A thought experiment

[Guest Eggcup The Daft]

In the More Snakeoil thread, @ put forward the proposition that some things can't be measured. This is a regular subject here.

It got me wondering. What of positions in the soundstage, inky black silences and so on?

 

Well, either they are there, or they are not. But, of course, everything we listen to has already been measured, because a recording is only a continuous measurement. And of course, we can repeat that measurement for the output of a system. All that is being done with conventional loudspeaker measurements, is that a controlled sound is played or generated and then recorded/re-recorded and the tones compared or some part of the result evaluated.

 

My conclusion is that any aspect of the sound, the movement of air, can be measured, and the input to a stereo system has already been.

 

So, we can't directly measure where myrantz hears a particular sound on playback. What we can do though, in theory, is:

play back a recording on systems where a number of audiophiles have identified a particular aspect to the sound or not: a change in location of a vocalist, an "inky black silence", an "upper midrange colouration" or such.

 

We could then use a high quality recording system to record the output of the system at the listening position for each system, and compare with the input. If there is a difference in sound that leads to that percieved change, it should show up in the differences between each recording and the playback. If there is no sound difference, we can assume that the change takes place in the listener's head.

 

Has anyone tried this experiment in the past, and if so, do we know what the results were?

 

Would it work or is there some reason why it shouldn't?

 

Would these differences be obscured by other differences, or be too fine to be picked up?

[Guest myrantz]

Well, either they are there, or they are not.

Schrödinger's cat?

 

My point is, and always will be, this

 

everything we listen to has already been measured...

because a recording is only a continuous measurement.

is absolutely wrong.. Unless you and the usual suspects can prove otherwise of coz (with real data, not just because "I said so")... 

 

JA has posted several blogs in his time, one of which just so happen to de-bunk your post: Linky

 

A comment by "cgh" in an online reprint of a Stereophile review caught my attention: "The [1990s] were probably the last real decade that we could reasonably bend the truth. Everything since is verifiable electronically."

 

Everything? After a quarter century of measuring the performance of audio components for this magazine, I'm not so sure that we have a firm handle on what makes audio products sound different from one another.

 

Of coz collective pack wisdom could prove him (and me wrong).. This thread may break new frontiers in sampling! A continuous sampling method that captures everything. Imagine the applications  .

I'd leave it to the others to dominate this discussion, but let me say this - I'd be amazed if folks can come up with a counter with the same level of quality and thought as JA to support the "everything" case.

[Super Mustud]

I decided a long time ago that the biggest problem is that we can't measure why sometimes we like something while at other times the same thing may not particularly attract.

 

Yep, measure the measurable for sure. It makes us feel good about being objective. A bonus, no doubt.

 

But how come sometimes I like Dianna Krall and other times I don't?

 

Measurements aren't that important to me these days. Not until I can measure Dianna Krall attractiveness.

[Satanica]

My point is, and always will be, this

 

Eggcup The Daft, on 24 Jun 2015 - 11:05 AM, said:

everything we listen to has already been measured...

because a recording is only a continuous measurement.

 

is absolutely wrong.. Unless you and the usual suspects can prove otherwise of coz (with real data, not just because "I said so")...

 

Thanks for brightening up my day with that little beauty!

Now I know there is more in a recording than what is in a recording and don't worry about the "real data", it's because you say so!

[davewantsmoore]

Measurements aren't that important to me these days. Not until I can measure Dianna Krall attractiveness.

 

As a musician, recording artist/engineer, and loudspeaker designer  (not gods-gift to any of those, mind you - although I do aspire to the later ***) .....   I could not begin to describe IMHO the important of recordings being played back on systems which have as few errors as possible.    When I make a change to my guitar, or to the microphone location ..... I want to know that people are able to hear that change in the same way I do.    I want people to have systems which reproduce the recording to a standard which ensures this.

 

When I make a change to a speaker (or any other aspect of a playback system) .... I want that change to be about removing errors in the reproduction.     What I "like" isn't a consideration.....    sometimes, on an improved playback system, I might be caused to go back and revisit (artistic) choices about the sound - either the instrument/musician, or the recording process 

 

..... and that does identify an interesting "circle of confusion", where by, if someone auditions their "sound" on a playback system with obvious flaws....   Then do they compensate for these flaws in the section of their sound?   If your studio monitors have lacking bass, do you they end up compensating for that by dialling in more bass in the studio?    It's a real problem.

 

OTOH, that isn't to say that people can't have enjoyment if they playback on system which has flaws.    People's ability to enjoy music of any 'quality' shouldn't be underestimated in my experience.

 

 

*** one day I may become professional, and relieve you all from my constant daytime dross  (I have a boring day-job)

[Super Mustud]

As a musician, recording artist/engineer, and loudspeaker designer  (not gods-gift to any of those, mind you - although I do aspire to the later ***) .....   I could not begin to describe IMHO the important of recordings being played back on systems which have as few errors as possible.    When I make a change to my guitar, or to the microphone location ..... I want to know that people are able to hear that change in the same way I do.    I want people to have systems which reproduce the recording to a standard which ensures this.

 

When I make a change to a speaker (or any other aspect of a playback system) .... I want that change to be about removing errors in the reproduction.     What I "like" isn't a consideration.....    sometimes, on an improved playback system, I might be caused to go back and revisit (artistic) choices about the sound - either the instrument/musician, or the recording process 

 

..... and that does identify an interesting "circle of confusion", where by, if someone auditions their "sound" on a playback system with obvious flaws....   Then do they compensate for these flaws in the section of their sound?   If your studio monitors have lacking bass, do you they end up compensating for that by dialling in more bass in the studio?    It's a real problem.

 

OTOH, that isn't to say that people can't have enjoyment if they playback on system which has flaws.    People's ability to enjoy music of any 'quality' shouldn't be underestimated in my experience.

 

 

*** one day I may become professional, and relieve you all from my constant daytime dross  (I have a boring day-job)

 

 

Good points. You remind me that I was rather simplistic in my earlier post. My comments are based on there being a reasonably good system already available, and with effort then being expended to obtain further incremental improvements.

 

I guess that after some years of doing this I realised I was spending more time listening for "flaws" than to the music, and when finding, or thinking I found something, declaring AHA! with some triumph. There comes a time when that becomes a distracting obsession. For me, anyway.

 

When I could have been listening to Dianna Krall all the time.

[Guest Eggcup The Daft]

@

I read your link. It doesn't debunk my claim at all. It shows that the standard measurement of two loudspeakers shows small differences, and describes subjective audio differences. It actually doesn't go very far in relating the measurement and the sound, either, other than some comments about bass response.

 

You can deny that the recording is a measurement as much as you like. It remains a fact.

 

@@davewantsmoore

I'm entirely aware that my thought experiment relates only to the entire system and its environment. It doesn't prove much about any one component.

And surely you are balancing errors rather than just reducing them, given the limitations and problems of speaker design?

[Guest myrantz]

@

I read your link. It doesn't debunk my claim at all. It shows that the standard measurement of two loudspeakers shows small differences, and describes subjective audio differences. It actually doesn't go very far in relating the measurement and the sound, either, other than some comments about bass response.

Reading cgh's comments, is your claim in the first post any different to the claim made by CGH? If yes, how?

Earlier on you're trying to tell me about cause and effect. I'm trying to make you understand what you're doing is only establishing correlation, not causation. If you understand JA's breakdown over several blog posts (he and his team made a couple around that time), you would have know your mistake in using the words everything.

 

You can deny that the recording is a measurement as much as you like. It remains a fact.

This is at least the second time I have to assert I did not said that in bold... Please give me a link where I actually said that, as I suspect a clarification to my words is needed.. Bombarded from all sides yesterday, some of the things I said may well be wrong. And so, I will correct it.

If you still insist on being right, I have more audiophile terms that you can't measure.. e.g. the quality of separation between the main singer voice vs the backup vocals, the quality of the timbre, the ease of the music to send my heart pounding, clarity of vocals/instruments/microdetails, scale of a voice (is it 1:1 scale like a normal person standing in front of me, or is it projected like the mouth is 3m wide), the ease of me to get into sadness with certain songs, the ease of me getting happy with others, hairs raising on the back of my neck, shivers up my spine, the solidness of a playing instrument or voice, the level of disappearing act a system can do, the time I can spend listening to a system without getting fatigued.

 

Go ahead trying to pin those down into measurements. See how firm a grip you can get. 

[TP1]

Measurements can be useful  up to a point  but everything that contributes to sound quality cannot be measured at the moment.

 

To quote Nelson Pass:  

 

"It is nevertheless possible to have a product that measures well but doesn't sound so good. It is still a mystery as to how this could be, but there it is. My experience is that there is a reasonable correlation between sound and measurement for simple Class A circuitry with minimal or no feedback. This relationship seems to disappear when the circuit becomes complex or has a lot of non-linearity corrected by feedback."

 

Having said that, I think too many manufacturers BS about their specs and  with the advent of  online reviews , there are not enough reviewers measuring equipment to keep the B's honest

 

 

There is more interesting reading in this  article:

 

http://www.herronaudio.com/images/Measurements.pdf

[ZEN MISTER]

What is it about boys and measuring things.????

ZM.

[Guest myrantz]

"It is nevertheless possible to have a product that measures well but doesn't sound so good. It is still a mystery as to how this could be, but there it is. My experience is that there is a reasonable correlation between sound and measurement for simple Class A circuitry with minimal or no feedback. This relationship seems to disappear when the circuit becomes complex or has a lot of non-linearity corrected by feedback."

Hey, Nelson Pass says correlation! Good point on Class A, in a sense it's less chaotic (a point I have raised before to illustrate the difficulty and the usual suspects just seem to glaze over)..  

In theory at least (as I have 0 real experience with Class A), it actually does makes sense to be worth a go. Will be interesting to see how the usual suspects (+ the OP)  go with this experiment (not just thought, actual work).

 

Be interesting to see if they can solve the impasse between so called engineer (concentrating more on the physical aspects), and so called audiophile (concentrating more on cognitive side).

 

What is it about boys and measuring things.????

 

To be fair listeners measures also, we just have difference references and methods of measuring.. Myself I definitely measure when I listen, although how I measure couldn't really be put down into words..

 

As to why boys like to measure things, I blame it on the women, and er, books with illustrations. Boys just have this urge to make sense and order out of chaos.

[TP1]

While I agree specs don't tell the full story, they still provide some guidance as to the capability of the unit.

There is a surprising (to me) amount of DIY gear being used by people for which it would be rare to even hazard a guess at the specs, let alone know what they are. Then there are manufacturers that don't publish specs or when they do they are not useful.

Listening tests are so system dependent they can't be relied upon to perform well in all systems

[Satanica]

While I agree specs don't tell the full story, they still provide some guidance as to the capability of the unit.

There is a surprising (to me) amount of DIY gear being used by people for which it would be rare to even hazard a guess at the specs, let alone know what they are. Then there are manufacturers that don't publish specs or when they do they are not useful.

Listening tests are so system dependent they can't be relied upon to perform well in all systems

 

There's two things to consider when reading specs.

1) Am I reading the complete specs of the unit? The answer is almost definitely NO.

2) Even if I am reading the complete specs of a speaker or sub-woofer, it says nothing about how it will work in my listening room.

[Guest Eggcup The Daft]

@

What I mean by you denying that a recording is a measurement, is that you deny that things in the recording cannot be measured, i.e. the positional representation of a singer's voice in the stereo image. My point is that it has been measured for every two channel recording ever made.

The recording is a measurement, therefore the position has been measured, or you could never hear the location in the recording. If you deny that it can be measured, then you by that fact deny that the recording is a continuous measurement.

 

I should allow for the deleterious effects (distortions) of the playback system and environment, of course. But that does not mean that we can measure those deleterious effects, if we wish, by comparing a recording of what you hear in your listening position to what is on the recording you are playing back. But the position in the mix has been measured at the time of recording (or possibly precisely placed by the engineer in mixing down to stereo); with the correct analysis it can be known from the input signal.

 

I will qualify this by saying that while work has been done on stereo placement, I don't know that anyone has ever attempted to place a particular sound from a stereo mix, other than by playing back and listening.

 

I am not suggesting that we can measure your precise emotional reaction by measuring the sound in the listening position, yet. That would require "big data". We would need to measure your reaction to lots of music, be able to tell what type of music is in a recording by examining the waveform, be able to recognise which musical techniques you react to in particular ways, and select recordings with similar attributes. We'd also need to analyse lyrical forms to which you respond as well. You're probably not as atypical or individual as you think you might be - in fact I think you posted some links about that in one of the recent threads here. Evidence with big data suggests that we can actually get closer by finding what people similar to you react to as well. We could also measure your individual responses using things like fMRI, though that technique is far from perfect, as seen by the dead fish measurement.

 

It's all measurable. Technologically, it can probably all be done now, apart from gaining the knowledge to interpret the measurements. I'm not sure I WANT to measure people's emotional reactions so well as to measure the personal responses you refer to. That should be your own property, in my opinion.

 

The one that had me stuck for a while was the "inky black silence", because silence is silence. However, we are probably being prepared for that "special" silence in the preceding fading of sound, perhaps less low-level distortion.

 

Some degree of correlation would have to be used. However, I expect that once the particular attributes are found that lead to a group of audiophiles concluding that an effect is present in a blind test, it will lead to the discovery of causation.

 

Some effects may be taking place in the mind of the listener or a response to non-audio stimuli. In those cases something other than the sound has to be measured.

 

I hope that formulation of words is more acceptable. If you still disagree, fair enough.

 

 

Measurements can be useful  up to a point  but everything that contributes to sound quality cannot be measured at the moment.

 

To quote Nelson Pass:  

 

"It is nevertheless possible to have a product that measures well but doesn't sound so good. It is still a mystery as to how this could be, but there it is. My experience is that there is a reasonable correlation between sound and measurement for simple Class A circuitry with minimal or no feedback. This relationship seems to disappear when the circuit becomes complex or has a lot of non-linearity corrected by feedback."

 

 

 

It depends on what you measure and how. My thought experiment is about measuring a playback system in its environment at the time of measurement, not any one component.

 

In the case of amplification, my wild guess would be that the form the distortion takes changes,and  that perhaps feedback is not as good at fixing non-linearity as conventionally thought. Others have certainly said similar things to this quote, or have stated a preference for systems with no feedback /no global feedback even with higher distortion, or at least higher THD.

 

 

There's two things to consider when reading specs.

1) Am I reading the complete specs of the unit? The answer is almost definitely NO.

2) Even if I am reading the complete specs of a speaker or sub-woofer, it says nothing about how it will work in my listening room.

 

In theory, we could measure the speaker so well in terms of its output that we could model how it works in any particular system and room. The acoustic models used for designing auditoriums can model what can be heard in any seat for a particular singer on the stage. Or at least, that is what is claimed.

[Guest myrantz]

@

What I mean by you denying that a recording is a measurement, is that you deny that things in the recording cannot be measured

Again stop putting words in my mouth. If you want to have an argument with me at least understand the basis of the points I make first before making wild claims about what I said (or not)...

My assertion is simple - The measurements you can do now cannot be mapped into the things audiophiles are looking for.

 

Please show me how you can turn this measurement: "any aspect of the sound, the movement of air, can be measured" into the quality of timbre associated with a Cello...

It's all measurable you say? Prove it.

[Satanica]

In theory, we could measure the speaker so well in terms of its output that we could model how it works in any particular system and room. The acoustic models used for designing auditoriums can model what can be heard in any seat for a particular singer on the stage. Or at least, that is what is claimed.

 

Sure, but in practice hardly a home enthusiast does this, nothing like it.

 

My comments were made in an attempt to squash comments from audiophiles that they can hear more than specs.

Because when if ever do they have close to all of them from the manufacturers and how many of them are even capable of interpreting them properly if they did?

Not very often and not many I presume.

 

And then there's the listening room.

 

But I guess it's still amusing to read such strong comments and opinions from our fellow audiophiles in a world of so much practical uncertainty.

[Guest Eggcup The Daft]

Is anyone prepared to take a shot at answering the three questions at the end of my original post?

Edited June 25, 2015 by Eggcup The Daft

[Guest myrantz]

Now that I got that out of my system.. Let me post from a more irreconcilable (hopefully) post.

 

I hope that formulation of words is more acceptable. If you still disagree, fair enough.

I'm afraid you are not aware just how limited measurements can be. It's good that you did not fail into the "I said so" trap and take my word for it.

That's why I encourage you to try what you suggest and realise for yourself just how limited in scope measurements can be. 

Do you reckon if we measure the earth crusts movements we can predict when and where the next earthquake will hit? Some seismologists were being charged in court for not doing exactly that (hopefully it's a April fool's joke). 

Measurements is basically a translation of the real world into a smaller subset of "finite" variables. It's describing a complex problem into similar (more managable terms).. It's all measurable? I have to disagree,  You simply cannot say all, everything, all aspects, etc.

 

If what you said is true, we would have known exactly where MH-370 is the moment it's lost.

 

As for your hypothesis, we actually discussed this with DWM before, the closest I guess would be this: https://en.wikipedia.org/wiki/Difference_in_differences

But the kicker is, people are after different things.. There is no clear and definite goal, reference is essentially undefined. Some will claim perfect reproduction (or as little error as possible) is actually something that's possible.. It's up to you to decide on your own when you start doing your own experimentation how true that is (as in, how to actually prove it).

[Satanica]

OK, for what it's worth I'll have a crack.

 

Has anyone tried this experiment in the past, and if so, do we know what the results were?

 

Would it work or is there some reason why it shouldn't?

 

Would these differences be obscured by other differences, or be too fine to be picked up?

 

I'm not sure if "anyone" has, I know I haven't.

 

I don't think you should measure at the listening position. I think you should measure in an Anechoic Chamber, probably at 1m, each channel separately and compared separately.

 

You're assuming there will be a difference. If there was then that would warrant further investigation.

[davewantsmoore]

I guess that after some years of doing this I realised I was spending more time listening for "flaws" than to the music

 

It's interesting that some people (I don't mean you) seem to think that anyone who would make a measurement of a speaker, is the type of person who would hear (or listen for) only flaws.... obsess (constantly) over their system.... or be the sort of person who would accept some type of "soulless" sound (because the measurements say so?!) .....  and aren't in it for any kind of musical enjoyment.

 

I don't "listen for flaws".     It isn't enjoyable... and it isn't reliable.

• I hear flaws which aren't there
• I hear flaws or a greater or lesser magnitude than are there
• I hear flaws but my subjective interpretation of the flaw leads me to think the cause is something that it is not

 

 

... and the kicker......   I measure the system (not with my ears) and I find flaws which were not obvious to me (by ear).....   when I reduce these flaws, often the difference is very audible (positively).

 

 

 

To trade something (like "listening") for something which is more objective (like methodical measurement and analysis) .....  takes one key thing.   The belief that there are measurements and analysis which correlate with positive subjective preferences.

 

This is something that I would not expect anyone to "just believe".

 

There is a lot of evidence which has been presented to support it ..... BUT, it is not necessarily something I would expect anyone to believe unless they had "seen it for themselves".

 

 

.... and it takes a lot of effort to "see for yourself".

[Guest Eggcup The Daft]

Now that I got that out of my system.. Let me post from a more irreconcilable (hopefully) post.

 

I'm afraid you are not aware just how limited measurements can be. It's good that you did not fail into the "I said so" trap and take my word for it.

That's why I encourage you to try what you suggest and realise for yourself just how limited in scope measurements can be. 

Do you reckon if we measure the earth crusts movements we can predict when and where the next earthquake will hit? Some seismologists were being charged in court for not doing exactly that (hopefully it's a April fool's joke). 

Measurements is basically a translation of the real world into a smaller subset of "finite" variables. It's describing a complex problem into similar (more managable terms).. It's all measurable? I have to disagree,  You simply cannot say all, everything, all aspects, etc.

 

If what you said is true, we would have known exactly where MH-370 is the moment it's lost.

 

As for your hypothesis, we actually discussed this with DWM before, the closest I guess would be this: https://en.wikipedia.org/wiki/Difference_in_differences

But the kicker is, people are after different things.. There is no clear and definite goal, reference is essentially undefined. Some will claim perfect reproduction (or as little error as possible) is actually something that's possible.. It's up to you to decide on your own when you start doing your own experimentation how true that is (as in, how to actually prove it).

Thanks for sticking with this. If our positions become more irreconsilable, it's at least helping me understand what my position is - I think I've shifted a little already.

 

Regarding MH370, as I understand it, what they got were a set of broadcasts. They were using a triangulation for those broadcasts that gave approximate positions at specific times. Had they had a continuous broadcast, they would know where the signal stopped, ie. where the plane crashed. They haven't got that. For all we know, a pilot may have left the computer on autopilot until the last of those timed signals, then changed course for the last 55 minutes before the plane went down. Or the second possible position is really the right one and the plane landed at a secret airstrip in Kazakhstan. In other words, the information is incomplete. Just like a mono sweep measurement of one of your loudspeakers won't tell you if your hair will stand on end listening to a particular song.

 

Clearly, as little error as possible (for a given state of technology) can be done, so I don't quite understand your last paragraph. People here are chasing that goal. It may be the wrong goal. But reproducing as little error as possible is fidelity to the recording being played back, and if we hear real differences in the sound, in the movement of air, they are either all (or all bar one!) being played back with errors from the original signal.

If we agree on that much, then we CAN compare the original with a recording taken at the listening position and determine the differences. All of them. If there is no difference, then the sound is no different and something else is influencing our preference for one system over another. If there is a difference, then it at least proves that there is something for audiophiles to hear.

Of course, it may be very, very difficult to show what aspects of those differences are in fact audible.

 

Edit - I've misinterpreted something @@Newman posted, removed that bit

Edited June 25, 2015 by Eggcup The Daft

[Guest Eggcup The Daft]

OK, for what it's worth I'll have a crack.

 

 

I'm not sure if "anyone" has, I know I haven't.

 

I don't think you should measure at the listening position. I think you should measure in an Anechoic Chamber, probably at 1m, each channel separately and compared separately.

 

You're assuming there will be a difference. If there was then that would warrant further investigation.

It depends on what you are trying to show. Part of the audiophile listening problem is that the system is in a non-anechoic environment. If I was actually equipped to do this sort of work, I would include the room initially. I'm trying to compare what we "hear" to the acutal sound present, not different components or systems.

 

If there is no difference, or no audible difference, that is a significant finding in itself, of course. It would mean that the speakers and the room, with all those variations, are no more important than sticking a picture of yourself in the freezer (if you don't know that reference, google Peter Belt and read on...)

[Guest myrantz]

Thanks for sticking with this. If our positions become more irreconsilable, it's at least helping me understand what my position is - I think I've shifted a little already.

Whoops!!! Where di that ir come from? I mean more reconcilable...

Regarding MH370, as I understand it, what they got were a set of broadcasts. They were using a triangulation for those broadcasts that gave approximate positions at specific times. Had they had a continuous broadcast, they would know where the signal stopped, ie. where the plane crashed. They haven't got that. For all we know, a pilot may have left the computer on autopilot until the last of those timed signals, then changed course for the last 55 minutes before the plane went down. Or the second possible position is really the right one and the plane landed at a secret airstrip in Kazakhstan. In other words, the information is incomplete. Just like a mono sweep measurement of one of your loudspeakers won't tell you if your hair will stand on end listening to a particular song.

The information is incomplete, but the analysis supports the notion that the plane crashed in the southern hemisphere, not the north. But that's all we have, based on the measurements, we don't know _exactly_ where.

Think of it in Tasso's words that it gives you some guidance. But there are some things/information that you can't gleam from it. To say everything, all aspects, or continuous measurements is incorrect.

Clearly, as little error as possible (for a given state of technology) can be done, so I don't quite understand your last paragraph. People here are chasing that goal. It may be the wrong goal, as Newman has stated. But reproducing as little error as possible is fidelity to the recording being played back, and if we hear differences, they are either all (or all bar one!) being played back with errors from the original signal.

If we agree on that much, then we CAN compare the original with a recording taken at the listening position and determine the differences. All of them. If there is no difference, then the sound is no different and something else is influencing our preference for one system over another. If there is a difference, then it at least proves that there is something for audiophiles to hear.

Of course, it may be very, very difficult to show what aspects of those differences are in fact audible.

So here you are trying to recreate a system that produces as little error as you can.. In order to do that, you need to design and invent a measuring system that is perfect in measurements. Accuracy vs precision, you'd be forever going around in circles. If your measurements can never be perfect, how can you claim you have a reproduction system that has little errors?

It's all measurable is akin to cgh's comments everything can be electronically verified... It is just not that simple. Even for the sake of argument that it's all measurable, what do the measurements mean? What do the differences in measurements mean? How to do the sort of "verification" in terms audiophiles are actually interested in?

[Satanica]

It depends on what you are trying to show. Part of the audiophile listening problem is that the system is in a non-anechoic environment. If I was actually equipped to do this sort of work, I would include the room initially. I'm trying to compare what we "hear" to the acutal sound present, not different components or systems.

 

If there is no difference, or no audible difference, that is a significant finding in itself, of course. It would mean that the speakers and the room, with all those variations, are no more important than sticking a picture of yourself in the freezer (if you don't know that reference, google Peter Belt and read on...)

 

I don't think I'm sticking my neck out by saying you're going to always find a difference.

Question(s): How many microphones would you use? Would you record in mono or stereo spaced mikes?

[davewantsmoore]

Has anyone tried this experiment in the past, and if so, do we know what the results were?

Would it work or is there some reason why it shouldn't?

Would these differences be obscured by other differences, or be too fine to be picked up?

 

This is an interesting type of test which has been used before by a few well known people.

 

 

One interesting variant of the test is one which I heard of through Danley, and which I've myself tried a number of times..... and the results are interesting.    I've never documented the results rigorously (and compared many setups to each other with any great precision) ..... but I have used it to compare a few speakers  (very good ones, VS very bad ones, for example)  .... to verify that it very much does point to what's best.

 

 

The method is:

 

Play some program material on the speaker

Record the output

Play that recording through the speaker

Record the output

Play the second recording through the speaker

Record the output

 

... keep repeating this process until the sound degrades a lot.

 

 

I've done this both outside, and in a room.

 

On less than good speakers, it may only take a recording of a recording, before the sound has turned to garbage......    on better speakers, a recording, of a recording, of a recording (etc.) can sound remarkably good.