WOW.... What a difference in Amps!! - Page 21 - AVS Forum
Forum Jump: 
 
Thread Tools
Old 12-17-2009, 02:15 PM
AVS Addicted Member
 
Chu Gai's Avatar
 
Join Date: Sep 2002
Location: NYC area
Posts: 14,968
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 319 Post(s)
Liked: 596
Quote:
Originally Posted by Magnefied View Post

So 5 people in an ABX test meet your criteria for statistically accurate tests with a diverse demographic? Wow! You're easy!

Of course not. No more so than a single person measuring the deflection of a star's position during an eclipse means that Einstein's equations were unequivocally correct or me saying that if you can't see the double star in the Big Dipper's handle your vision isn't 20/20. The tests which are done with these small sample sizes illustrate something else entirely that was done years ago.

Quote:


I used 2 people for simplicity to illustrate a point. Yes, results get better with a larger sample size. But, like you, every ABX test I've ever seen was conducted by a couple of geeks on a handful of their geek buddies. Hardly a significant or randomly chosen sample of the market and a pretty lousy demographic slice.

Certainly and the demographic that's being tested, people signficantly removed from their teens and having been exposed to sounds and noises for many years, have significantly worse hearing on a multiple of levels than young males and females. They do have money though.
Quote:


Then I don't know what we're arguing about. I agree that ABX testing is a conceptually & practically flawed technique to assess the differences in something like audio equipment. Unless you can control all the variables the test is flawed. And you can't.

I suggest you look up the work of people like Florian when it comes to the concept of Just Noticeable Differences which were run on very large sample sets using test tones which are a more sensitive probe than say music where one has to deal with masking effects and the problems with auditory focus. You'll find that our hearing is most sensitive in and around that 1KHz region which is why we look to control variations between amps to within 0.1 dB or so there. Then after you're finished with that you might be curious to read about the audibility of jitter then follow that up with work done by Benjamin & Gannon of Dolby which tested hundreds of people subjected to jitter using music.

Plenty of work is done on developing audio codecs using music of various sorts and tested by relatively small groups which are then replicated by other small groups or even single individuals in order to establish things like audible transparency. A good place for that is hydrogenaudio.

"I've found that when you want to know the truth about someone that someone is probably the last person you should ask." - Gregory House
Chu Gai is online now  
Sponsored Links
Advertisement
 
Old 12-17-2009, 05:22 PM
Senior Member
 
Tonmeister2008's Avatar
 
Join Date: Dec 2008
Location: Westlake Village, CA, USA
Posts: 291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 25
Quote:
Originally Posted by Magnefied View Post

Scientific does not = valid in every case.

Let's say you're partially deaf or you just like a certain frequency response. You sit through a double blind listening test of two speakers and proclaim the winner to be behind button "A".

Then I sit through the same listening test in the same room with the same equipment. But, I've got tinnitus or maybe I like a warmer or brighter frequency response. I pick the speaker behind button "B".

The double blind test was administered properly and adhered to scientific principles.

Who's right? Which is the better speaker?

The answer is speaker "B" because that's the one I chose. So much for science.

You've described a listening test that from my viewpoint does not adequately control an important nuisance variable: the hearing of the listener.

Toole has shown that broadband hearing loss can produce unreliable and unpredictable loudspeaker preference ratings. Since hearing loss increases naturally with age (presbycusis) and exposure to loud noise/music, some of less desirable listeners in terms of discrimination and reliability are older listeners, rock n' rollers, and sound engineers/audiophiles who have had prolonged exposure to loud sound. Audio has its occupational hazards which is damaged hearing.

That is why we test the hearing of our listeners regularly, and select them on the basis of their hearing and performance in various critical listening tasks.

Cheers,
Sean Olive
My Audio Blog
Tonmeister2008 is offline  
Old 12-17-2009, 05:34 PM
Advanced Member
 
par4's Avatar
 
Join Date: Jul 2005
Posts: 956
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 19
Quote:
Originally Posted by Magnefied View Post

Uh oh! You're challenging the scientists with logic. It...does...not...com...pute! Science is God! You have just challenged their faith.

Indeed. This is an outrage!
par4 is offline  
Old 12-17-2009, 05:53 PM
Advanced Member
 
Magnefied's Avatar
 
Join Date: Nov 2009
Posts: 595
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Tonmeister2008 View Post

You've described a listening test that from my viewpoint does not adequately control an important nuisance variable: the hearing of the listener.

Toole has shown that broadband hearing loss can produce unreliable and unpredictable loudspeaker preference ratings. Since hearing loss increases naturally with age (presbycusis) and exposure to loud noise/music, some of most undesirable listeners based on discrimination and reliability are older listeners, rock n' rollers, and sound engineers/audiophiles who have had prolonged exposure to loud sound. Audio has its occupational hazards which is damaged hearing.

That is why we test the hearing of our listeners regularly, and select them on the basis of their hearing and performance in various critical listening tasks.

Well how would you propose that we "control" the hearing of the listeners? I suppose that you would hand select your "golden eared" listening panel and pronounce their judgement to be "the word".

Personally, I think the "golden ears" should represent no more than .0000000001% of the sample so that they are weighted properly in relation to their market share. I think that the sample should represent all ages and genders, all cultures, all musical tastes, all income levels, etc. Then you may be able to start to draw some conclusions. A panel of hand chosen geeks with "golden ears" may hear things one way but the average Jane/Joe that relies on their opinion may be in for a rude awakening when they rely on the opinion of your panel of experts chosen because they all share certain characteristics. That's hardly science.

Of course if you don't want to sell to older wealthier customers, or musicians, or sound engineers, or constructions workers, or factory workers, or truck drivers, or airline crews, or military, or the Nascar crowd, or oil rig operators, or rail employees, or well, you get the point.....then you're on exactly the right track. If you're only trying to market to golden eared geek wannabe's with perfect hearing you'll do well with the .0000000001% of the market you're targeting. No wonder mainstream audio is going belly up and high end brands are folding on a weekly basis. After all these decades, the geeks still don't have a clue.
Magnefied is offline  
Old 12-17-2009, 06:01 PM
AVS Addicted Member
 
Chu Gai's Avatar
 
Join Date: Sep 2002
Location: NYC area
Posts: 14,968
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 319 Post(s)
Liked: 596
Quote:
Originally Posted by Tonmeister2008 View Post

You've described a listening test that from my viewpoint does not adequately control an important nuisance variable: the hearing of the listener.

Toole has shown that broadband hearing loss can produce unreliable and unpredictable loudspeaker preference ratings. Since hearing loss increases naturally with age (presbycusis) and exposure to loud noise/music, some of most undesirable listeners based on discrimination and reliability are older listeners, rock n' rollers, and sound engineers/audiophiles who have had prolonged exposure to loud sound. Audio has its occupational hazards which is damaged hearing.

That is why we test the hearing of our listeners regularly, and select them on the basis of their hearing and performance in various critical listening tasks.

Have you performed a sufficient number of tests with people that do have hearing loss and attempted to correlate that with speaker preference, Sean? Perhaps segmenting the the population into various bins? What I'm getting at is that while people with satisfactory hearing may generally agree upon a particular speaker, might further study suggest that different characteristics with say respect to FR might appeal to a different segment of the population? Just curious.

"I've found that when you want to know the truth about someone that someone is probably the last person you should ask." - Gregory House
Chu Gai is online now  
Old 12-17-2009, 06:16 PM
Advanced Member
 
Magnefied's Avatar
 
Join Date: Nov 2009
Posts: 595
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Chu Gai View Post

Of course not. No more so than a single person measuring the deflection of a star's position during an eclipse means that Einstein's equations were unequivocally correct or me saying that if you can't see the double star in the Big Dipper's handle your vision isn't 20/20. The tests which are done with these small sample sizes illustrate something else entirely that was done years ago.

Certainly and the demographic that's being tested, people signficantly removed from their teens and having been exposed to sounds and noises for many years, have significantly worse hearing on a multiple of levels than young males and females. They do have money though.
I suggest you look up the work of people like Florian when it comes to the concept of Just Noticeable Differences which were run on very large sample sets using test tones which are a more sensitive probe than say music where one has to deal with masking effects and the problems with auditory focus. You'll find that our hearing is most sensitive in and around that 1KHz region which is why we look to control variations between amps to within 0.1 dB or so there. Then after you're finished with that you might be curious to read about the audibility of jitter then follow that up with work done by Benjamin & Gannon of Dolby which tested hundreds of people subjected to jitter using music.

Plenty of work is done on developing audio codecs using music of various sorts and tested by relatively small groups which are then replicated by other small groups or even single individuals in order to establish things like audible transparency. A good place for that is hydrogenaudio.

Chu Gai, I don't buy audio equipment to listen to test tones. Few people do. A few really weird people. If a "difference" is masked by music then by definition it is irrelevant because it can't be heard. If a tweeter has a resonance at 1 GHz I don't really care. I suggest you read up on psychoacoustics and take a look at where the money has been going in the music industry for the last five years.

As for the old geezers with money, I fully agree. I'm one of them. That was my point. And I'll bet a hip hop fan in their 30's would assess a music system from a fundamentally different viewpoint than me.
Magnefied is offline  
Old 12-17-2009, 06:35 PM
AVS Addicted Member
 
Chu Gai's Avatar
 
Join Date: Sep 2002
Location: NYC area
Posts: 14,968
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 319 Post(s)
Liked: 596
Perhaps a person such as yourself can suggest some good reading material on pyschoacoustics. I see the money going to more and more compressed material and digital downloads as well as providing the music industry with the increasing ability to control what the listener can listen to, how often, copy, and share.

"I've found that when you want to know the truth about someone that someone is probably the last person you should ask." - Gregory House
Chu Gai is online now  
Old 12-17-2009, 08:41 PM
Senior Member
 
Tonmeister2008's Avatar
 
Join Date: Dec 2008
Location: Westlake Village, CA, USA
Posts: 291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 25
Quote:
Originally Posted by Magnefied View Post

Well how would you propose that we "control" the hearing of the listeners? I suppose that you would hand select your "golden eared" listening panel and pronounce their judgement to be "the word".

Personally, I think the "golden ears" should represent no more than .0000000001% of the sample so that they are weighted properly in relation to their market share. I think that the sample should represent all ages and genders, all cultures, all musical tastes, all income levels, etc. Then you may be able to start to draw some conclusions. A panel of hand chosen geeks with "golden ears" may hear things one way but the average Jane/Joe that relies on their opinion may be in for a rude awakening when they rely on the opinion of your panel of experts chosen because they all share certain characteristics. That's hardly science.

Of course if you don't want to sell to older wealthier customers, or musicians, or sound engineers, or constructions workers, or factory workers, or truck drivers, or airline crews, or military, or the Nascar crowd, or oil rig operators, or rail employees, or well, you get the point.....then you're on exactly the right track. If you're only trying to market to golden eared geek wannabe's with perfect hearing you'll do well with the .0000000001% of the market you're targeting. No wonder mainstream audio is going belly up and high end brands are folding on a weekly basis. After all these decades, the geeks still don't have a clue.

Our definition of normal hearing is that the listener has no more than 20 HL at any audiometric frequency between 250 Hz to 8 kHz. That criterion includes a fairly large percentage of the population, particularly below 50 years of age.

I'm not sure how you define "golden ears" because that's mostly a self-reported, unquantified term that doesn't have much meaning to me. We train, measure and select employees as listeners based on their proven ability to discriminate and rate sound quality attributes in a consistent fashion.

We have compared the sound quality preferences of our listening panel to different populations of consumers that we're interested in. From this, we have learned exactly how the results from our trained listening panel map to the consumer responses. So, while your concerns about whether our test results can be generalized to consumers are warranted, it is an issue that we have already addressed and resolved.

Cheers,
Sean Olive
My Audio Blog
Tonmeister2008 is offline  
Old 12-17-2009, 08:54 PM
Senior Member
 
Tonmeister2008's Avatar
 
Join Date: Dec 2008
Location: Westlake Village, CA, USA
Posts: 291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 25
Quote:
Originally Posted by Chu Gai View Post

Have you performed a sufficient number of tests with people that do have hearing loss and attempted to correlate that with speaker preference, Sean? Perhaps segmenting the the population into various bins? What I'm getting at is that while people with satisfactory hearing may generally agree upon a particular speaker, might further study suggest that different characteristics with say respect to FR might appeal to a different segment of the population? Just curious.

No, I feel that not enough controlled tests have been done on hearing impaired listeners to be able to determine if they tend to prefer a certain type of loudspeaker, and how to characterize its performance.

Hearing loss is very nonlinear and it's only the lower 30-40 dB part of the dynamic range of music is not heard. Once the level of reproduced music gets above their elevated hearing threshold, the hearing of the listener is more or less the same as a normal listener (a process known as recruitment).

Maybe the heavily compressed, limited dynamic range of today's music is actually beneficial to the hearing impaired (an increasing segment of the population)? Personally, I'd rather the music industry restore the dynamic range of recorded music and let our hearing aides do the work for us. Or else put the compressor in the receiver/ipod/head-unit and let the end user decide how compressed they want their music.

Cheers,
Sean Olive
My Audio Blog
Tonmeister2008 is offline  
Old 12-17-2009, 09:42 PM
AVS Special Member
 
CharlesJ's Avatar
 
Join Date: Jan 2006
Posts: 3,413
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 190 Post(s)
Liked: 117
Quote:
Originally Posted by Tonmeister2008 View Post

....
Hearing loss is very nonlinear and it's only the lower 30-40 dB part of the dynamic range of music is not heard. ....

But, isn't a good part of that range also below many room's noise floor as well? And, how much well recorded music is into those ranges?
Certainly not talking about your testing chamber's floor here.
CharlesJ is offline  
Old 12-17-2009, 10:06 PM
AVS Special Member
 
Fanaticalism's Avatar
 
Join Date: Jun 2008
Posts: 4,394
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 17
Quote:
Originally Posted by Magnefied View Post

No, simply to illustrate a point. You seem to want to take everything literally.

Not so much that I take everything literally, but wanting to clarify for the unsuspecting because we are having a conversation, I feel it is important that those who may just be reading for different POV's understand every aspect.


Quote:
So 5 people in an ABX test meet your criteria for statistically accurate tests with a diverse demographic? Wow! You're easy!

It was an example, and an attempt to drive home a point which you obviously understood by your comment below. Talk about taking things literally.


Quote:
I'll say it once again, you seem to take a literal interpretation of every statement as long as it suits your purpose.

And you seem to respond with much more than just reasoning, but emotion, which I don't get.

Quote:
Since we're typing a conversation here, it gets a bit tedious to say or read a thorough discussion of anything.

I do not disagree

Quote:
I used 2 people for simplicity to illustrate a point. Yes, results get better with a larger sample size. But, like you, every ABX test I've ever seen was conducted by a couple of geeks on a handful of their geek buddies. Hardly a significant or randomly chosen sample of the market and a pretty lousy demographic slice.


Then I don't know what we're arguing about. I agree that ABX testing is a conceptually & practically flawed technique to assess the differences in something like audio equipment. Unless you can control all the variables the test is flawed. And you can't.

From what I've read from your posts and response to Sean Olive, it seems to me (correct if my wrong) that you feel that DBT's are flawed because they do not take into account hearing ability/disability, environment, and other anomalies.

The point is not to see if PEOPLE can hear a difference, because we all know people believe they do, but, IS there actually a difference. The only way to know for sure if these differences are a proponent of the equipment, or just our perception/placebo is by removing bias.
Fanaticalism is offline  
Old 12-21-2009, 11:44 PM
AVS Special Member
 
NIN74's Avatar
 
Join Date: May 2001
Location: Skoghall, va, Sweden
Posts: 3,572
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 14
Quote:
Originally Posted by Magnefied View Post

Scientific does not = valid in every case.

Let's say you're partially deaf or you just like a certain frequency response. You sit through a double blind listening test of two speakers and proclaim the winner to be behind button "A".

Then I sit through the same listening test in the same room with the same equipment. But, I've got tinnitus or maybe I like a warmer or brighter frequency response. I pick the speaker behind button "B".

The double blind test was administered properly and adhered to scientific principles.

Who's right? Which is the better speaker?

The answer is speaker "B" because that's the one I chose. So much for science.

You have not really thought about this right? So let me get this, DBT on medicine is not working because the side-effects are not universal on all people that use the drug?

Still waiting for ANY proof why DBT is not scientific.

Sound and video is not magic, it is pure physics. Physics that can be magical
NIN74 is offline  
Old 12-22-2009, 08:36 AM
Advanced Member
 
Lhasa-lover's Avatar
 
Join Date: Feb 2009
Location: Georgia
Posts: 737
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Not offering this up as factual but only as interesting data. What is factual is that I suffer from tinnitus in my left ear from 40 years of guitar playing. Significant ringing. In sighted or unsighted tests, and ABX tests, I have never been able to consistantly pick out any differences in sound between ss amplifiers or receivers. On the other hand, without doubt I constantly and reliably pick speakers in sighted and unsighted tests that others consider "bright" or "forward." I have owned or listened to hundreds of speakers ranging from inexpensive to very expensive. And without any exception always select bright speakers as sounding the best. Blind or sighted. (Triangle, older vintage Klipsch, montior, etc.) at almost a 100% hit rate. Don't know what that proves but I do think that hearing loss affects the way I perceive what I hear. If amps sounded audibly different and my loss was significant enough to render those differences in amps negligible to me, then perhaps the same might prove true for speakers and I should not be able to hear the differences that I do?
Lhasa-lover is offline  
Old 12-22-2009, 03:44 PM
Member
 
serenity737's Avatar
 
Join Date: Jul 2007
Location: Colorado Springs, CO
Posts: 45
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Double blind test design combined with randomized assignment - sampling is the Gold Standard for experimental design to assess differences between groups. There are a variety of methodologies including ABAB design used for single case design. All of the above methods (and other designs) are used as part of a scientific approach with human subjects.

So let's move on ... and get back to the issue of whether they are discernable differences between amps.

Best,

Mike
serenity737 is offline  
Old 12-23-2009, 09:05 AM
Member
 
milpai's Avatar
 
Join Date: Jun 2009
Location: USA
Posts: 46
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Funny thread. Look at where it started and where it is going.
For people who mention that all amps should sound the same, provided configuration A, B, C, etc in all the amps are the same - that's like saying if God created all human with same specifications, then all men would be either Arnold Schwarzenegger or Albert Einstein!! That would be scary And we would either have all days in the week either as Sundays or Mondays!! Luv ur concept guys!!
milpai is offline  
Old 12-23-2009, 09:34 AM
AVS Addicted Member
 
penngray's Avatar
 
Join Date: Feb 2006
Location: Florida
Posts: 26,779
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 30
Quote:
Originally Posted by milpai View Post

Funny thread. Look at where it started and where it is going.
For people who mention that all amps should sound the same, provided configuration A, B, C, etc in all the amps are the same - that's like saying if God created all human with same specifications, then all men would be either Arnold Schwarzenegger or Albert Einstein!! That would be scary And we would either have all days in the week either as Sundays or Mondays!! Luv ur concept guys!!

Nice try!!

Do you know anything about Tom Nousaine? How about the likes of Sean Olive, Geddes or Linkwitz?

You can just ask an engineer like Bob Lee from QSC about what matters in amps and the fact that if they are built PROPERLY the actually do not alter sound and SHOULD sound very similar. Its a simple fact that people exaggerate the differences daily and the biggest problem is that those people do not understand how to control their listening experience.

If you can see the name, see the price, see the amp, see the salesman's expressions during the demo you are simply tainting the test and your conclusions are not going be valid.

btw, no one has ever posted that all amps sound the same so get off the dead horse!
Heck, I will dumb it down...

I have a Behringer A500 amp that we can put against similar class amps 10x its price tag. I welcome any $$ bets on if you can spot the better amp alteast 70% of the time. Also before you think I have never owned expensive amps. I have Sunfire amps too and Outlaw monoblocks. The whole idea that one amp is better then another sonically constantly proves the ignorance in audio. Speakers, room design, room treatements will ALWAYS impact sonic performance 1000 times more then any amp change once you find a 100% working amp.

It is not "open-minded" to reject knowledge - Bob Lee
penngray is offline  
Old 12-23-2009, 09:35 AM
AVS Addicted Member
 
penngray's Avatar
 
Join Date: Feb 2006
Location: Florida
Posts: 26,779
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 30
Those who want differences really just want flaws......It doesn't make sense to me but then again the audio world is full of subjective ignorance anyways so why should this flawed mindset be different.

It is not "open-minded" to reject knowledge - Bob Lee
penngray is offline  
Old 12-23-2009, 09:52 AM
Advanced Member
 
par4's Avatar
 
Join Date: Jul 2005
Posts: 956
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 19
Thanks for dumbing it down........
par4 is offline  
Old 12-23-2009, 10:49 AM
AVS Special Member
 
A9X-308's Avatar
 
Join Date: Mar 2008
Location: Australia; now run by adults.
Posts: 5,468
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 181 Post(s)
Liked: 115
Quote:
Originally Posted by milpai View Post

Funny thread. Look at where it started and where it is going.
For people who mention that all amps should sound the same, provided configuration A, B, C, etc in all the amps are the same - that's like saying if God created all human with same specifications, then all men would be either Arnold Schwarzenegger or Albert Einstein!! That would be scary And we would either have all days in the week either as Sundays or Mondays!! Luv ur concept guys!!

No, you have simply misunderstood. The only 'same' specifications I recall are reasonable distortion, signal to noise, bandwidth, output impedance and that the amps not be clipping. The first three are usually trivially easy to engineer into an amplifier, the fourth also for SS amps but not for tube amps. All of these if outside 'reasonable' parameters will make one device obvious to identify simply because of that aspect. Output impedance makes changes to the frequency response of a non flat load such as that of a speaker and is easy to see from the results in published measurements, eg from Stereophile and these variations alone are greater than what can easily be detected as FR differences, so enough to ID two different units from that alone.

Amps when clipping into brief transients to not sound audibly clipped or distorted but they can sound different enough to be easily detectable for that reason alone because of the shape of individual transfer functions, how the circuit reacts to overload etc. Bob Cordell did some testing on this aspect, reported here. So just keep them within their normal working parameters for the test, a not unreasonable aim.
A9X-308 is offline  
Old 12-23-2009, 11:04 AM
AVS Addicted Member
 
penngray's Avatar
 
Join Date: Feb 2006
Location: Florida
Posts: 26,779
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 30
Quote:
Originally Posted by par4 View Post

Thanks for dumbing it down........

I need to for you guys once in awhile since you are not into audio education

It is not "open-minded" to reject knowledge - Bob Lee
penngray is offline  
Old 12-23-2009, 11:09 AM
Advanced Member
 
Magnefied's Avatar
 
Join Date: Nov 2009
Posts: 595
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by NIN74 View Post

You have not really thought about this right? So let me get this, DBT on medicine is not working because the side-effects are not universal on all people that use the drug?

Still waiting for ANY proof why DBT is not scientific.

I haven't thought about it? You're kidding right? If this is your response after "thinking about it", your short bus is waiting.

So, you'd equate ABX testing in the audio industry to drug trials in the pharmaceutical industry. I don't think so.

Pharma Drug Trial Characteristics
  • Oversight by a regulatory agency, the FDA
  • Subject to peer review
  • Time frame of the tests can be years
  • Subjects numbers in the hundreds if not thousands with a broad demographic selected for a carefully chosen set of measurable criteria.
  • DOE (design of experiment) is applied to block known sources of variation.

Audio Trial Characteristics
  • Zero oversight by any regulatory body
  • Seldom any peer review
  • Time frame of a few hours
  • Subjects number in the single digits and usually consist of geeks testing other geeks with "superior hearing".
  • DOE (design of experiment) can't help much with blocking for room variation since that's one of the things you want to know about and one of the things you can't reliably quantify up front.
  • Usually takes place in a single carefully prepared (or not) room with a single set of equipment.
  • Fundamentally cannot take into account the impact one of the most the important variables in the listening process for a key component.....room interactions with speakers!

The last one is the killer Ace. Room interactions will obliterate any infinitesimal difference in electronic distortion or digital errors of source components and amplifiers. So explain how your ABX testing will remove the room from the equation. Explain how a test done on say a speaker in one room with a specific set of equipment will be totally valid in another room with a totally different set of equipment. And explain why your subjective judgement in that room will match my subjective judgement in another room. ABX testing is fine in theory but it has problems when it comes to practical application of the theory.

Drug trials certainly have to consider environmental conditions but seldom is the environment so directly impacting each and every test subject so dramatically as it does with audio testing.

Of course, this is just me not thinking. What are your thoughts?
Magnefied is offline  
Old 12-23-2009, 11:11 AM
Advanced Member
 
par4's Avatar
 
Join Date: Jul 2005
Posts: 956
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 19
Quote:
Originally Posted by penngray View Post

I need to for you guys once in awhile since you are not into audio education

Much gratitude for passing along that great wealth of knowledge, professor. Always enlightening to the slow witted and those who frequent this forum and are unable to read or comprehend.........
par4 is offline  
Old 12-23-2009, 11:17 AM
AVS Addicted Member
 
penngray's Avatar
 
Join Date: Feb 2006
Location: Florida
Posts: 26,779
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 30
Quote:
Originally Posted by par4 View Post

Much gratitude for passing along that great wealth of knowledge, professor. Always enlightening to the slow witted and those who frequent this forum and are unable to read or comprehend.........

I do like to come over to the trailer park once in awhile just to remember how the uneducated masses live

It is not "open-minded" to reject knowledge - Bob Lee
penngray is offline  
Old 12-23-2009, 11:18 AM
AVS Addicted Member
 
penngray's Avatar
 
Join Date: Feb 2006
Location: Florida
Posts: 26,779
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 30
Quote:
Originally Posted by Magnefied View Post

I haven't thought about it? You're kidding right? If this is your response after "thinking about it", your short bus is waiting.

So, you'd equate ABX testing in the audio industry to drug trials in the pharmaceutical industry. I don't think so.

Pharma Drug Trial Characteristics
  • Oversight by a regulatory agency, the FDA
  • Subject to peer review
  • Time frame of the tests can be years
  • Subjects numbers in the hundreds if not thousands with a broad demographic selected for a carefully chosen set of measurable criteria.

Audio Trial Characteristics
  • Zero oversight by any regulatory body
  • Seldom any peer review
  • Time frame of a few hours
  • Subjects number in the single digits and usually consist of geeks testing other geeks with "superior hearing".
  • Usually takes place in a single carefully prepared (or not) room with a single set of equipment.
  • Fundamentally cannot take into account the impact one of the most the important variables in the listening process for a key component.....room interactions with speakers!

The last one is the killer Ace. Room interactions will obliterate any infinitesimal difference in electronic distortion or digital errors of source components and amplifiers. So explain how your ABX testing will remove the room from the equation. Explain how a test done on say a speaker in one room with a specific set of equipment will be totally valid in another room with a totally different set of equipment. And explain why your subjective judgement in that room will match my subjective judgement in another room.

Drug trials certainly have to consider environmental conditions but seldom is the environment so directly impacting each and every test subject so dramatically as it does with audio testing.

Of course, this just me not thinking. What are your thoughts?

Full treated rooms are very controlled and the same speaker testing two amps isnt going to be altered by the room for one amp and not the other so Im not sure what your point about the room is?


No test is perfect but don't you guys think its important to remove all those uncontrollable variables that constantly distort the brains ability to make a valid conclusion??

It is not "open-minded" to reject knowledge - Bob Lee
penngray is offline  
Old 12-23-2009, 12:31 PM
Member
 
serenity737's Avatar
 
Join Date: Jul 2007
Location: Colorado Springs, CO
Posts: 45
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
If you look at some of Sean Olive's work well-controlled using hundreds of non-expert regular subjects he has found little or no difference (except for being able to describe their experience in a nuanced way) between their ability to distinguish between or preference for sound characteristics of audio components and responses of "experts." His conclusion, well supported by research with many subjects, is expert panels of audiophyles can be used reliably to predict consumer preference for audio components.

Although the neuroscience literature suggests they may be some very select groups, especially expert professional musicians, who at a neurological level simply process sound - music better than the rest of us, the conclusion for most people holds - by and large as long as our hearing is intact, we tend to process sound, to perceive it and develop acoustic preferences in similar ways.

So what difference might this make for amplifyer design? Not much from my take on it except for the issue of short-term dynamics and potential clipping. Rather than a demonstrate at a convention, I would like to see some of the interesting work cited replicated in a more controlled setting.

Best,

Mike
serenity737 is offline  
Old 12-23-2009, 05:32 PM
AVS Special Member
 
NIN74's Avatar
 
Join Date: May 2001
Location: Skoghall, va, Sweden
Posts: 3,572
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 14
Quote:
Originally Posted by Magnefied View Post

I haven't thought about it? You're kidding right? If this is your response after "thinking about it", your short bus is waiting.


Waiting for you, yes.


Quote:
Originally Posted by Magnefied View Post

Audio Trial Characteristics[list][*]Zero oversight by any regulatory body


Don't needed. All is needed is a good knowledge how to do the test correctly. That is not hard to do at all, well maybe for audiophiles that are scared to have their pants down



Quote:
Originally Posted by Magnefied View Post

[*]Seldom any peer review


Wrong again. It is easy and are often done, that people listen to the equipment that they will test openly before, to hear, talk and see if they can hear anything difference. That they later on test if they can hear it in the DBT.
Are you sure you really knows ANYTHING about this?


Quote:
Originally Posted by Magnefied View Post

[*]Time frame of a few hours


Absolutely wrong!!! Nothing is said that the time frame must be a couple of hours. One can do an openly test for 4 weeks and then do the blind test for as long as one wants. NOTHING about DBT says that it must be done within a couple of hours.



Quote:
Originally Posted by Magnefied View Post

[*]Subjects number in the single digits and usually consist of geeks testing other geeks with "superior hearing".


Faulty thinking. If "Steve" say he can hear a difference between amp A and amp B, then we have a subject that can be one of the people in the test. Do Steves result give objective result what ALL will hear? No off course not, nobody have said that. But it show if he can still hear the difference, that he already have said he hears, when he don't know what amp is playing.

But now to the point. If Steve fail to hear any difference in the DBT, it is just one test and don't say anything about different sound between amps, just that he could not hear the difference and it was probably placebo that made him think he could hear a difference in open test.
BUT, now for the kicker, when you have 1000 persons, that hear difference in open tests, fail the DBT, we can make some interesting points:

*That many people can only hear the difference when they know what they are listning to.
*If all 1000 persons fail the test, we can NOT say that all amps sound the same, but the difference between their open test and DBT and that all fail, we can guess that if more would do the test, it would give a similar result.



Quote:
Originally Posted by Magnefied View Post

[*]DOE (design of experiment) can't help much with blocking for room variation since that's one of the things you want to know about and one of the things you can't reliably quantify up front.


Wrong again! Nothing about DBT says it must be done in a prison and that the inmates will rape all that don't succeed
The fact is that many of those test are made in the home of the person that make the claim of difference. So if he have already listen to the amps, for some time, in his own system, own room and will do the test with his own system, in his own room and whatever time he wants, and still fail, what can you tell about that?


Quote:
Originally Posted by Magnefied View Post

[*]Usually takes place in a single carefully prepared (or not) room with a single set of equipment.


Wrong! Nothing with DBT say it must or will be done so. What have you "learned" that nonsense from?


Quote:
Originally Posted by Magnefied View Post

[*]Fundamentally cannot take into account the impact one of the most the important variables in the listening process for a key component.....room interactions with speakers!


And AGAIN you say something that have NOTHING to do with DBT. DBT is a M-E-T-H-O-D, it do not say anything how long you must do it, where you must do it, and with what system you must do it. NOTHING!


Quote:
Originally Posted by Magnefied View Post

Of course, this is just me not thinking. What are your thoughts?


That you don't know a thing about DBT and you should probably listen and read more before making everybody see that you don't know anything about it.

So whats your thoughts on this lesson I gave you?

Sound and video is not magic, it is pure physics. Physics that can be magical
NIN74 is offline  
Old 12-23-2009, 05:45 PM
Advanced Member
 
Magnefied's Avatar
 
Join Date: Nov 2009
Posts: 595
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by penngray View Post

Full treated rooms are very controlled and the same speaker testing two amps isnt going to be altered by the room for one amp and not the other so Im not sure what your point about the room is?


No test is perfect but don't you guys think its important to remove all those uncontrollable variables that constantly distort the brains ability to make a valid conclusion??

And every room in the world is "full treated", right? Well, most rooms, right? Well how about some rooms? I think we may have to settle for a few rooms. A few rooms are designed and constructed to minimize unwanted reflections and resonances at the listening position at most frequencies. So if you perform your testing in your "full treated" room, the results are completely transferable to every other room in the world, right? And your subjective interpretation of any differences you hear are exactly the same as my subjective response, right? And the equipment you used to power the speakers and produce the signal is the exact same equipment that I use, right? That's my issue.

People seem to think that ABX testing results are equivalent to some immutable law of nature on the order of E=M*C^2. It's not. At least not when you're dealing with people trying to convey their subjective reaction to some stimulus. Unfortunately, it is about the best tool available now or for the foreseeable future. But it's an imperfect tool when used to gather people's qualitative opinions about "good" and "bad".

You can't get a person to tell you how much "jitter" they hear. They can't tell you how much time alignment is off. They can't tell you they can hear 18,237Hz but not 18, 275Hz. They can't give you anything quantifiable other than "I liked that one better than the other ones in this room, with this equipment". How does that help me?
Magnefied is offline  
Old 12-23-2009, 06:06 PM
Advanced Member
 
Magnefied's Avatar
 
Join Date: Nov 2009
Posts: 595
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by NIN74 View Post


So whats your thoughts on this lesson I gave you?

I'm thinking you're gonna miss the short bus if you don't stop hanging around internet forums babbling incoherent pseudo-science.

So dig up one double blind test on any piece of equipment of your choosing
  • That involved thousands of people.
  • Where the tests were performed in each of the subject's own listening rooms.
  • Where the test for each subject lasted for weeks.
  • And the data was reviewed by a peer group not connected in any way to the study or the equipment manufacturers.

I'll wait till the short bus brings you home from school. But since you assert that this is how audio equipment reviews are performed, I have no doubt you'll deliver the goods. No doubts at all.
Magnefied is offline  
Old 12-23-2009, 06:26 PM
AVS Special Member
 
David James's Avatar
 
Join Date: Dec 2000
Location: Longmont, CO
Posts: 1,945
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 22 Post(s)
Liked: 72
Quote:
Originally Posted by Magnefied View Post

They can't give you anything quantifiable other than "I liked that one better than the other ones in this room, with this equipment". How does that help me?

hmmm, I thought the issue was telling if there is difference, not one is better then the other? Or has the thread veered off topic?

So...the issue of treated becomes less of an issue, as it's still an apples to apples comparison. Unless someone would argue that boutique amp A requires room treatments X, Y & Z, but not X & Y or A & B etc. etc. to show off it's boutique-ness.
David James is offline  
Old 12-23-2009, 07:07 PM
AVS Special Member
 
NIN74's Avatar
 
Join Date: May 2001
Location: Skoghall, va, Sweden
Posts: 3,572
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 14
Quote:
Originally Posted by Magnefied View Post

I'm thinking you're gonna miss the short bus if you don't stop hanging around internet forums babbling incoherent pseudo-science.


You are joking, right? You just making yourself look like a fool.



Quote:
Originally Posted by Magnefied View Post

[*]That involved thousands of people.


There have been extremly many BT over the past 25 years on amps, just try google, you maybe learn somthing.


Quote:
Originally Posted by Magnefied View Post

[*]Where the tests were performed in each of the subject's own listening rooms.


Already done. A classic one is the one that Tom Nousaine did at some guys house that showed that he could not tell the difference between his 20 000 amplifier and a 200 reciever



Quote:
Originally Posted by Magnefied View Post

[*]Where the test for each subject lasted for weeks.


Don't know if anyone have done it so long but you can if you want. What you don't understand is the fact that soundmemory is a short term memory, and that shorter testcycles are the best way to find a difference.
But again, please understand this simple fact, that DBT do NOT say anything about the time, place, system, or music. It is all up to the people that do the test. It is a fact and unless you show me some proof a would say that you don't know what you are saying.


Quote:
Originally Posted by Magnefied View Post

[*]And the data was reviewed by a peer group not connected in any way to the study or the equipment manufacturers.


Again, all that is needed is a correctly done DBT. It looks like you don't know what DBT is. Maybe you should look up that.[/list]

Quote:
Originally Posted by Magnefied View Post

I'll wait till the short bus brings you home from school. But since you assert that this is how audio equipment reviews are performed, I have no doubt you'll deliver the goods. No doubts at all.


You are very childish and still have not proven a thing.

Sound and video is not magic, it is pure physics. Physics that can be magical
NIN74 is offline  
 
Thread Tools


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off