Summary Of Opinions Here
This thread has been interesting. I went back through the 700 posts and the opinions here seem to roughly divide into 3 groups as follows:
Those like SY, mikek, Steve Eddy, Christer, mefinnis, Brad Kizer, thylantyr, bratislav, kilentra, blmn, purplepeople and nw_avphile all seem to believe that blind listening is a valid indicator of real audible differences. Some of this group believes amps that measure reasonably well should sound so similar as to be hard to distinguish in a blind test (kept within their power limits of course). Some of these folks have been in blind tests, others appear to believe in the results of the many documented blind tests.
Those like tube_dude, ashok, sam9, moamps, vic, nania, mrfeedback and olsonsys seem to believe there are some audible differences between even amps that measure well, but they're caused by subtle things that can still be measured. They believe if two amps measure sufficiently close, they should be audibly indistinguishable (i.e. as in the Carver Challenge).
Those like Peter Daniel, fdegrove, pan, carlsosfm, Fred Dieckmann, traderbam, planet10 and analog_sa seem to believe there are things we cannot measure that make amplifiers sound different. They trust their ears above all else but at least some don't trust them in a blind test. They appear to dismiss all the evidence to the contrary and insist that what they hear is factual and real. Some of this group, like Fred Dieckmann, admit that psychological bias does alter listening perceptions. But nobody has made any other suggestions on how to eliminate it or why it doesn't invalidate what they hear.
The other contributors didn't really taken a stand on the issues present. They've contributed other things (like statistical measurement information, opinions on side topics, etc.). I'm sure I missed a few folks and if you think I put you in the wrong group, let us know.
This thread has been interesting. I went back through the 700 posts and the opinions here seem to roughly divide into 3 groups as follows:
Those like SY, mikek, Steve Eddy, Christer, mefinnis, Brad Kizer, thylantyr, bratislav, kilentra, blmn, purplepeople and nw_avphile all seem to believe that blind listening is a valid indicator of real audible differences. Some of this group believes amps that measure reasonably well should sound so similar as to be hard to distinguish in a blind test (kept within their power limits of course). Some of these folks have been in blind tests, others appear to believe in the results of the many documented blind tests.
Those like tube_dude, ashok, sam9, moamps, vic, nania, mrfeedback and olsonsys seem to believe there are some audible differences between even amps that measure well, but they're caused by subtle things that can still be measured. They believe if two amps measure sufficiently close, they should be audibly indistinguishable (i.e. as in the Carver Challenge).
Those like Peter Daniel, fdegrove, pan, carlsosfm, Fred Dieckmann, traderbam, planet10 and analog_sa seem to believe there are things we cannot measure that make amplifiers sound different. They trust their ears above all else but at least some don't trust them in a blind test. They appear to dismiss all the evidence to the contrary and insist that what they hear is factual and real. Some of this group, like Fred Dieckmann, admit that psychological bias does alter listening perceptions. But nobody has made any other suggestions on how to eliminate it or why it doesn't invalidate what they hear.
The other contributors didn't really taken a stand on the issues present. They've contributed other things (like statistical measurement information, opinions on side topics, etc.). I'm sure I missed a few folks and if you think I put you in the wrong group, let us know.
nw_avphile
Now you're going to get into trouble speaking for others. I was very specific in what I said and how I said it. Here it is again for those who missed it the first time:
Now you're going to get into trouble speaking for others. I was very specific in what I said and how I said it. Here it is again for those who missed it the first time:
I have listened to the same recording, on the same system and have heard it sound different. Perhaps the mains was noisier or I just didn't have the ears that day, who knows? The point is that if someone claims to hear things others can't, it doesn't necessarily make them a liar but the idea that they should be able to prove their claim in a DBCRT is absolutely plausable. The human body is capable of remarkable sensitivity and I'll wager that it can see, hear and feel things no machine can measure. So in this regard, you have definitely misinterpreted my opinion. There are people that can hear spatial and timing nuances even when their hearing is not as sensitive on a decibal per decibel chart comparison, just like a person with 20/20 vision may be incapable of seeing the aforementioned "magic eye" stereographic prints. Don't forget that the eyes and ears don't work independently from the brain. The whole system processes the analogue information presented to it so it would be difficult to find a machine that can credibly claim to be more sensitive.
Thanks for the correction. I edited my original post above to say the first group believes in blind testing and only some of them have stated that amps with reasonably good measurements tend to be hard to distinguish from each other.nania said:
Am I correct, from your post, that you believe in the objectivity of blind testing?
I think you greatly oversimplify my position. I believe that blind testing, when appropriate, can be a valid way to establish whether or not differences in a variable are perceptable. The words "can be" and "when appropriate" are extremely important. And the generality of my statement is meant to encompass sensory research in general, not the specific limitation of testing boxes of gain.
I would add that anyone having an opinion on controlled, valid sensory testing who has not participated in a well-structured test regime is analogous to a nun having an opinion on the relative merits of different sexual positions.
I would add that anyone having an opinion on controlled, valid sensory testing who has not participated in a well-structured test regime is analogous to a nun having an opinion on the relative merits of different sexual positions.
THE LABEL MAN.
Hi,
I agree with this entirely.
nw_avphile,
It seems labeling people and things seems to comfort you in your beliefs.
I object to being labelled and I notice you resort to insult as soon as you don't understand what is being said.
And I'm still having trouble figuring out what I said offended you this much.
More importantly I see little point in pursuing the subject of this thread anyway.
Must you really take it all in a personal way even when you're not addressed?
You claim to have asked me questions directly, I can't find any.
You claim to be offended by me pointing out a possible impedance mismatch in the soundcard levelmatching suggestion which you didn't suggest in the first place.
The cherry on the cake is your suggestion of cabledesigners putting out cables that would intentionally add distortions??
Come on, get real please.
Are you intentionally misreading peoples' replies?
This thread is just one big misleading mess and if you would have stated your intentions clearly from the start it would have been solved in just a few post.
Ciao,
Hi,
I agree with this entirely.
nw_avphile,
It seems labeling people and things seems to comfort you in your beliefs.
I object to being labelled and I notice you resort to insult as soon as you don't understand what is being said.
And I'm still having trouble figuring out what I said offended you this much.
More importantly I see little point in pursuing the subject of this thread anyway.
Must you really take it all in a personal way even when you're not addressed?
You claim to have asked me questions directly, I can't find any.
You claim to be offended by me pointing out a possible impedance mismatch in the soundcard levelmatching suggestion which you didn't suggest in the first place.
The cherry on the cake is your suggestion of cabledesigners putting out cables that would intentionally add distortions??
Come on, get real please.
Are you intentionally misreading peoples' replies?
This thread is just one big misleading mess and if you would have stated your intentions clearly from the start it would have been solved in just a few post.
Ciao,
comparasion test
I have 4 amps in my home, a marantz 2238 a pioneer tube PP amp a kenwood (new ) and my thornson gainclone.I can hear the difference between all of them. The kenwood is the worst and cant even begin to compare to the marantz ( the kenwood is rated at 140 wpc , the marantz 38wpc.next in worst is the tube PP amp,it as a great mid but has weak lows and a lack of high definition the marantz is next as it has the power and definition but on some material has a rather confused mid. The gainclone i love the best as it is a more of a combinaton in sound of the marantz and the tubes. Clear powerful bass very fast in all aspects, smooth defined mids and clear highs.I can tell all of them apart.
ron
I have 4 amps in my home, a marantz 2238 a pioneer tube PP amp a kenwood (new ) and my thornson gainclone.I can hear the difference between all of them. The kenwood is the worst and cant even begin to compare to the marantz ( the kenwood is rated at 140 wpc , the marantz 38wpc.next in worst is the tube PP amp,it as a great mid but has weak lows and a lack of high definition the marantz is next as it has the power and definition but on some material has a rather confused mid. The gainclone i love the best as it is a more of a combinaton in sound of the marantz and the tubes. Clear powerful bass very fast in all aspects, smooth defined mids and clear highs.I can tell all of them apart.ron
Re: Re: Re: Re: Re: We Got Both Kinds..
Hi Peter,
I didn't quite get the story about power cord and RCA/XLR. Are you saying that someone MEASURED differences or HEARD them ? There is a myriad of reasons that I can give here for measuring something that is not really there at all; thus we talk about controlled environment and peer reviews in a scientific method.
In any case, your dismissal of my knowledge about high quality audio reproduction is unfair, unfounded and rather childish.
How do you know about my experiences ?
What _is_ obvious is your ignorance about what signal to noise ratio really means and what is possible an what is not. If I told you that I could lift my car with one hand, would you believe it ?
Or that I can see 12th magnitide stars with my naked eye ?
Retrieval of signal below the noise using correlation techniques is possible and indeed it is a standard methodology in for examle astronomy, where people traditionally deal with incredibly tiny signals (from ~28th magnitide object we recive one photon every several MINUTES for example). But there is limit to it. You can go below threshold of noise (using best techniques and with known noise distribution) some 10-20 dB, maybe a bit better in special cases.
In a typical listening environment there will be at least 40dB of background noise. Your typical listening session will hover on average around 90dB. To HEAR residuals that are say 80dB down (and in case of cable or capacitor they will be WAY lower than this) your ear/brain "DSP" would have to be able to retrieve distortion that is 30+dB below noise floor ! And your room background noise, unlike that of a night sky, will not "autocorrelate out" very effectively. I won't even mention how much will speakers and room acoustics (resonances, diffraction and interference) add to the noise floor.
I guess that this may be overly academic, but there is no other way to put it. And I won't resort to foot stomping and name calling, sorry.
Bratislav
Pan said:
I wellcome "null tets" as these can proove what "golden ears" do hear is actually there.
One guy did a nulling test only swapping power cord to A CDP, results showed that there was a difference in what was coming out of the RCA/XLR terminalks depending on which power cord was there. HAAAA!
And by the way, the fact that you state that audible differences between various caps and cables are "deeply burried" by speaker and room distortion/noise only tells how little you know about high performande audio reproduction. The errors from cables (as an example) is smaller than the errors from the room, that is correct, but they are audible with good speakers and in a room with decent acoustics.
Writing this I assume (maybe wrong of me) that "burried" as you write it means "masked and non-audible"?
/Peter
Hi Peter,
I didn't quite get the story about power cord and RCA/XLR. Are you saying that someone MEASURED differences or HEARD them ? There is a myriad of reasons that I can give here for measuring something that is not really there at all; thus we talk about controlled environment and peer reviews in a scientific method.
In any case, your dismissal of my knowledge about high quality audio reproduction is unfair, unfounded and rather childish.
How do you know about my experiences ?
What _is_ obvious is your ignorance about what signal to noise ratio really means and what is possible an what is not. If I told you that I could lift my car with one hand, would you believe it ?
Or that I can see 12th magnitide stars with my naked eye ?
Retrieval of signal below the noise using correlation techniques is possible and indeed it is a standard methodology in for examle astronomy, where people traditionally deal with incredibly tiny signals (from ~28th magnitide object we recive one photon every several MINUTES for example). But there is limit to it. You can go below threshold of noise (using best techniques and with known noise distribution) some 10-20 dB, maybe a bit better in special cases.
In a typical listening environment there will be at least 40dB of background noise. Your typical listening session will hover on average around 90dB. To HEAR residuals that are say 80dB down (and in case of cable or capacitor they will be WAY lower than this) your ear/brain "DSP" would have to be able to retrieve distortion that is 30+dB below noise floor ! And your room background noise, unlike that of a night sky, will not "autocorrelate out" very effectively. I won't even mention how much will speakers and room acoustics (resonances, diffraction and interference) add to the noise floor.
I guess that this may be overly academic, but there is no other way to put it. And I won't resort to foot stomping and name calling, sorry.
Bratislav
I'm actually in all 3 categories...
To clarify:
I believe that blind (single, double, multiple) tests have merit.
I also believe that subjective test have merit because they are my main source of information when deciding my choices.
I believe that 2 different amps that measure well can have small differences that also can be measured.
I also believe there are differences that can be sensed (as opposed to heard, a more limited activity) and not measured.
This may seem contradictory, but in my view (below) it is not.
The scientific method has been fairly reliable for the past several hundred years so I expect its application in "blind" testing here to be both worthwhile and useful. However, as has been proven many times over the years, advances in technology and technique can re-write the state of the art. What one scientist "proves" one year against the "instinct" of another can be reversed with new information or another experiment. Sometimes, the first proof turns out to be reliable over the long term.
We are limited to measuring those things that our instruments are capable of, including not only the accuracy, repeatability and resolution of the instruments, but the units being measured. The analogy is that a ruler can only measure length in one direction at a time to an accuracy of the smallest marking on its face. Some things are just impossible. PI has a mathematical value but we'll never be able to measure it.
One paper (AES?) I've read suggests that musical instruments produce harmonics well beyond 20K. No, we can't "hear" this, but I will not yet agree that it cannot be sensed by humans. Sony claims this is the basis of SACD. Wonderful. If only they could record everything. All frequencies from 0Hz to UV rays and play it back. Maybe we can actually re-live the experience. Until then, we have our equipment.
Take bass notes. They are not only heard, but felt and usually in our stomachs. Why? Because the air pocket in or torso resonates with the frequency and amplitude of the wave.
Why do we sense someone behind us sometimes? Maybe we hear the person, maybe we feel their breath, maybe we smell the pheromones, maybe we see the slight changes in shadows nearby. I do not know, but I suggest that, similarly, there may be things we cannot hear that affect our experience of sound, both live and recorded.
Have I made this complicated enough? No? How about this...
Physicists have proven with experiements that energy travels as both waves and particles. It is both analog and digital. This does not only apply to atomic sized stuff, but sound. Take a sine-wave sound at f=1000Hz on a SACD into an amp and out of a 6-inch driver. Source is digital. 2.8Mhz / 1000Hz = 2800 bits peak to peak. Speaker moves back and forth 1000 times/sec creating an analog wave of air. Air particles have a pressure of 1 atm. There are 10^X particles of air between driver and my ear. Those particles transmit the sound wave through collision, just like billiard balls. Each collision is both digital and analog. Etc. etc. etc.
My original point (earliest post) is that I believe the argument is futile and that both subj and obj have merit and flaws. To use only one and not the other limits our ability to make the best possible decision at any time. I try to use both if I can. Both instruments and my body take measurements and I hope that they can agree. Sometimes, instinct is right. Sometimes not.
ensen.
To clarify:
I believe that blind (single, double, multiple) tests have merit.
I also believe that subjective test have merit because they are my main source of information when deciding my choices.
I believe that 2 different amps that measure well can have small differences that also can be measured.
I also believe there are differences that can be sensed (as opposed to heard, a more limited activity) and not measured.
This may seem contradictory, but in my view (below) it is not.
The scientific method has been fairly reliable for the past several hundred years so I expect its application in "blind" testing here to be both worthwhile and useful. However, as has been proven many times over the years, advances in technology and technique can re-write the state of the art. What one scientist "proves" one year against the "instinct" of another can be reversed with new information or another experiment. Sometimes, the first proof turns out to be reliable over the long term.
We are limited to measuring those things that our instruments are capable of, including not only the accuracy, repeatability and resolution of the instruments, but the units being measured. The analogy is that a ruler can only measure length in one direction at a time to an accuracy of the smallest marking on its face. Some things are just impossible. PI has a mathematical value but we'll never be able to measure it.
One paper (AES?) I've read suggests that musical instruments produce harmonics well beyond 20K. No, we can't "hear" this, but I will not yet agree that it cannot be sensed by humans. Sony claims this is the basis of SACD. Wonderful. If only they could record everything. All frequencies from 0Hz to UV rays and play it back. Maybe we can actually re-live the experience. Until then, we have our equipment.
Take bass notes. They are not only heard, but felt and usually in our stomachs. Why? Because the air pocket in or torso resonates with the frequency and amplitude of the wave.
Why do we sense someone behind us sometimes? Maybe we hear the person, maybe we feel their breath, maybe we smell the pheromones, maybe we see the slight changes in shadows nearby. I do not know, but I suggest that, similarly, there may be things we cannot hear that affect our experience of sound, both live and recorded.
Have I made this complicated enough? No? How about this...
Physicists have proven with experiements that energy travels as both waves and particles. It is both analog and digital. This does not only apply to atomic sized stuff, but sound. Take a sine-wave sound at f=1000Hz on a SACD into an amp and out of a 6-inch driver. Source is digital. 2.8Mhz / 1000Hz = 2800 bits peak to peak. Speaker moves back and forth 1000 times/sec creating an analog wave of air. Air particles have a pressure of 1 atm. There are 10^X particles of air between driver and my ear. Those particles transmit the sound wave through collision, just like billiard balls. Each collision is both digital and analog. Etc. etc. etc.
My original point (earliest post) is that I believe the argument is futile and that both subj and obj have merit and flaws. To use only one and not the other limits our ability to make the best possible decision at any time. I try to use both if I can. Both instruments and my body take measurements and I hope that they can agree. Sometimes, instinct is right. Sometimes not.
ensen.
What Noise Annoys An Oyster ?.
You misunderstand me too.
What I say is that IF you go to the trouble of modifying an on-test amplifier using the Carver null test to match that of a particular 'reference' amplifier, then by definition they ought to be difficult to distinguish.
But in real life this does not happen, and I do not expect any two 'factory' condition amplifiers to sound the same.
Measurements do not categorise the nature of the noise inherent in any amplifier, and the ear is the arbiter here.
By listening deeply into the noise, and by process of correlation, all amplifiers can be distinguished in my experience, PROVIDED that the room, source, line stages, all cables, speakers and mains power supply provide sufficient resoloution.
Different amplifiers react differently to what they are fed and according to how they are loaded, and this puts characteristics and patterns into the noise, and this can be distinguished reliably by the trained/learned ear.
IME, it is these patterns in the noise (thermal noise, distortions and errors) that categorise amplifiers, and are the distinction that can make an amplifer cold/warm, bright/dull, interesting/boring, alive/dead, pleasing/irritating etc, etc.
Until test instruments can categorise the natures of the noise under DYNAMIC conditions, the ears and the soul will be the judges, and probably always will be regardless.
IMO, because of all the variables involved (static, dynamic and audience/venue), DBL testing will never be a reliable indicator, and neither will standard test methods.
Standard test methods go a long way, but they do not go the last bit.
Eric.
As a polite question - How many DBL tests have been done with female only audience, and what were the outcomes ?.
You misunderstand me too.
What I say is that IF you go to the trouble of modifying an on-test amplifier using the Carver null test to match that of a particular 'reference' amplifier, then by definition they ought to be difficult to distinguish.
But in real life this does not happen, and I do not expect any two 'factory' condition amplifiers to sound the same.
Measurements do not categorise the nature of the noise inherent in any amplifier, and the ear is the arbiter here.
By listening deeply into the noise, and by process of correlation, all amplifiers can be distinguished in my experience, PROVIDED that the room, source, line stages, all cables, speakers and mains power supply provide sufficient resoloution.
Different amplifiers react differently to what they are fed and according to how they are loaded, and this puts characteristics and patterns into the noise, and this can be distinguished reliably by the trained/learned ear.
IME, it is these patterns in the noise (thermal noise, distortions and errors) that categorise amplifiers, and are the distinction that can make an amplifer cold/warm, bright/dull, interesting/boring, alive/dead, pleasing/irritating etc, etc.
Until test instruments can categorise the natures of the noise under DYNAMIC conditions, the ears and the soul will be the judges, and probably always will be regardless.
IMO, because of all the variables involved (static, dynamic and audience/venue), DBL testing will never be a reliable indicator, and neither will standard test methods.
Standard test methods go a long way, but they do not go the last bit.
Eric.
As a polite question - How many DBL tests have been done with female only audience, and what were the outcomes ?.
Re: Intentional Distortion ?.
Contrary to audiophile beliefs, it's not hard to get an AC signal under 20khz through a meter of cable without distorting it in any meaningful way. That might be why Nousaine in his article found a $3 interconnect cannot be distinguished from any of the much more expensive ones in the test once the psychological bias was removed.
Just like with those single-ended tube amps, a few manufactures have made cables that DO distort the audio signal in plainly measurable ways. There are cables with "magic boxes" in them. The manufacture usually claims the box is to assist with terminating the cable. I can try to find the reference, but in at least one case the "magic box" was found to gently roll off the highs to make a cable that would stand out as being "warm", "smooth", etc. (pick your adjective).
Likewise, even without a box, if you make a cable that has a ton of parasitic inductance and/or capacitance--through really odd construction--it can also alter the audio signal. So here we have a case of the cheap cable being far more accurate than the expensive one.
Try to dig up the Nousaine article I quoted. You can read about it for yourself and see what you think.
It's a lot like amplifiers. You can design an amp for minimum distortion--as SY would put it: "a box with gain". Or, you can design one that has rather high levels of distortion (like a single-ended tube amp) that will stand out from other amplifiers.mrfeedback said:
Contrary to audiophile beliefs, it's not hard to get an AC signal under 20khz through a meter of cable without distorting it in any meaningful way. That might be why Nousaine in his article found a $3 interconnect cannot be distinguished from any of the much more expensive ones in the test once the psychological bias was removed.
Just like with those single-ended tube amps, a few manufactures have made cables that DO distort the audio signal in plainly measurable ways. There are cables with "magic boxes" in them. The manufacture usually claims the box is to assist with terminating the cable. I can try to find the reference, but in at least one case the "magic box" was found to gently roll off the highs to make a cable that would stand out as being "warm", "smooth", etc. (pick your adjective).
Likewise, even without a box, if you make a cable that has a ton of parasitic inductance and/or capacitance--through really odd construction--it can also alter the audio signal. So here we have a case of the cheap cable being far more accurate than the expensive one.
Try to dig up the Nousaine article I quoted. You can read about it for yourself and see what you think.
Re: What Noise Annoys An Oyster ?.
That's consistent with what I posted. We're saying the same thing.mrfeedback said:
Well Bob turned the "modified Carver" into a production version that Stereophile re-tested but the results were inconclusive. But this IS a DIY forum, so if Bob Carver can make a cheap amp indistinguishable from a really expensive one in a few days, I suspect people here could as well.mrfeedback said:
Well as I've said before, the null test IS under dynamic conditions--real music driving real speakers. You can, for example, null both channels against each other, swap an expensive cap for a cheap one in one channel, and re-check the null. If it doesn't change, it's a safe bet the expensive cap is indistinguishable from the cheaper one.mrfeedback said:
The wife of the owner of Sunshine Audio couldn't pick out the cheap Yamaha from the $15,000 monoblocks. But, in general, I think most of the widely documented tests have used "trained ears" that supposedly have all the special listening skills that you and others have referenced. They tend to be attached to males.mrfeedback said:
- Status
- Not open for further replies.