Can you hear 16/44.1 from 24/96?

[Audiop]

Originally Posted by MichaelPatrick

Peter, I wish you had read my words before telling me how an improperly set up supertweeter works, because then I'd know that you at least read what I said.

I did read your words but maybe I did not pay enough attention to that part.

That said, what is a proper set up of a supertweeter?

IMO there is no and either way you must do some kind of comparison to get anywhere.

Studies in sonic perception and psycho-acoustics are still going on because measurement systems have not fully mapped out how humans hear.

Measurements are made to collect data. The mapping is done by people. No repeatable controlled results has been brought to the table to support the idea of ultra sound being audible with music at typical levels.


/Peter

[Ozpeter]

The results of the Boston study indicate that the vast majority of listeners evidently have sub-wooden-ear hearing:-

"The test results for the detectability of the 16/44.1 loop
on SACD/DVD-A playback were the same as chance:
49.82%. There were 554 trials and 276 correct answers.
The sole exceptions were for the condition of no signal
and high system gain, when the difference in noise floors
of the two technologies, old and new, was readily audible.
As the tests progressed, we repeatedly sorted the data
for correlations with age, sex, upper frequency hearing
limit, or experience. No such correlations have emerged.
Specifically, on music at normal levels as defined here,
audiophiles and/or working recording-studio engineers got
246 correct answers in 467 trials, for 52.7% correct. Females
got 18 in 48, for 37.5% correct. Those subjects able
to hear tones above 15 kHz got 116 in 256 trials, for 45.3%
correct; listeners aged 14–25 years old (who were, as it
turned out, the same group), also got 116 correct in 256
trials, 45.3%. The “best” listener score, achieved one
single time, was 8 for 10, still short of the desired 95%
confidence level. There were two 7/10 results. All other
trial totals were worse than 70% correct.
Furthermore, none of the more elaborate and expensive
playback systems (for which the subjects were all dedicated
amateur audiophiles, active students in a professional
recording program, and/or experienced working
professionals) revealed detectable differences on music,
again at levels as defined previously."

However, the authors do paradoxically concede that

"Though our tests failed to substantiate the claimed advantages
of high-resolution encoding for two-channel audio,
one trend became obvious very quickly and held up
throughout our testing: virtually all of the SACD and
DVD-A recordings sounded better than most CDs—
sometimes much better. Had we not “degraded” the sound
to CD quality and blind-tested for audible differences, we
would have been tempted to ascribe this sonic superiority
to the recording processes used to make them."

They suggest that the difference is not down to the use of 24/96 recording standards per se.

"Plausible reasons for the remarkable sound quality of
these recordings emerged in discussions with some of the
engineers currently working on such projects. This portion
of the business is a niche market in which the end users are
preselected, both for their aural acuity and for their willingness
to buy expensive equipment, set it up correctly,
and listen carefully in a low-noise environment.
Partly because these recordings have not captured a
large portion of the consumer market for music, engineers
and producers are being given the freedom to produce
recordings that sound as good as they can make them,
without having to compress or equalize the signal to suit
lesser systems and casual listening conditions. These recordings
seem to have been made with great care and
manifest affection, by engineers trying to please themselves
and their peers. They sound like it, label after label.
High-resolution audio discs do not have the overwhelming
majority of the program material crammed into the top 20
(or even 10) dB of the available dynamic range, as so
many CDs today do.

Our test results indicate that all of these recordings
could be released on conventional CDs with no audible
difference. They would not, however, find such a reliable
conduit to the homes of those with the systems and listening
habits to appreciate them. The secret, for two-channel
recordings at least, seems to lie not in the high-bit recording
but in the high-bit market."

That seems to me to have the ring of truth about it - though of course it's an unproven assertion!

[Dale]

the boston study methodology is in question.
seems they had a pa system running in the same area simultaneously. not a controlled environment for hearing the subtle differences! this was just one of the major flaws.

[Ozpeter]

seems they had a pa system running in the same area simultaneously. not a controlled environment for hearing the subtle differences!

I've never seen that assertion before. And the tests were conducted in a number of environments using several systems. However, anyone wanting to spend about an hour on the matter might care to grind through Audio Asylum Thread Printer where the subject was recently resurrected (and, as always, not put to rest).

The point the Boston study made is I think a compliment to recording professionals. The whole 24/96 thing gives rise to the danger that people will hear recordings made that way and say "Great recording - you must be using very good equipment". Like the person who sees a superb photo and turns to the photographer and says "what a great camera you have". It's the engineer, not the equipment. A carefully and skillfully made recording will sound better than an incompetently made recording (almost) regardless of what was used. The Boston study underlines that. Forget the numbers and listen to the sound. Obsessing about the numbers does the profession no favours, IMHO.

[audio ergo sum]

Originally Posted by Dale

the boston study methodology is in question.
seems they had a pa system running in the same area simultaneously. not a controlled environment for hearing the subtle differences! this was just one of the major flaws.

"seems they had..." Do you have any evidence fort that claim or is it just personal opinion? Please elaborate.
In the Boston "area" ?

Another much earlier study in Germany has come to the very same results many years ago, and that one was definitely not next to a pa system in the same area.

96 kHz contra 48 kHz - Hörtests am Erich-Thienhaus-Institut in Detmold
96 kHz contra 48 kHz - Hörtests am Erich-Thienhaus-Institut in Detmold

[Teddy Ray]

temporal resolution of hearing limited by bandwidth restriction...

Information for prospective students

[aracu]

Originally Posted by ISedlacek

So what prevents you to do it ?? What are your own observations after doing a proper comparison ? It is so simple. Even one practical experiment and comparison has thousand times higher value than hundreds pages of these theoretical arguments ... (where no one ever cares to listen to a real thing) Just remotely telling the others they are doing completely wrong ... It should be done like this, it should not be done like this ... Just do it first and report ... Only then something valuable will be added to the topic.

I see your point. However, the reason that I think testing should be done
blind is not based purely on a theoretical idea. I try not to take theories too
seriously.

I made some simple, limited tests using Sound Devices pres and converters, of nylon string guitar recordings at various sample rates. Knowing the sample rates of each file, I thought I could easily percieve the difference.

With the same files in a blind test I scored about 50% correct.

I was surprised to find out how much of my perception was influenced by
belief and imagination.

[boojum]

Originally Posted by aracu

I see your point. However, the reason that I think testing should be done
blind is not based purely on a theoretical idea. I try not to take theories too
seriously.

I made some simple, limited tests using Sound Devices pres and converters, of nylon string guitar recordings at various sample rates. Knowing the sample rates of each file, I thought I could easily percieve the difference.

With the same files in a blind test I scored about 50% correct.

I was surprised to find out how much of my perception was influenced by
belief and imagination.

This is why I also believe that double-blind is the safest as it does eliminate the subtle psychological cues that one is sure are not there. In a double-blind test they are definitely not there. ~50% is what guessing will get you. However double-blind is also a big red flag for many. I am not sure why responsible scientific testing is rejected. It seems to be most disbelieved in the audio realm if my experience is correct.

At the end-up I think we will just have to agree to disagree.

[Audiop]

Originally Posted by Teddy Ray

temporal resolution of hearing limited by bandwidth restriction...

Information for prospective students

Those tests (or at least "that test", the first I think) are very questionable as have been adressed many times allready in various threads here at GS.


/Peter

[MichaelPatrick]

Originally Posted by Audiop

Those tests (or at least "that test", the first I think) are very questionable as have been adressed many times allready in various threads here at GS.

Peter, the physics professor at USC, Kunchur, cites 4 publications, 5 presentations at scientific association meetings and at 22 scientific and academic colloquia. Please, if you will, cite some of the GS threads where "those tests...have been addressed." Were they conclusive in any way? What persuades you that Kunchur is wrong?

Originally Posted by Audiop

No repeatable controlled results has been brought to the table to support the idea of ultra sound being audible with music at typical levels.

I think you are trying to prove a negative, but if you can do it I'm all ears. If your study observes the same rigor as Prof. Kunchur's work, I may even donate to the cause. Perhaps other GS members will too.

[James Lehmann]

Originally Posted by Petrus

In my test file the ONLY difference is the resolution... SRC was done with Sox utility using Shibata dither.

In mathematical terms maybe, but in this case 'resolution' relates to the two significant variables you mention above that require you to qualify your extremely vague initial question - "Can you hear 16/44.1 from 24/96?" - before it becomes a useful one to ask.

If you'd asked: "Can you hear the effect of SRC and dither on this recording?" then fair enough, but this isn't the same as asking "Can you hear the difference between these two recordings (of the same performance) made at different bit depths and sample rates?"

Which one are you asking?

[Petrus]

Originally Posted by James Lehmann

In mathematical terms maybe, but in this case 'resolution' relates to the two significant variables you mention above that require you to qualify your extremely vague initial question - "Can you hear 16/44.1 from 24/96?" - before it becomes a useful one to ask.

If you'd asked: "Can you hear the effect of SRC and dither on this recording?" then fair enough, but this isn't the same as asking "Can you hear the difference between these two recordings (of the same performance) made at different bit depths and sample rates?"

Which one are you asking?

Many people also on this site claim they can instantly hear the difference between 16/44 and higher resolution formats. This test file is only designed to help those people to show that this is indeed the case. The only difference between the different random samples is the fact that some of them have gone through a double SRC where their resolution has been dropped to 16/44.1 level.

If you care to take the test 2 different outcomes are possible and we can draw some conclusions from those:

1) No difference can be heard: recording at 24/96 and carefully downconverting to 16/44.1 seems to work extremely well and RBCD is good enough for final delivery. This does not prove that recording nativelly at 16/44.1 works as well. I would think it would, but somebody else can test it. Besides, if this route gives a "perfect" result, why not use it?

2) Difference can be heard with enough certainty: it seems that at least this method of conversion is not good enoug. I would also venture to say that any 16/44.1 native recording is not good enough, but somebody else can do the testing as there are countless converters and dithering alogrithms.

I am not claiming this test is the ultimate answer to all digital recording questions, it is just an easy way for those who have golden ears to prove that the difference can be heard in real life.

If the difference can be heard, fine. If not, the way the 16/44.1 sections in this test were constructed really does not matter, as the result is good enough.

[boojum]

Has anyone identified which is which in the test sample??

[Yannick]

I don't think anyone will.

For me the only difference between 24/96 and 24/44.1 is how well the AD converters perform.

I've never heard the difference between 24/96 and a SRC 16/44.1 version of it.

Changing interconnects or speaker cables has two orders of a magnitude more difference to it.

Or using balanced power supply, makes 1000x more difference than sampling rate.

Conclusion: a well-designed AD converter will sound as good at 44.1KHz as 96KHz.
I would NEVER record at 16 bit, however.

[aracu]

Originally Posted by boojum

I am not sure why responsible scientific testing is rejected. It seems to be most disbelieved in the audio realm if my experience is correct.

The mind is capable of believing, literally, anything, given the right circumstances and incentive. I know paranormal investigators who will record sound in a cemetary which is surrounded by a sidewalk. The voices of far off children walking by are picked up by the microphone, believed by the researchers to be the voices of children long passed away, communicating through the audio recorder. I have also seen that in blind experiments with musicians listening to recordings, real recorded monastery reverb can be percieved as being artificial, and artificial crafted monastery reverb can be percieved as being real.

[James Lehmann]

In the red corner...

Originally Posted by Yannick

...a well-designed AD converter will sound as good at 44.1KHz as 96KHz.

In the blue corner...

Originally Posted by ISedlacek

...if we record in 96k and then we record the same in 44k, the result will sound much better in 96k.

Gotta love Gearslutz!

[Soundwave]

Interviews

Tim de Paravicini: King of Tubes
By Steve Harris • November, 2007

Stereophile.com



When CD arrived in the 1980s, Tim de Paravicini was among the first to explain the shortcomings of the new format's sound quality by pointing out that existing analog media were superior when analyzed in terms of sampling rate. He argued then that a digital medium would need a much higher sample rate than 44.1kHz (and a higher bit rate than 16) to match the resolution of analog tape or vinyl. I asked him to explain this again.
"When it comes to digital, it's how to operate it, how many bits we devote to it, and the sampling frequency, as to how we store that information. The original digital system of CD, with 16 bits and 44.1kHz sampling, was what the mathematicians deemed to be the minimum acceptable to human hearing for so-called hi-fi. They never looked at all the artifacts and all the problems. And they never did enough analysis of the human hearing mechanism to realize that we don't stop hearing at 20kHz—people can discern and detect sound up to 45kHz. We have, as I say to people, an equivalent risetime of 11 microseconds in the hearing mechanism. And the ability to resolve detail in those digital systems wasn't quite good enough.
"In analog, you can change the thing and keep on aspiring to perfection without a compatibility issue. With digital, once you change any parameter, you've got a compatibility issue. Now, you can record on ProTools at 24-bit/192kHz, but it's not compatible with CD. I did my own summation—and this is from 20 years ago—that if we did 384kHz at 24-bit, we'll have a system that will resolve on a par with the best analog.

http://www.stereophile.com/interviews/1107parav/index1.html

[Ozpeter]

That's really hardly worth replying to, but to take a couple of points -

The original digital system of CD, with 16 bits and 44.1kHz sampling, was what the mathematicians deemed to be the minimum acceptable to human hearing for so-called hi-fi.

Wrong, actually. 48kHz was the originally chosen sampling frequency but this was deemed by the marketing people to provide too short a running time for a CD.

And they never did enough analysis of the human hearing mechanism to realize that we don't stop hearing at 20kHz—people can discern and detect sound up to 45kHz.

That can quickly be knocked on the head. Take a typical 96kHz sampling rate recording of musical material, run it through a frequency analysis display and note the average (or max if you really want) level of the signal at (say) 25kHz. Generate a sine wave at that frequency at that level and replay it at the original level you would play the original recording, with sections of silence interspersed. Raise you hand on hearing the 25kHz tone. You won't move. You might also care to set up your recording system in the room to see whether your speakers are in fact reproducing the tone and if so, what tiny movement of the mics makes how much difference to the re-recorded level, and then make sure you can keep both ears at the right point(s) without moving for the length of a symphony.

"In analog, you can change the thing and keep on aspiring to perfection without a compatibility issue. With digital, once you change any parameter, you've got a compatibility issue.

Now that's certainly very true.

[ISedlacek]

Originally Posted by James Lehmann

In the red corner...



In the blue corner...



Gotta love Gearslutz!

I don't want to sound like a broken record - but anyone can try it and compare for himself ... So easy ...

[audio ergo sum]

Originally Posted by ISedlacek

I don't want to sound like a broken record - but anyone can try it and compare for himself ... So easy ...

Actually doing a proper ABX test is not THAT easy. After you did that and scored at least 95% you can say it is easy to hear a difference.

Without that, whatever you hear and satisfies you, means nothing to anybody else.
If many people who believe the same thing but can't proof it come together, it is usually called a religion.

[boojum]

foobar2000 will do ABX comparisons. There must be the two separate files. foobar2000 is free for the download.

Download foobar2000 and optional components

[joelpatterson]

Originally Posted by audio ergo sum

...
If many people who believe the same thing but can't proof it come together, it is usually called a religion.

I would usually call that "The Politics Forum."

But anyway-- is there some unspoken assumption that "the listening experience is superior" with greater resolution, bit depths, file sizes, all of it?

Maybe just maybe "total listening satisfaction" breaks down as some small percentage for "realism compared to the original source" and overwhelmingly in favor of emotional factors, the ability to relate to the music, engineering savvy?

Greater realism is all to the good, but it just may not "matter" in the big picture. I know I sure feel like 24/96 is "way, way better" than 16/48... but who besides me really cares. You know?

[ISedlacek]

Originally Posted by audio ergo sum

Actually doing a proper ABX test is not THAT easy. After you did that and scored at least 95% you can say it is easy to hear a difference.

Without that, whatever you hear and satisfies you, means nothing to anybody else.
If many people who believe the same thing but can't proof it come together, it is usually called a religion.

Actually THIS already sounds like a kind of "religion" to me ... If (in general) a difference is such negligible that you need to perform a laboratory test under strict conditions to be able to identify it, it has no meaning for a normal use and it really does not matter.

The only difference that matters is the one that is clearly audible to everyone who has ears, taste and experience. I don't need to perform an ABX test and to score 95% to tell that I like one meal or one ice cream better than the other. Tasting it repeatedly many times and in a systematic way, may actually rather mess the real straight experience and result in a kind of nausea.

Similarly, I don't need the same to tell that I like say Earlybird preamp better than Behringer etc. (or at least that there is a clear difference between these two, irrespective even of some minor performance or level differences). If you are going to tell me that this is just my imagination and without a proper ABX test etc. - I can just smile at you and recommend you to taste a real life a bit more ... Actually, the hearing perception may get easily tired or less focused when you force it repeatedly, under some tension (of a "laboratory" experiment) etc. So what you call a "scientific ABX test", is actually very much dependent on a subjective hearing, that gets off its natural fresh perception and may start being tensed, messed, tired and unreliable if being forced too much. So where is anything objective in it ?

Sure, one always tries and compares few times but then he knows. Other one cannot distinguish, because his ears are less trained or he is less experienced and does not know what to focus at.
It is easy to compare 44k and 96k recordings and to hear or not to hear a difference.

If you don't hear a difference, good for you ... then you have no problem to decide. If others do hear a difference, why do you think they listen through your ears ? Most people hear a difference and that's why they use higher sample rates. That's all.

How can you "prove" that you like one woman better than the other ? Do you usually perform a repeated blindfold ABXY test, waiting to reach at least 95% score ? Good luck ...

But a large number of men (me included) may tell you that all the women are more or less the same in the end

One could just wonder what your point is. You probably do not claim that there is no difference between 96 and 44 recordings. Because you yourself said that you have compared and "prefered 96k". And to reach that conclusion, you must have for sure duly performed all the strict and complex rituals (cannot imagine otherwise), so your conclusion should be taken as a final word and clear the doubts for everyone else, I suppose.

[Ozpeter]

I don't have a shred of problem with your likes and dislikes, Ivo, nor would I presume to contest your tastes. As long as we are talking about subjective individual preferences there's no contest. But as soon as we talk about one thing being demonstrably better or different from another, then a demonstration is necessary. And then we're into a whole can of worms.

As for "why not try it", well, I suspect that most people here have by now.

They may have decided that they can tell the difference themselves, or that they can't tell the difference but their clients believe they can, or that they might as well use 24/96 because it's there and there's no downside for them and it's what some sectors of the business expect, so they've continued to use it.

Others may have tried it and gone back to lower standards because they have not perceived a difference, or their clients don't care or don't know (for instance for FM broadcast it's really hard to justify), or they feel the gains are outweighed by the storage and processing downsides.

It's bound to be a personal thing, and a matter on which disagreement will be endless - until more research is conducted along lines which leave no scope for dissent, and I suspect the cows will have come home long before that happens.

Because of the times in which we've lived, we are perhaps inclined to assume that there's something universally acknowledged to be better (in whatever field) round the corner. Supersonic air travel is bound to come - but where's the next Concorde (or even the last one)? Better than CD audio will be in mass market production any day soon - but how's sales of SACD going? In the case of audio, there is something which the market seems to consider better than CD - and that's downloadable mp3 for replay on portable devices. "Better" is seen as smaller delivery size and convenient playback methods - not lower noise or wider frequency response.

So perhaps some who don't bother even to try 24/96 are simply people who regard it as of academic interest in that context, and life's too short to put in the hours required really to reach a decision in which they can have confidence, given the (surprising) lack of authoratitive data on which they might otherwise base their decision.

[12ax7]

Originally Posted by ISedlacek

How can you "prove" that you like one woman better than the other ? Do you usually perform a repeated blindfold ABXY test, waiting to reach at least 95% score ?

Ah...Well, not exactly "usually."

But thanks for prompting me into remembering the good times!



(...And by the way, in my opinion it was closer to 98.6!)

[Petrus]

Originally Posted by ISedlacek

It is easy to compare 44k and 96k recordings and to hear or not to hear a difference.

Yes, with the test file I provided.

Originally Posted by ISedlacek

If you don't hear a difference, good for you ... then you have no problem to decide. If others do hear a difference, why do you think they listen through your ears ? Most people hear a difference and that's why they use higher sample rates. That's all.

If so, how come NOBODY has come out with the right sequence?

[Bradlo]

Maybe it would help to think of things this way.

At 44k, you will have 2 points to "draw out" a 22khz wave cycle. We don't hear that very well, so who cares, right?

But, we hear 11khz quite well, which is "drawn out" by only 4 points. Not such an accurate picture... pretty much a square wave. Now think of a 12khz wave. How are four points going to draw out 12khz to look very different from 11khz? Even worse, how about 11.1khz?

Yes, I know that high frequency square waves are rounded out by LPFs in the D-A stage, plus the actual physical limitations of the speakers themselves, but you should still be able to distinguish all the different HF sounds in a mix from each other.

Just my opinion

[James Lehmann]

OK. Let's say, for the sake of argument, that we could all agree on the parameters and logistics of an ABX test that features two WAV files of the same performance recorded by identical equipment at 24/96k and 16/44.1k respectively.

And let's say, for the sake of argument, no-one here is able to correctly separate the two files more than a random 50% guess would.

Where would this leave your opinion?

Would it prove that 'there is no difference between 24/96k and 16/44.1k'?

or

Would it prove that in a particular ABX test of that particular recording it wasn't possible for the people who took the test to reliably tell the two apart?

or

Would it prove that the particular converter used for the test did not behave audibly better at 24/96k than it did at 16/44.1k according to the ears of those who took the test?

or

[...insert conclusion that merely affirms your existing opinion]

[audio ergo sum]

Originally Posted by ISedlacek

Actually THIS already sounds like a kind of "religion" to me ... If (in general) a difference is such negligible that you need to perform a laboratory test under strict conditions to be able to identify it, it has no meaning for a normal use and it really does not matter.

The only difference that matters is the one that is clearly audible to everyone who has ears, taste and experience. I don't need to perform an ABX test and to score 95% to tell that I like one meal or one ice cream better than the other. Tasting it repeatedly many times and in a systematic way, may actually rather mess the real straight experience and result in a kind of nausea.

Similarly, I don't need the same to tell that I like say Earlybird preamp better than Behringer etc. (or at least that there is a clear difference between these two, irrespective even of some minor performance or level differences). If you are going to tell me that this is just my imagination and without a proper ABX test etc. - I can just smile at you and recommend you to taste a real life a bit more ... Actually, the hearing perception may get easily tired or less focused when you force it repeatedly, under some tension (of a "laboratory" experiment) etc. So what you call a "scientific ABX test", is actually very much dependent on a subjective hearing, that gets off its natural fresh perception and may start being tensed, messed, tired and unreliable if being forced too much. So where is anything objective in it ?

Look, peace upon us first of all. We should not argue. You don't accept the scientific method and there is not much else to talk about then.

The argument you give is purely sensual and without knowledge. You don't know about the scientific method and also not about the neuroscience involved in hearing and so you can not argue about it.

You have all the right to say that you enjoy this or that more and nobody will argue it.
When you start to say "A is better than B" then you leave that secure room of personal believe and you enter the arena of fact and rational with that factual statement.

If you follow the phenomenological method you would know that what your "is" is not what an "is" in the real world means.

In short you can't state what is without agreeing to an ABX test.

...
How can you "prove" that you like one woman better than the other ? Do you usually perform a repeated blindfold ABXY test, waiting to reach at least 95% score ? Good luck ...

But a large number of men (me included) may tell you that all the women are more or less the same in the end

So many jokes and no time...
Seriously, that comparison is flawed. We are not trying to win A/D converters for us. We don't expect them to love us either. We buy them and we measure them rationally, based on a return of investment principle. Is it worth it basically... (I can see the jokes coming :-)
If you personally have inherited a fortune and are not bound to such limitations I say good for you, enjoy the ride and keep sponsoring the Audio industry.

One could just wonder what your point is. You probably do not claim that there is no difference between 96 and 44 recordings. Because you yourself said that you have compared and "prefered 96k". And to reach that conclusion, you must have for sure duly performed all the strict and complex rituals (cannot imagine otherwise), ...

Yes, because there is a big difference between the canvas I paint on and the final frame I present the picture in.
For production I want a transparent as possible format. Forget about dynamic limitations. Smooth transparent Anti-Aliasing filters etc. For technical reasons. Basically for the same reason a professional photographer uses a Hasselbald with a much higher resolution than needed in the end, but leaving the option to crop a smaller range of the picture (e.g. raise the level or do lots of processing in audio) later. No limitation of bandwidth and dynamic at the source with unknown or mediocre real time equipment, but later in mastering with the best available algorithms, if necessary non-realtime.

In final delivery - this is what this thread is about - I could not reliably tell 44 and 96 apart in ABX testing, even when I thought I could. So couldn't all of my highly trained colleagues worldwide, not a single one of them. You would be the first exception...

BTW, congratulations on the new Geithain 901K, my favorite speaker also. You are also a dealer for them now?

How do you perceive the bass response compared to B&W 801? In my experience the Bass cardioid principle changes the room response quite drastically. Building new rooms for it would require less bass treatment. Rooms that have already been treated extensively for radial bass speakers could even sound too weak in the basses then.

[ISedlacek]

Originally Posted by James Lehmann

OK. Let's say, for the sake of argument, that we could all agree on the parameters and logistics of an ABX test that features two WAV files of the same performance recorded by identical equipment at 24/96k and 16/44.1k respectively.

And let's say, for the sake of argument, no-one here is able to correctly separate the two files more than a random 50% guess would.

Where would this leave your opinion?

Would it prove that 'there is no difference between 24/96k and 16/44.1k'?

or

Would it prove that in a particular ABX test of that particular recording it wasn't possible for the people who took the test to reliably tell the two apart?

or

Would it prove that the particular converter used for the test did not behave audibly better at 24/96k than it did at 16/44.1k according to the ears of those who took the test?

or

[...insert conclusion that merely affirms your existing opinion]

Yes, exactly. Since this uses again the human subjective/erring factor, this time being prone to defocusing the perception due to repeated set of inputs, it is not any more reliable than anything else.

Now - if one person finds something in a set of "ABX", the other the opposite and the other just something between (it can easily happen), it can prove nothing. Even if one person scores 95% (in either direction), it does not mean much. How can we know that this person can hear well at all etc.