difference between yoghurt and
mayonnaise. That's not a joke.
Arny Krueger
news:5p4ru5l7e03l5rlg4...@4ax.com
Speaking as someone who was pretty heavily involved with high end consumer
audio at the time..
I think that saying that J. Gordon Holt created the audiophool market misses
the point. I think he did unintentionally set the stage for the creation of
that market, which was clearly created while he was an important factor in
high end audio.
Saying that J Gordon Holt is responsible for the overall high quality of
modern audio components is a grotesque insult to the thousands of very
creative, skilled, and industrious people who "wrote the plays", "set the
stages", "sold the tickets", and "performed" what Holt critiqued.
Critics and reviewers have their place, but they are not the sole creators
of whole industries and markets no matter what some (ex) reviewers and
critics may sincerily believe.
Arny Krueger
> D-BT is not "science".
No single element of science is "science".
> It is a testing protocol which might or might not be useful.
Audio DBTs have been immensely useful. They have been highly instrumental in
the creation of much of the technology that enables our modern AV industry.
(Well, its our industry if we are currently engaged in it and not sitting on
the sidelines).
There is a small problem with a number of highly vocal wannabees and
foot-draggers don't want to face up to the fact that this isn't the 1980s
any more.
I feel a lot better about selling Ray Dolby on DBTs than I feel bad about
not selling DBTs to the likes of JGH or JA.
BTW I don't even know if Ray Dolby needed selling. ;-)
Arny Krueger
>>> DBTs reveal what is and is not audible under the
>>> conditions of D-B testing. This may or may not have any
>>> relation with what one hears when one sits down to
>>> listen.
>> DBTs are just like sitting down and listening, except
>> that listener bias has been removed as a strong
>> influence.
> DBT is nothing of the sort.
And you base this on how many DBTs you've done yourself?
> It is a different listening experience.
Every listening experience is different.
> Next thing, youlll be telling me there's no difference
> between yoghurt and mayonnaise. That's not a joke.
Williiam, I'm a professional recordist with over a thousand recordings in
the hands of happy clients. I weekly do live sound and recording supporting
up to 50 musicans who are performing for an audience of over 300 who pay an
average of over $30 a head including teens and children. Who are you besides
someone that JA fired a few decades back? ;-)
Laurence Payne
<grizzle...@comcast.net> wrote:
>> Let's try a thought experiment. Set up a system at home, listen for
>> as long as you like. Weeks even. Include a box with a switch that,
>> when operated, may or not swap in an alternative component. Give
>> someone else a key to your house. When you were out, maybe he
>> visited. Maybe he turned the switch. Maybe it did something.
>
>> If your log of "something has changed - even my wife noticed" failed
>> to correlate, how would you attack the methodology of the experiment?
>
>I wouldn't. In fact, you made my case for me. Perfectly. Thank you.
>
>D-BT is not "science". It is a testing protocol which might or might not be
>useful.
Well, what I described is a DB test, though a more laborious one than
is generally found practical. It would be worth setting up, if only
to shut up the True Believers. Except that when results failed to
correlate with their beliefs, they'd of course find a hole to wriggle
through :-)
William Sommerwerck
result of being forced to rethink my beliefs. I've learned almost nothing by
being supplied with accurate, valid explanations, because such are few and
far between.
The most-useful was the extended discussion of AD and DA conversion. There
were several math and philosophical holes in my understanding, which were
filled in, mostly by the doctoral thesis someone supplied. (The repeated
"explanations" that the output of a DAC is analog because the device is
/called/ a digital-to-analog converter had zero influence.)
I sometimes feel like an extra-terrestrial, wondering why humans have
intelligence they don't use. You believe what you read in books or what some
"expert" tells you, without questioning it. You think that because you know
something you understand it.
I've had my semi-annual shot at Arny's misunderstanding of "science". It's
unlikely he'll ever have his "Aha!" moment.
Laurence Payne
wrote:
>Williiam, I'm a professional recordist with over a thousand recordings in
>the hands of happy clients. I weekly do live sound and recording supporting
>up to 50 musicans who are performing for an audience of over 300 who pay an
>average of over $30 a head including teens and children. Who are you besides
>someone that JA fired a few decades back? ;-)
...er....what the congregation put in the collection plate doesn't
really say anything about the audio or musical quality they're
getting! We all know you're working in an environment where
intention is valued way above results, and that the only sample of
your work we've ever heard is pretty bad both musically and
technically. Waving your "professional credentials" around is a risky
game.
Soundhaspriority
"philicorda" <phili...@dontspamme.com> wrote in message
news:Dc%Gn.67091$x54....@newsfe10.ams2...
> I've always been of the opinion that there should not be an obvious
> audible difference between recording at 44.1 and 96Khz. When I have
> experimented with average converters, both sound fairly similar.
>
> But... the other day a friend brought in a Sony PCM-D50 and claimed there
> was a huge difference at 96K. We tried a few bind tests using the
> internal mics on voice and piano, and sure enough the 96Khz sounded
> obviously 'better'. Clearer, more present etc.
>
> The listening setup was nothing special, a Teac A-X3000 amplifier and
> Tannoy Mercury MX2 speakers in an average living room.
>
> I'm not totally convinced that the extra bandwidth should make so much
> difference, so what else might be going on, and how can I test for it?
Phil, good to know. I have one of these, but have not examined it so
critically. I'll be sure to record at 96 with this unit.
Bob Morein
(310) 237-6511
William Sommerwerck
> in the hands of happy clients. I weekly do live sound and recording
> supporting up to 50 musicans who are performing for an audience
> of over 300 who pay an average of over $30 a head including teens
> and children. Who are you besides someone that JA fired a few
> decades back? ;-)
Not funny. Who are you but someone who's repeatedly lied about his
accomplishments and published papers? Ask JA, who's had to rebut your claims
on several occasions.
You have never given a straight answer to any of my difficult questions.
Because you don't know them. (In fairness, no one does.)
Have you ever stood up in front of a manufacturer and called him a liar? I
have. (The head of Micro Acoustics, if it's of any interest.) My only regret
as a reviewer is, after having been insulted to my face by Peter Aczel ("I
didn't know they let people like you in here."), I didn't go over and start
smashing his speakers. I refrained partly because I didn't want to embarrass
Gordon, but more because I didn't want a felony arrest on my police record.
I quit Stereophile after a startling listening experience in which I
discovered just how appallingly unreliable short-term subjective testing
was. I no longer felt I was a reliable reviewer. * (I'm still not sure.)
John will tell you that he fired me because I accepted bribes from
manufacturers. Not true. (JA's memory is the most-flexible of any human I've
ever known.)
In case you're not aware of it, bribery is common. The simplest form is the
availability of accommodation purchases. These allow reviewers to purchase
equipment for (usually) 50% of retail. The reviewer can sell it after a year
or so, often making a profit. Although accommodation purchases are also to
the advantage of the manufacturer, they do slightly distort the reviewing
process. **
Another form of bribery is the long-term loan, which JA strongly encouraged.
The argument is that if you feel a product is of reference quality, you
should have a sample for comparison. Nothing wrong with that, but these
loans often turn into ownership, especially as products are discontinued. I
possess a JVC hall synthesizer, plus STAX headphones, amplifier, and
equalizer that were obtained under these conditions. Shure was also nice to
me over the years, probably because I gave their surround processors the
rave reviews they deserved. ***
Yet another form of bribery is the factory junket, in which the manufacturer
pays most or all of the expenses. These exist primarily to impress the
reviewer, and should probably be verboten. What should definitely be
prohibited is allowing reviewers to bring along wives or family. True, the
reviewer usually pays for their fare and accommodations. However, few
reviewers are wealthy, and these junkets make possible a relatively
inexpensive vacation or getaway that might otherwise not be affordable.
* The fact that the amount of money I was paid for the amount of time I
spent on a review was insufficient. I was also not pleased when I trashed
the AKG K1000 headphones, and John told me that my FFT measurements proving
that their design was defective were probably wrong. "How close were you to
the wall?" JA is not much-interested in the truth. Like you, he would make a
lousy scientist.
** For obvious psychological reasons, you should not review what you own, or
vice versa.
*** I also have Yamaha orthodynamic headphones. Yamaha wouldn't take them
back, apparently for health reasons. It's a shame they aren't made any more.
Moshe
You keep moving the goal posts.
I never said it was a science.
I simply claimed it is a method that can be useful for determining
what some people claim to hear and in a setting that is as
unbiased as possible.
IOW take one of these "Golden Eared" audiophile's, put him in a
room with the sources and see if he can in a statistically
significant manner, choose one source over the other.
You may have to set out some traps to catch one of them though and
physically drag him or her to the testing site because those
types, I call them "Stereophiles" after the magazine name, are
difficult to catch and pin down when it comes down to fish or
cutting bait.
For the record, I'm more along the lines of if differences can
heard,under a specific set of conditions, they can most likely be
measured.
Ethan Winer has a couple of threads going over on Gearslutz that
are similar to this one. Very interesting reading, not that I
agree with everything Ethan proposes.
William Sommerwerck
> I simply claimed it is a method that can be useful for determining
> what some people claim to hear and in a setting that is as
> unbiased as possible.
But D-BT is not unbiased, because it isn't the way you normally listen.
> IOW take one of these "Golden Eared" audiophile's, put him
> in a room with the sources and see if he can in a statistically
> significant manner, choose one source over the other.
I'd be delighted to do it. Just let me pick the equipment to be tested.
I'll be specific. You provide a Krell KSA-series amplifier, and a Crown
K-series amplifier. I will provide a Parasound A-series amplifier. They
sound so different from each other, I can quickly and easily distinguish
them. In fact, /you/ will be able to tell which is which, /without/
comparing them, simply from my description of their sound characteristics.
I'll list those characteristics here... The Krell produces a very wide and
deep soundfield, with a lot of "space". It's liquid-sounding, and has the
"slamming" bass many people describe as characteristic of Krell amplifiers.
JGH might have described as having an unnaturally "backed-off" midrange.
The Parasound is noticeably "flatter", with rather less depth and space. It
is "drier", and not so liquid. The bass is less euphonically emphatic. There
also seems to be a bit more detail, though this might be the side-effect
The Crown amplifier sounds like dog feces. It's coarse and hashy sounding.
You don't need to compare it with anything to hear how bad it is.
Note that these amplifiers were designed by three of the most-famous (and
respected) designers in the history of high-end audio -- Dan D'Agostino,
John Curl, and Gerry Stanley, respectively. And they sound different.
> For the record, I'm more along the lines of if differences can
> heard, under a specific set of conditions, they can most likely
> be measured.
Of course. If something is audible, it /has/ to be measurable.
Laurence Payne
<grizzle...@comcast.net> wrote:
>> You keep moving the goal posts. I never said it was a science.
>> I simply claimed it is a method that can be useful for determining
>> what some people claim to hear and in a setting that is as
>> unbiased as possible.
>
>But D-BT is not unbiased, because it isn't the way you normally listen.
I've just offered you a D-BT method that suits your normal listening
habits perfectly. You're arguing as if the implementation of D-BT is
set in stone. It isn't.
William Sommerwerck
>>> I simply claimed it is a method that can be useful for determining
>>> what some people claim to hear and in a setting that is as
>>> unbiased as possible.
>> But D-BT is not unbiased, because it isn't the way you normally listen.
> I've just offered you a D-BT method that suits your normal listening
> habits perfectly. You're arguing as if the implementation of D-BT is
> set in stone. It isn't.
But everybody /else/ to seem to think it is. I have /long/ argued for
long-term (weeks or months) D-BT, under "living-room" conditions. No one
wants to do such tests, because they're too complex and expensive.
Laurence Payne
Well, say so then! You've wasted an entire thread arguing against a
narrow definition which you don't agree with, and you have no way of
knowing whether we do either!
"I don't like shoes"
....long argument....
"Well, of course I only meant GREEN shoes!"
Moshe
>> You keep moving the goal posts. I never said it was a science.
>> I simply claimed it is a method that can be useful for determining
>> what some people claim to hear and in a setting that is as
>> unbiased as possible.
>
> But D-BT is not unbiased, because it isn't the way you normally listen.
Laurence suggested a variation which does simulate the way you
listen.
The longer you perform the experiment, the more accurate the
results should be.
>
>> IOW take one of these "Golden Eared" audiophile's, put him
>> in a room with the sources and see if he can in a statistically
>> significant manner, choose one source over the other.
>
> I'd be delighted to do it. Just let me pick the equipment to be tested.
>
> I'll be specific. You provide a Krell KSA-series amplifier, and a Crown
> K-series amplifier. I will provide a Parasound A-series amplifier. They
> sound so different from each other, I can quickly and easily distinguish
> them. In fact, /you/ will be able to tell which is which, /without/
> comparing them, simply from my description of their sound characteristics.
Maybe you can.
Maybe you can't.
I dunno.
If they measure differently, you probably can hear a difference.
I don't particularly like the sound of Crown amplifiers, but you
can drop one from the back of a truck and they will probably still
function.
Irrelevant to the discussion however.
> I'll list those characteristics here... The Krell produces a very wide and
> deep soundfield, with a lot of "space". It's liquid-sounding, and has the
> "slamming" bass many people describe as characteristic of Krell amplifiers.
> JGH might have described as having an unnaturally "backed-off" midrange.
Why do I feel like I am reading Stereophile here?
> The Parasound is noticeably "flatter", with rather less depth and space. It
> is "drier", and not so liquid. The bass is less euphonically emphatic. There
> also seems to be a bit more detail, though this might be the side-effect
>
> The Crown amplifier sounds like dog feces. It's coarse and hashy sounding.
> You don't need to compare it with anything to hear how bad it is.
>
> Note that these amplifiers were designed by three of the most-famous (and
> respected) designers in the history of high-end audio -- Dan D'Agostino,
> John Curl, and Gerry Stanley, respectively. And they sound different.
Do they measure differently in a significant manner?
If so, you have your answer.
BTW how does one measure, "flatter, drier and liquid" ?
I'm a little behind the times as they still taught tubes when I
was in college earning my BSEE, but have things changed that much?
>
>> For the record, I'm more along the lines of if differences can
>> heard, under a specific set of conditions, they can most likely
>> be measured.
>
> Of course. If something is audible, it /has/ to be measurable.
Ahh, but you are assuming that the person, looking at the unit and
it's fancy brand is actually hearing a difference.
That's where the DBT is useful.
I used to sell high end equipment in various NYC audio salons back
in the 1970's and in fact it put me through engineering school.
Some of that gear easily cost more than an automobile at the time
and we had no problem unloading, errr, selling it.
In fact we didn't even have to sell it because the high end
magazines at the time did it for us. These stores were real smart
in that they had a very liberal exchange policy and like clock
work as soon as a new piece of gear was tested by the Stereophile
type magazines, the suckers, errr, customers, would be lined up at
the door wanting to exchange last month's "star of the show".
They swore they could hear differences and there was no trickery
going on at the store.
After hours the staff used to sit in the A room and compare gear
as we were all gear nuts as well.
Funny thing was, the more wine and cheese we scarfed down, the
more we "heard" differences between the gear.
Those were fun days.
Laurence Payne
<goldee_l...@gmail.com> wrote:
>> I'll be specific. You provide a Krell KSA-series amplifier, and a Crown
>> K-series amplifier. I will provide a Parasound A-series amplifier. They
>> sound so different from each other, I can quickly and easily distinguish
>> them. In fact, /you/ will be able to tell which is which, /without/
>> comparing them, simply from my description of their sound characteristics.
>Maybe you can.
>Maybe you can't.
>I dunno.
>
>If they measure differently, you probably can hear a difference.
>I don't particularly like the sound of Crown amplifiers, but you
>can drop one from the back of a truck and they will probably still
>function.
Ok, so you confirm there is a "sound" of the Crown. What about the
other two?
Moshe
Some Crown amplifiers yes.
Not all of them. I should have been more specific.
The PSA series in particular.
I suspect it will be able to be measured as well which is my
point.
Haven't heard on in years though although the last gig I did a
couple of weeks ago was using one for the sub cabinet, I think.
Les Cargill
> I've always been of the opinion that there should not be an obvious
> audible difference between recording at 44.1 and 96Khz. When I have
> experimented with average converters, both sound fairly similar.
>
> But... the other day a friend brought in a Sony PCM-D50 and claimed there
> was a huge difference at 96K. We tried a few bind tests using the
> internal mics on voice and piano, and sure enough the 96Khz sounded
> obviously 'better'. Clearer, more present etc.
>
> The listening setup was nothing special, a Teac A-X3000 amplifier and
> Tannoy Mercury MX2 speakers in an average living room.
>
> I'm not totally convinced that the extra bandwidth should make so much
> difference, so what else might be going on, and how can I test for it?
You probably want to do the usual measures of frequency response - white
noise, swept tones, individual octave tones ( 1k, 2k, 4k,
8k, 16k ). Hopefully, you can go around the microphones.
Being able to separate the recorded waveform from the signal
as played back would be a good thing. I see it has something like
USB connectivity, which should help with that.
--
Les Cargill
Scott Dorsey
>On 14 May 2010 09:39:39 -0400, klu...@panix.com (Scott Dorsey) wrote:
>
>>You might. Try a Panasonic SV-3700 and listen to the difference between
>>44.1 ksamp/sec and 48 ksamp/sec operation.... they use the same anti-aliasing
>>filters for both and there is substantial aliasing at 44.1. Mind you, it
>>sounds dreadful at 48, but it sounds even worse at 44.1. The effect is not
>>subtle.
>
>But in this case what you are hearing is not the difference between
>44.1 and 96kHz, but the difference between a good and a poor
>implementation. To be useful, the test should use exemplary
>implementations of both.
We have had good tests with exemplary implementations, which did not show
audible differences.
However, the original poster can hear audible differences. Therefore,
we attribute his problem to a poor implementation.
This is why he asks the original questions that started the thread: just
what precisely is poor about his implementation? How can he measure the
problem effectively? And what should he do about it?
I can't answer any of these yet. Can you? They are valid and important
questions.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Scott Dorsey
>> Ok, so you confirm there is a "sound" of the Crown. What about the
>> other two?
>
>Some Crown amplifiers yes.
>Not all of them. I should have been more specific.
>
>The PSA series in particular.
>
>I suspect it will be able to be measured as well which is my
>point.
Dunno about the PSA, but the PS1 has a crossover distortion issue at low
levels and slew rate issues at high levels. Hopefully things have improved
since then (although the D75 makes me suspect they have not).
On the other hand the MacroTechs sound pretty clean.
Peter Larsen
> This is why he asks the original questions that started the thread:
> just what precisely is poor about his implementation? How can he
> measure the problem effectively? And what should he do about it?
Testing border-conditions can be a fruitful strategy, so how about testing
its frequency response near upper cut-off for peakyness and testing how it
behaves when provoked with tones that may cause aliasing. That ought to be
reasonably simple and repeatable.
Kind regards
Peter Larsen
Don Pearce
I'm not sure I can summon as much interest in the question "why does a
poor implementation sound bad?" as I could about the 44.1/96kHz
question, if it could indeed be shown to reveal differences.
d
Arny Krueger
> I'll be specific. You provide a Krell KSA-series
> amplifier, and a Crown K-series amplifier. I will provide
> a Parasound A-series amplifier. They sound so different
> from each other, I can quickly and easily distinguish
> them. In fact, /you/ will be able to tell which is which,
> /without/ comparing them, simply from my description of
> their sound characteristics.
OK, so these are the only 3 power amps in the world that actually sound
different.
I can live with that!
>why would I care?<
I dunno, maybe one or more of them has some kind of serious technical flaw.
I've never said that all amplifiers sound different. Of course they don't
all sound different. Some are broke, some are bad designs.
Arny Krueger
message news:hsmc80$sg5$1...@news.eternal-september.org
> But D-BT is not unbiased, because it isn't the way you
> normally listen.
Non sequitor.
The fallacy is that any method of listening that does not exactly duplicate
some personal methodology of yours is for sure biased.
The fact is that there are any number of variations on the basic process of
listening that are in fact far more unbiased than your sighted evaluations.
Arny Krueger
>> Williiam, I'm a professional recordist with over a
>> thousand recordings in the hands of happy clients. I
>> weekly do live sound and recording supporting up to 50
>> musicans who are performing for an audience
>> of over 300 who pay an average of over $30 a head
>> including teens
>> and children. Who are you besides someone that JA fired
>> a few
>> decades back? ;-)
> Not funny.
But true.
> Who are you but someone who's repeatedly lied
> about his accomplishments and published papers?
What lies?
> Ask JA, who's had to rebut your claims on several occasions.
I have no idea about that. Its a free country, he can say what he wants to
say. If I don't know about it, I pay it no mind.
> You have never given a straight answer to any of my
> difficult questions.
I am unaware of that.
> Because you don't know them. (In fairness, no one does.)
More like I don't know what the questions are.
> Have you ever stood up in front of a manufacturer and
> called him a liar?
Name calling is not one of those things that I want to hang my hat on.
Arny Krueger
> You believe what
> you read in books or what some "expert" tells you,
> without questioning it. You think that because you know
> something you understand it.
Prove it.
William Sommerwerck
>> normally listen.
> Non sequitor.
> The fallacy is that any method of listening that does not exactly
> duplicate some personal methodology of yours is for sure biased.
No, I am going by "science" -- the test procedures have to match the
real-world conditions. The type of D-BT you advocate does not.
> The fact is that there are any number of variations on the basic process
of
> listening that are in fact far more unbiased than your sighted
evaluations.
Even other people here, Arny, understand that you can have D-BT that matches
the way one usually listens. You've never acknowledged that, because you're
afraid that it would produce different results from those you obtained. It
might, it might not. The only way to find out is to do it.
William Sommerwerck
>> tells you, without questioning it. You think that because
>> you know something you understand it.
> Prove it.
The fact that I raise points or ask questions, and (almost) always fail to
get a clear, coherent answer proves it.
There is a difference between knowing and understanding. I might be many
things, but I am not guilty of confusing the two.
Moshe
A band I used to play with used the PSA with both EAW (before they
were stolen) and later JBL cabinets and I never liked the sound of
those amps. Maybe it was the speaker/amplifier combination, maybe
it was me I really don't know but it just did not sound right with
my keyboards going through them.
Brittle is the best description I can come up with.
I'm convinced, measurements would show faults in these amps.
Board was a "vintage, at least these days" Yamaha analog board
which had a wonderful. smooth sound to it. Effects were all dbx
and Lexicon.
When my Bryston 4b (home system) developed a problem and needed
servicing, the shop loaned me a Crown Microtech (not Macro)
because I was having a party that weekend and that's all they had
around.
It sounded fine hooked to my KeF 105 speakers.
Maybe a little to much partying, but I really never noticed a
difference.
Bryston repaired the unit free of charge BTW.
All this is totally and completely subjective of course.
Scott Dorsey
>
>I'm not sure I can summon as much interest in the question "why does a
>poor implementation sound bad?" as I could about the 44.1/96kHz
>question, if it could indeed be shown to reveal differences.
So, rather than answer the original poster's question, you'd rather answer
a different and possibly unrelated one?
In fact, the question about why a poor implementation sounds bad is actually
a very interesting one, because it's the converse of asking what makes a
good implementation sound good.
Moshe
I agree.
Basically, and simplistically maybe, what I am saying is that if
users can hear a difference, measurements will reveal a
difference.
Combinations of amps and speakers are the first place to look.
Esoteric cabling is another place.
etc.
Some people just have better ears than others.
My significant other has ears like a cat.
Can't tell a good tune from a bad one though....
Moshe
Personally I love it when a knowledgeable person, like yourself
Scott, comes along and posts something like "xxyyzz converters are
bad because aabbcc is a poor design" because at least this way, if
I have indeed been hearing something that maybe I don't like, I
have some kind of starting point for data to back up my claims, or
my ranting's, frustrations etc.
There are a zillion reasons why one piece of gear can sound better
than another, but my basic POV is that if a piece of gear can be
statistically picked as better, or different from another piece of
gear, even if by a single person, then either the testing set up
is flawed, the ancillary gear (cables, speakers etc) favors one
piece of gear, or the differences can be measured and in the case
of cables, speakers etc this too can be measured although not
quite as easily.
It's tough being schooled as an engineer yet being a musician at
the same time.
William Sommerwerck
> is that if users can hear a difference, measurements
> will reveal a difference.
To paraphrase Mozart... Which mesurements would you have me make?
25 years ago, I asked JA to take me on as (in effect) Stereophile's
technical editor. As was and is his wont, he instantly refused. Had he not
done so, we might have made some progress. But objective testing with
more-or-less sure answers does not sell audiophile magazines.
Moshe
All your Stereophile baggage aside, and I don't even know what it
is as I must have missed that chapter, but times have changed.
The standard measurements that technology allows.
Frequency response, slew rate, power, distortion, phase etc.
There is always the chance that physics has not caught up with the
human being's abilities, I am not doubting this, however that
would easily be picked up by complete double blind tests.
I read Stereophile and I LMAO.
For me it's like reading Mad Magazine.
Substitute wine for electronic gear, taste for sound and you would
swear you were reading a snobby wine magazine.
Bascomb King is a riot to read.
I'd love to get that cracked pot into a studio with some gear and
a true double blind test.
I'd bet a case of his favorite wine that he couldn't tell the
$10,000 converter from the $100.00 variety in a manner that was
statistically significant.
BTW he doesn't get to look at the labels, nor the price tag before
hand.
Ever notice how these golden eared people suddenly either
disappear or start moving the goal posts when it comes down to
pinning them to the floor with actual, verifiable by independent
sources, tests?
See the Ethan Winer threads in gearslutz for details.
Let's just say I have my doubts.
Scott Dorsey
> Kludge writes:
>> My personal suspicion is that all of the audible differences between high
>> sample rates have to do with conversion artifacts and extended bandwidth
>> producing beat notes in the audible band. But by the same token,
>> if going to a different rate alters the artifacts in a good way, why not
>> just go with it?
>
>Or even if it just alters the client's expectation of the sound. If the
>client thinks it sounds better, that's good enough for me.
And that's really my argument for not having gone the wideband route yet
myself... the customers haven't demanded it. Interestingly, I have had
requests for DSD.
William Sommerwerck
>> 25 years ago, I asked JA to take me on as (in effect) Stereophile's
>> technical editor. As was and is his wont, he instantly refused. Had he
not
>> done so, we might have made some progress. But objective testing with
>> more-or-less sure answers does not sell audiophile magazines.
> All your Stereophile baggage aside, and I don't even know what it
> is as I must have missed that chapter, but times have changed.
No, they haven't. Nothing has changed in the past 60 years. People are still
arguing over matters of which they have neither practical or philosophical
understanding.
> The standard measurements that technology allows.
And how do they correlate with what we hear? What, for example, is the
threshold of harmonic or intermodulation distortion? How does the audibility
of distortion relate to its spectrum or orde?. Can you tell me? If you
can't, ask Arny. He knows /all/ these answers, right?
> Frequency response, slew rate, power, distortion, phase, etc.
> There is always the chance that physics has not caught up with
> the human being's abilities, I am not doubting this, however that
> would easily be picked up by complete double blind tests.
Prove it. I say that D-BT, as promoted by its leading adherents (who sell
"scientific" snake oil, but snake oil, nonetheless), is at least incomplete,
and at best grossly misleading. It is true only up to a point. And that
point is rather distantly removed from the as-yet-unknown truth.
> I read Stereophile and I LMAO.
> For me it's like reading Mad Magazine.
You would have felt differently 45 years ago. "The Stereophile" was founded
by a man who broke away from the mass-market magazines because they were
paid by manufacturers to lie about the quality of their products. They not
only lied about bad products, they lied about good products.
There is a story (which I cannot verify) that states that "High Fidelity"
got very, very nervous when J Gordon Holt wanted to publish a review stating
that the KLH Nine was, overall, the best speaker then available.
> Ever notice how these golden-eared people suddenly either
> disappear or start moving the goal posts when it comes down
> to pinning them to the floor with actual, verifiable by independent
> sources, tests?
Which "independent" sources? What verifiable tests? The ones /you/ think are
independent and verifiable? How much audio equipment have you ever sat down
and carefully auditioned? Not much, I suspect, because you're deathly afraid
that you might hear differences. It's so much more-pleasant to use ABX
testing, where all but the grossest differences magically vanish. You're
finally safe, and "science" has rescued you from beliefs that you don't want
to believe. How nice.
I don't have the money to run "proper" tests. And if I did, you wouldn't
have the guts to sit in on them. Nor would Arny nor JA.
Moshe
>>> To paraphrase Mozart... Which measurements would you have me make?
>>> 25 years ago, I asked JA to take me on as (in effect) Stereophile's
>>> technical editor. As was and is his wont, he instantly refused. Had he
> not
>>> done so, we might have made some progress. But objective testing with
>>> more-or-less sure answers does not sell audiophile magazines.
>
>> All your Stereophile baggage aside, and I don't even know what it
>> is as I must have missed that chapter, but times have changed.
>
> No, they haven't. Nothing has changed in the past 60 years. People are still
> arguing over matters of which they have neither practical or philosophical
> understanding.
And yet when pressed for testing, etc the people who make these
claims are often elusive.
>> The standard measurements that technology allows.
>
> And how do they correlate with what we hear? What, for example, is the
> threshold of harmonic or intermodulation distortion? How does the audibility
> of distortion relate to its spectrum or orde?. Can you tell me? If you
> can't, ask Arny. He knows /all/ these answers, right?
Nobody knows.
However, properly conducted double blind tests go a long way
toward proving or disproving these claims.
>
>> Frequency response, slew rate, power, distortion, phase, etc.
>> There is always the chance that physics has not caught up with
>> the human being's abilities, I am not doubting this, however that
>> would easily be picked up by complete double blind tests.
>
> Prove it. I say that D-BT, as promoted by its leading adherents (who sell
> "scientific" snake oil, but snake oil, nonetheless), is at least incomplete,
> and at best grossly misleading. It is true only up to a point. And that
> point is rather distantly removed from the as-yet-unknown truth.
That's funny....
The snake oil is really the people in rags like Stereophile
subscribing to stuff like marbles wrapped around a line cord and
such.
The clue to their idiocy is usually statements like "the
difference was astounding".
Come on already.
>
>> I read Stereophile and I LMAO.
>> For me it's like reading Mad Magazine.
>
> You would have felt differently 45 years ago. "The Stereophile" was founded
> by a man who broke away from the mass-market magazines because they were
> paid by manufacturers to lie about the quality of their products. They not
> only lied about bad products, they lied about good products.
Well I can't go back 45 years, but I can go back to the 70's and
in particular 1973 on....
> There is a story (which I cannot verify) that states that "High Fidelity"
> got very, very nervous when J Gordon Holt wanted to publish a review stating
> that the KLH Nine was, overall, the best speaker then available.
They said the same about the Advent speaker back then.
Next year it was the Allison One.
Then the Dahlquist or whatever that flat speaker was called, I
can't remember.
The year after that it was the ESS and their Heil Air Motion
thing.
Then we moved on to Genesis.....etc...
>
>> Ever notice how these golden-eared people suddenly either
>> disappear or start moving the goal posts when it comes down
>> to pinning them to the floor with actual, verifiable by independent
>> sources, tests?
>
> Which "independent" sources? What verifiable tests? The ones /you/ think are
> independent and verifiable? How much audio equipment have you ever sat down
> and carefully auditioned? Not much, I suspect, because you're deathly afraid
> that you might hear differences. It's so much more-pleasant to use ABX
> testing, where all but the grossest differences magically vanish. You're
> finally safe, and "science" has rescued you from beliefs that you don't want
> to believe. How nice.
Are you kidding?
I used to sell the stuff.
The shop I worked at had one of the first ABX comparator kits in
the USA.
The so called "golden ears" were exposed literally 100 percent of
the time.
I spent many hours listening to super high end gear that I
couldn't afford, at the time.
The best I could come up with, was it looks great and the
construction was second to none.
> I don't have the money to run "proper" tests. And if I did, you wouldn't
> have the guts to sit in on them. Nor would Arny nor JA.
Head on over to Gearslutz and talk to Ethan.
He is doing such a test and is looking for people.
Funny thing is the golden eared people seem to be going into
hiding.
This is typical....
If you know Bascom King personally, ask him to contact Ethan and
let's see if he can pick out a high end Lynx converter from a run
of the mill Delta 1010.
My money is on him failing miserably.
And BTW I don't agree with everything Ethan says...
I am as much a skeptic as you are and that's kool...
William Sommerwerck
> >
> > And how do they correlate with what we hear? What, for example, is the
> > threshold of harmonic or intermodulation distortion? How does the
audibility
> > of distortion relate to its spectrum or orde?. Can you tell me? If you
> > can't, ask Arny. He knows /all/ these answers, right?
>
> Nobody knows.
> However, properly conducted double blind tests go a long way
> toward proving or disproving these claims.
This is an unproven assumption. The validity & predictability of D-BT
remains in question.
>> There is a story (which I cannot verify) that states that "High Fidelity"
>> got very, very nervous when J Gordon Holt wanted to publish a review
stating
>> that the KLH Nine was, overall, the best speaker then available.
> They said the same about the Advent speaker back then.
No, neither Stereophile nor Absolute Sound said that. Harry Pearson was
rather taken with Double Advents, but to say that he considered it the best
speaker available would be stretching it a bit.
> Next year it was the Allison One.
> Then the Dahlquist or whatever that flat speaker was called, I
> can't remember.
> The year after that it was the ESS and their Heil Air Motion
> thing. Then we moved on to Genesis.....etc...
Baloney. No audiophile magazine ever said these were "the best". Besides,
you're deliberately ignoring the point.
In the early 60s, there were two speakers that, in terms of accurately
reproducing the signal, could have been considered "the best available".
These were the QUAD ESL-57 and the KLH Nine, both electrostatic. At that
time, these really were terrifically good speakers. There are people today
who consider the ESL-57 the best speaker ever made, and no serious modern
listener would say that it /isn't/ a good speaker. It has kept its high
reputation for 53 years (!!!!!!!!!!!!!!), something you cannot say about any
other speaker. * (Don't mention Klipschorns, which have changed quite a bit
in that time.)
Unlike Stereophile, Abso!ute Sound, et al, High Fidelity was frightened at
what might happen if they said "This is almost certainly the best speaker
you can buy." They would likely have lost a lot of advertising.
* I've never heard the A7. I'm sure it's fun to listen to, but how accurate
is it?
>> Which "independent" sources? What verifiable tests? The ones /you/ think
are
>> independent and verifiable? How much audio equipment have you ever sat
down
>> and carefully auditioned? Not much, I suspect, because you're deathly
afraid
>> that you might hear differences. It's so much more-pleasant to use ABX
>> testing, where all but the grossest differences magically vanish. You're
>> finally safe, and "science" has rescued you from beliefs that you don't
want
>> to believe. How nice.
> Are you kidding?
> I used to sell the stuff.
> The shop I worked at had one of the first ABX comparator kits in
> the USA.
> The so called "golden ears" were exposed literally 100 percent of
> the time.
> I spent many hours listening to super high end gear that I
> couldn't afford, at the time.
> The best I could come up with, was it looks great and the
> construction was second to none.
Thank you for confirming that ABX testing masks audible differences!
If you think that, say, Audio Research equipment of that era was sonically
indistinguishable from, say, Crown -- then there is/was something very wrong
with your hearing.
Stuart Richards
DSD? In that case I'd love to know if you've heard the Grimm Audio AD1:
http://www.grimmaudio.com/pro_converters_ad1.htm
Moshe
>>>> The standard measurements that technology allows.
>>>
>>> And how do they correlate with what we hear? What, for example, is the
>>> threshold of harmonic or intermodulation distortion? How does the
> audibility
>>> of distortion relate to its spectrum or orde?. Can you tell me? If you
>>> can't, ask Arny. He knows /all/ these answers, right?
>>
>> Nobody knows.
>> However, properly conducted double blind tests go a long way
>> toward proving or disproving these claims.
>
> This is an unproven assumption. The validity & predictability of D-BT
> remains in question.
That's the classic argument offered up by the so called golden
ears who always seem to disappear when these tests are offered
up..
Again, see Ethan Winer's offer in Gearslutz and notice the gear
snobs running and hiding behind their overpriced converters.
They are trying to find every reason why they won't submit to the
tests.
Reason?
They are terrified of picking a $100.00 Sound Blaster over their
$4000.00 Lynx.
>
>>> There is a story (which I cannot verify) that states that "High Fidelity"
>>> got very, very nervous when J Gordon Holt wanted to publish a review
> stating
>>> that the KLH Nine was, overall, the best speaker then available.
>
>> They said the same about the Advent speaker back then.
>
> No, neither Stereophile nor Absolute Sound said that. Harry Pearson was
> rather taken with Double Advents, but to say that he considered it the best
> speaker available would be stretching it a bit.
Could be, I'm just talking of the press in general and their
fickle nature which seems to somehow be attached to a money
trail....
At least that's what I suspect.
>> Next year it was the Allison One.
>> Then the Dahlquist or whatever that flat speaker was called, I
>> can't remember.
>> The year after that it was the ESS and their Heil Air Motion
>> thing. Then we moved on to Genesis.....etc...
>
> Baloney. No audiophile magazine ever said these were "the best". Besides,
> you're deliberately ignoring the point.
Best?
Nothing is "best" in this business.
At least not for more than a month or so.
Stereo Review, the bane of the Stereophile crowd used to expose
this kind of bunk all the time.
> In the early 60s, there were two speakers that, in terms of accurately
> reproducing the signal, could have been considered "the best available".
> These were the QUAD ESL-57 and the KLH Nine, both electrostatic. At that
> time, these really were terrifically good speakers.
I have heard the Quad's and I would agree they sounded very, very
nice IMHO.
The only KLH I have heard were bookshlef variety and sounded
aqfully muddy.
Still, many people bought the advent, the Allison etc.
Best?
No..Maybe I am gilding the lilly a bit, but let's just say "great
reviews".
There are people today
> who consider the ESL-57 the best speaker ever made, and no serious modern
> listener would say that it /isn't/ a good speaker. It has kept its high
> reputation for 53 years (!!!!!!!!!!!!!!), something you cannot say about any
> other speaker. * (Don't mention Klipschorns, which have changed quite a bit
> in that time.)
I hated Klipschorns and as a musician who at the time was working
on Broadway, I got to go to enough cast parties where the
producers had them.
Honk city as far as I was concerned.
Best speakers at the time IMHO were KeF and maybe B&W.
Maggies were also very nice as long as you weren't into disco :)
Mcintosh also made a real killer, sleeper speaker.
> Unlike Stereophile, Abso!ute Sound, et al, High Fidelity was frightened at
> what might happen if they said "This is almost certainly the best speaker
> you can buy." They would likely have lost a lot of advertising.
Yet those same rags have reviews of products and then
advertisements on the very next page for those products.
They are not alone, most rags are like that.
High Fidelity went way downhill IMHO.
So did Stereo Review.
Still, I always liked Julian Hirsch's approach to things.
Still even he was off base sometimes, like when he would slam a
phono pre-amp (MM) for having a 300mv or so overload even though
that figure would rarely if ever be reached in real life.
> * I've never heard the A7. I'm sure it's fun to listen to, but how accurate
> is it?
>
>
>>> Which "independent" sources? What verifiable tests? The ones /you/ think
> are
>>> independent and verifiable? How much audio equipment have you ever sat
> down
>>> and carefully auditioned? Not much, I suspect, because you're deathly
> afraid
>>> that you might hear differences. It's so much more-pleasant to use ABX
>>> testing, where all but the grossest differences magically vanish. You're
>>> finally safe, and "science" has rescued you from beliefs that you don't
> want
>>> to believe. How nice.
>
>> Are you kidding?
>> I used to sell the stuff.
>> The shop I worked at had one of the first ABX comparator kits in
>> the USA.
>
>> The so called "golden ears" were exposed literally 100 percent of
>> the time.
>> I spent many hours listening to super high end gear that I
>> couldn't afford, at the time.
>> The best I could come up with, was it looks great and the
>> construction was second to none.
>
> Thank you for confirming that ABX testing masks audible differences!
That's a cop out.
Did it ever occur to you that maybe the gear sounds the same?
> If you think that, say, Audio Research equipment of that era was sonically
> indistinguishable from, say, Crown -- then there is/was something very wrong
> with your hearing.
Do scientific measurements show a difference?
If they do, then you are right.
If they don't I'm sure I can find a few people with fat wallets
who will hear the difference.
Just make sure and show them the nameplate first.
Scott Dorsey
>On Sat, 15 May 2010 16:29:20 -0700, William Sommerwerck wrote:
>
>>> The standard measurements that technology allows.
The problem is that the standard measurements aren't complete. And a lot
of the measurements that are very powerful at characterizing perceived
sound quality, like distortion spectra, haven't really been standardized
although today in the Audio Precision world they are not that hard to
make accurately.
>> And how do they correlate with what we hear? What, for example, is the
>> threshold of harmonic or intermodulation distortion? How does the audibility
>> of distortion relate to its spectrum or orde?. Can you tell me? If you
>> can't, ask Arny. He knows /all/ these answers, right?
>
>Nobody knows.
>However, properly conducted double blind tests go a long way
>toward proving or disproving these claims.
A lot of this stuff IS known. There is an incredible amount of research
on psychoacoustics out there, much of which is regularly ignored by the
audio community.
There are also some things that everybody knows (ie. low order odd harmonics
sound buzzy, low order even harmonics sound 'full' and 'rounded) but which
haven't really been backed up by proper research. No doubt some of these
things will be found not to be true, but most of them seem okay.
>>> Frequency response, slew rate, power, distortion, phase, etc.
>>> There is always the chance that physics has not caught up with
>>> the human being's abilities, I am not doubting this, however that
>>> would easily be picked up by complete double blind tests.
>>
>> Prove it. I say that D-BT, as promoted by its leading adherents (who sell
>> "scientific" snake oil, but snake oil, nonetheless), is at least incomplete,
>> and at best grossly misleading. It is true only up to a point. And that
>> point is rather distantly removed from the as-yet-unknown truth.
The problem is that as soon as we have good measures to quantify colorations
in the current technology, somebody comes along with a new kind of coloration
and we need to first of all realize it exists and then figure out what it
is before we can figure out how to measure it.
When amplifier topologies were all the same and therefore amplifier distortion
spectra were all pretty much the same, the THD measure was a huge step forward
in identifying why different amplifiers sounded different. It was designed to
measure a particular effect in a particular situation.
As amplifiers advanced, the THD measure became effectively useless. Many
amplifiers exhibited high order distortion products that were very audible
even though the THD measure was extremely low. It became necessarily to
talk about distortion spectra and weighted distortion measures.
But now we have lossy compression systems... and most of them sound pretty
bad. But we don't yet have a measure to explain why they sound bad or to
explicitly state which one will be more colored on a given source material.
As the technology changes, the required measurements change. Folks are
working on this particular measurement as I type this.
>The snake oil is really the people in rags like Stereophile
>subscribing to stuff like marbles wrapped around a line cord and
>such.
There's a lot of that crap in the high end audio world. The weird thing
is that occasionally it turns out there's something to some of it. Don't
blame Stereophile, though. While Stereophile has promoted some of the
silliness they are by no means the worst offender and if anything they have
tried to rein in a bit of it.
Scott Dorsey
I have only heard it at the AES show, which is not a decent place to
audition anything. I'll say that it's beautifully made and that Eelco
Grimm (who once used to post here back in the heyday of this newsgroup)
knows the distortion sources in sigma-delta converter design better than
just about anyone else out there.
If I had the money to take the serious DSD route, I'd look at his box
and also at the Meitner converters. But I haven't seen the money, and
I tend not to invest in any technologies until they are already obsolete....
Moshe
> The problem is that the standard measurements aren't complete. And a lot
> of the measurements that are very powerful at characterizing perceived
> sound quality, like distortion spectra, haven't really been standardized
> although today in the Audio Precision world they are not that hard to
> make accurately.
I don't really disagree with that because it goes along with my
statement that maybe there are things we just don't know about
yet.
I don't discount that anymore than I discount the possibility that
God, or a supreme being created us.
Who knows?
Where did the first one come from?
A paradox I know, but just an example to show how little we may or
may not know.
>>> And how do they correlate with what we hear? What, for example, is the
>>> threshold of harmonic or intermodulation distortion? How does the audibility
>>> of distortion relate to its spectrum or orde?. Can you tell me? If you
>>> can't, ask Arny. He knows /all/ these answers, right?
>>
>>Nobody knows.
>>However, properly conducted double blind tests go a long way
>>toward proving or disproving these claims.
>
> A lot of this stuff IS known. There is an incredible amount of research
> on psychoacoustics out there, much of which is regularly ignored by the
> audio community.
Mostly because of $$$ attachments and fanaticism and fear of being
tarred and feather by ones peers.
Do golen eared people exist?
I believe they do but they are few and far between.
As a personal example, when I was about 15, a child prodigy at the
piano, I was asked by a particular church to pick out a grand
piano for them as some parishioner donated a ton of money upon her
death.
I went to a showroom along with the people holding the checkbook
and played the various pianos.
Everyone I picked, and no price tags were visible, I found out
later they were in the bench, was a $30k or more piano.
Why?
Those pianos sang a song to me when I played them.
To the listeners they all looked alike.
To me, a totally non snob musician, I picked the big buck models
every time.
Go figure.
> There are also some things that everybody knows (ie. low order odd harmonics
> sound buzzy, low order even harmonics sound 'full' and 'rounded) but which
> haven't really been backed up by proper research. No doubt some of these
> things will be found not to be true, but most of them seem okay.
If true, all I ask is that the golden eared people pick them out.
>>>> Frequency response, slew rate, power, distortion, phase, etc.
>>>> There is always the chance that physics has not caught up with
>>>> the human being's abilities, I am not doubting this, however that
>>>> would easily be picked up by complete double blind tests.
>>>
>>> Prove it. I say that D-BT, as promoted by its leading adherents (who sell
>>> "scientific" snake oil, but snake oil, nonetheless), is at least incomplete,
>>> and at best grossly misleading. It is true only up to a point. And that
>>> point is rather distantly removed from the as-yet-unknown truth.
>
> The problem is that as soon as we have good measures to quantify colorations
> in the current technology, somebody comes along with a new kind of coloration
> and we need to first of all realize it exists and then figure out what it
> is before we can figure out how to measure it.
Agreed, but stuff like "brittle, lack of soundstage, etc" need to
be quantified, and I used these very terms myself in this thread.
I can tell you right now, the sound stage of the Event ASP8 I am
listening to right now (Stacy Kent) blows away both the Mackie
HR824 and the Genelec 8040 I have sitting right next to them.
If I ever posted that on Gearslutz they would crucify me.
Yet to *me* the Event sounds *better*, as a monitor not as a
casual listening speaker.
There is so much snobbery in the business.
> When amplifier topologies were all the same and therefore amplifier distortion
> spectra were all pretty much the same, the THD measure was a huge step forward
> in identifying why different amplifiers sounded different. It was designed to
> measure a particular effect in a particular situation.
I have always suspected that IM distortion is the key to all of
this.
Just a thought.
> As amplifiers advanced, the THD measure became effectively useless. Many
> amplifiers exhibited high order distortion products that were very audible
> even though the THD measure was extremely low. It became necessarily to
> talk about distortion spectra and weighted distortion measures.
Negative feedback at work.
The "power wars" of the late 70's were famous for this.
As I recall Sansui won with some massive receiver although Pioneer
was real close.
I always had a place in my heart for Marantz and still have a
2285B that is in need of pwr supply filter caps and dial lamps.
Any idea where to get these?
TIA!!!
> But now we have lossy compression systems... and most of them sound pretty
> bad. But we don't yet have a measure to explain why they sound bad or to
> explicitly state which one will be more colored on a given source material.
> As the technology changes, the required measurements change. Folks are
> working on this particular measurement as I type this.
I've done a number of informal tests on this with mp3 vs ogg and
have reached the conclusion that although the curves, via
Soundforge etc are different, most lay persons can't hear any
difference.
For me, it's the phasing noise (that's what I call it, the
flanging effect)and the screwy artifacts you get, especially with
heavy percussion type music, that kills it for me.
>
>>The snake oil is really the people in rags like Stereophile
>>subscribing to stuff like marbles wrapped around a line cord and
>>such.
>
> There's a lot of that crap in the high end audio world. The weird thing
> is that occasionally it turns out there's something to some of it. Don't
> blame Stereophile, though. While Stereophile has promoted some of the
> silliness they are by no means the worst offender and if anything they have
> tried to rein in a bit of it.
> --scott
I admit I haven't kept up so my information is somewhat dated.
I learned a long time ago to give others the benefit of doubt, but
at the same time I am always on my guard.
As an example I don't believe at all in psychics or mediums etc
but yet when a person who knows nothing about me at all tells me I
skipped 8th grade, my first car was red colored, I grew up in a
certain city and that one of my grandmothers died of a certain
disease I start to consider that there might be something more to
this than the engineer in me has thought of.
True story BTW and I never even met the person face to face.
And most certainly never said a word and none of this is on the
net anywhere.
Scared the hell out of me.
William Sommerwerck
> I believe they do but they are few and far between.
Most people have better hearing than they think. My experience is that most
people will pick "good" equipment if they're simply encouraged to listen
critically and carefully. This process includes keeping them away from
magazine reviews.
> As a personal example, when I was about 15, a child prodigy at the
> piano, I was asked by a particular church to pick out a grand
> piano for them as some parishioner donated a ton of money upon her
> death.
> I went to a showroom along with the people holding the checkbook
> and played the various pianos.
> Every one I picked, and no price tags were visible, I found out
> later they were in the bench, was a $30k or more piano.
> Why?
> Those pianos sang a song to me when I played them.
> To the listeners they all looked alike.
> To me, a totally non snob musician, I picked the big buck models
> every time.
> Go figure.
I'm not the least surprised. As a talented musician, you were more aware of
the subtleties of instrumental sound.
>> There are also some things that everybody knows (ie. low order odd
harmonics
>> sound buzzy, low order even harmonics sound 'full' and 'rounded) but
which
>> haven't really been backed up by proper research. No doubt some of these
>> things will be found not to be true, but most of them seem okay.
> If true, all I ask is that the golden eared people pick them out.
God, I wish you could hear a Crown K-1 or K-2. You would run from the room,
screaming.
> I have always suspected that IM distortion is the key to all of
> this. Just a thought.
It probably is, as IM is not harmonically related to the signal.
I learned about a year ago that I learned the harmonic distortion of
junction transistors varies with the /instantaneous/ base current. Oh,
God...
Don Pearce
>Don Pearce <sp...@spam.com> wrote:
>>
>>I'm not sure I can summon as much interest in the question "why does a
>>poor implementation sound bad?" as I could about the 44.1/96kHz
>>question, if it could indeed be shown to reveal differences.
>
>So, rather than answer the original poster's question, you'd rather answer
>a different and possibly unrelated one?
>
Well, looking at the thread's title, I think we can agree that the
answer is "there isn't one - not even a small one".
>In fact, the question about why a poor implementation sounds bad is actually
>a very interesting one, because it's the converse of asking what makes a
>good implementation sound good.
If you substitute the words inaccurate and accurate for poor and good,
the question, or at least its answer, becomes fairly self evident.
d
Stuart Richards
>
> Stuart Richards <sjr...@yahoo.com> wrote:
> >
> >DSD? In that case I'd love to know if you've heard the Grimm Audio AD1:
> >http://www.grimmaudio.com/pro_converters_ad1.htm
>
> I have only heard it at the AES show, which is not a decent place to
> audition anything. I'll say that it's beautifully made and that Eelco
> Grimm (who once used to post here back in the heyday of this newsgroup)
> knows the distortion sources in sigma-delta converter design better than
> just about anyone else out there.
>
> If I had the money to take the serious DSD route, I'd look at his box
> and also at the Meitner converters. But I haven't seen the money, and
> I tend not to invest in any technologies until they are already obsolete....
Judging from Sony's attitude to SACD I'd say the technology is already
obsolete.
Arkansan Raider
Unfortunate, but true.
---Jeff
Laurence Payne
<goldee_l...@gmail.com> wrote:
>The snake oil is really the people in rags like Stereophile
>subscribing to stuff like marbles wrapped around a line cord and
>such.
>
>The clue to their idiocy is usually statements like "the
>difference was astounding".
>Come on already.
Yeah, that's where they shoot themselves in the foot. William at
least admits that the differences are very subtle - given absolutely
ideal testing conditions over a long term he MIGHT be able to tell one
component from another. These imbeciles bring in an overpriced wooden
acorn and claim the difference is "night and day".
Laurence Payne
>> Next year it was the Allison One.
>> Then the Dahlquist or whatever that flat speaker was called, I
>> can't remember.
I had a pair of Dalquists at one time. They weren't particularly
accurate, but did give a stunning stereo image.
William Sommerwerck
>> is already obsolete.
Tell that to 2L, Alia Vox, BIS, Hanssler, et al. They don't seem to know.
The best-sounding commercial recordings I've ever heard (that is, the ones
that sound most like "live") have been multi-ch SACDs. The recording system
is only one factor, but it seems to be a significant one. Listen, for
example, to the recent Britten "War Requiem" on Hanssler. Living Stereo?
More like comatose.
If SACD is "obsolete", what will it be replaced with? Phonograph records?
William Sommerwerck
> >subscribing to stuff like marbles wrapped around a line cord and
> >such.
> >The clue to their idiocy is usually statements like "the
> >difference was astounding".
> >Come on already.
> Yeah, that's where they shoot themselves in the foot. William at
> least admits that the differences are very subtle -- given absolutely
> ideal testing conditions over a long term he MIGHT be able to tell one
> component from another. These imbeciles bring in an overpriced wooden
> acorn and claim the difference is "night and day".
I said nothing of the sort. In the case of the amplifiers I described, the
differences are gross and instantly audible. There's nothing at all subtle
about it.
However... The tendency of reviewers (my former self included) is to assume
that there are always differences, and to hear differences when they don't
exist, or exaggerate ones that do.
When buying equipment, you need to listen at a time of day when your hearing
is most-acute (for me, it's the morning), relaxed, and with minimal
distractions. Kick back, listen for extended periods, and see what you
think. Rapid switching among components will tend to confuse you.
William Sommerwerck
> particularly accurate, but did give a stunning stereo image.
I think they were pretty accurate -- for an electrodynamic speaker. There's
no question, though, that the exceptionally good imaging greatly helped the
illusion of "reality".
Arny Krueger
>> Basically, and simplistically maybe, what I am saying
>> is that if users can hear a difference, measurements
>> will reveal a difference.
>
> To paraphrase Mozart... Which mesurements would you have
> me make?
I take the typo as a Freudian slip Bill, because you've never shown me that
you could measure anything, or understand what it could possibly mean in
terms of sound quality.
Arny Krueger
>>> To paraphrase Mozart... Which measurements would you
>>> have me make? 25 years ago, I asked JA to take me on as
>>> (in effect) Stereophile's technical editor. As was and
>>> is his wont, he instantly refused. Had he not done so,
>>> we might have made some progress. But objective testing
>>> with more-or-less sure answers does not sell audiophile
>>> magazines.
>
>> All your Stereophile baggage aside, and I don't even
>> know what it is as I must have missed that chapter, but
>> times have changed.
>
> No, they haven't. Nothing has changed in the past 60
> years. People are still arguing over matters of which
> they have neither practical or philosophical
> understanding.
What hasn't changed Bill would appear to be your lack of understanding of
audio technology.
What's changed in the last 60 years is that some very bright and hard
working people have learned a great deal about how the ear works. You
probably don't even know their names, let alone what they accomplished.
Laurence Payne
<grizzle...@comcast.net> wrote:
>> Yeah, that's where they shoot themselves in the foot. William at
>> least admits that the differences are very subtle -- given absolutely
>> ideal testing conditions over a long term he MIGHT be able to tell one
>> component from another. These imbeciles bring in an overpriced wooden
>> acorn and claim the difference is "night and day".
>
>I said nothing of the sort. In the case of the amplifiers I described, the
>differences are gross and instantly audible. There's nothing at all subtle
>about it.
Where you've picked out some models which you DO believe sound grossly
different (and you may well be right) the discussion of blind testing
is irrelevant. Blind, sighted, "ask the wife" all will obviously give
the same result.
I'm talking about the comparisons where you feel a perfunctory DB
session is inadequate. Therefore differences must be extremely
subtle.