What objectivists and subjectivists can learn from each other

[Ethan Winer]

I was at an audio event 2 years ago and we did an A/B test between two different things. On one [A], I noticed a full sound stage. On , I noticed a complete drop into oblivion of only the extreme left of the sound stage. Are you saying that a null test would be able to pick up on this?

First, you'll have to tell me what "things" were being compared. Nulls tests aren't so useful for acoustic events. But to be considered valid, a listening test of sound stage requires your head staying in the exact same place, within literally one or two millimeters:

The above was meant as a joke, but it's literally true in this case.

--Ethan

 

[Ethan Winer]

Its simply that two channel stereo can not replicate a live performance

Yes, and this has nothing to do with audio fidelity, or subjectivism versus objectivism. That microphones and speakers can't precisely emulate a live event is a red herring IMO. Right up there with "You can't measure a symphony" which I've also seen argued many times when I talk about measuring audio gear.

BTW, it is possible to emulate a live performance pretty well, but not the way most recordings are made. If every instrument is mic'd very close, and everything in the signal path is very accurate, the output from the speakers will be close to the original. So you could reproduce the sound of a string quartet playing in your living room. But not the sound of the quartet playing in an auditorium.

--Ethan

 

[Phelonious Ponk]

OTOH, you also have made some comments recently that are just about as preposterous in the opposite direction, although I'll allow for the possibility that you might be exaggerating for effect.

I do, sometimes, exaggerate for effect, but I'm not aware of having ever gone that far. Examples?

Tim

 

[treitz3]

First, you'll have to tell me what "things" were being compared. Nulls tests aren't so useful for acoustic events. But to be considered valid, a listening test of sound stage requires your head staying in the exact same place, within literally one or two millimeters:

It was a test between an original CD and a lossless copy. I'd have to look up my notes to find out what equipment was used but the test might have well as been blind, as it was done through a remote control behind my back and I can assure you, my head did not move. The end result as to what hit my ears was rather profound, even if my head did move. I was able to pick out the difference 100% of the time in front of a room full of people, some of which heard the same exact thing off axis after it was pointed out to them as to what it was I heard.

 

[Ethan Winer]

can you give some detail about this "complete nulling" of two music tracks & tell us something about the resolution that this operates at?

When nulling Wave files, the resolution is that of the files. So if two 16-but files null to silence, that means any differences are more than 96 dB down. A total null also means both files are identical, proving that any perceived differences are due entirely to faulty perception and not a real physical difference.

I presume from your statement that the nulling available 50+ years ago was the equivalent of today's nulling efforts & hence your statement?

No, old H-P analyzers play a pure sine wave through the device being measured, then filter out the sine wave. Anything remaining at the output of the device is noise and/or distortion added by the device. So that's different from nulling two wave files. But the concept of nulling still applies, because the analyzer's residual output can easily be analyzed for level and spectral content. The notion that there's more to audio fidelity than is currently known is easily disproved, even with the old-school analyzers. I've seen the residual output of that type of analyzer many times, and all I ever saw was harmonics and noise, as expected.

--Ethan

 

[Ethan Winer]

It was a test between an original CD and a lossless copy.

This is a perfect application of the null test because the comparison is between two time-locked data streams. A null test is vastly more reliable than a listening test, whether sighted or blind. If you extract the Wave files from both the original and copy and they null completely, that proves the two sources are identical and any perceived difference is imagined. Even better, do a bit compare on the two files to see if they hold identical data. I know some people hate when I say this, but it's the truth. Indeed, what other explanation makes sense if identical bits are played back yet it sounds like they're different?

Now, had that test you were part of nulled the two files and found a difference, then we'd have a basis to investigate what went wrong with the lossy encoding or decoding.

--Ethan

 

[NorthStar]

Ethan, I am reading all of your posts, and they are a breeze of fresh air, thanx! :b

--Bob

 

[Ethan Winer]

A null test is vastly more reliable than a listening test, whether sighted or blind.

I should have elaborated more: A null test is better here because it avoids the added complication of amps and loudspeakers, the room, and human perception. In other words, if a difference really exists between an original Wave file and a copy after a lossy compression / decompression cycle, that difference can be determined just by examining the data.

And thanks bob!

--Ethan

 

[Phelonious Ponk]

This implies either that the ADC and DAC used were essentially perfect or that the test methods were unable to detect a difference (leading me to question all of the test's results). If the first thought is correct then no further work needs to be done on developing new ADCs or DACs; that seems to me an unlikely proposition.

It doesn't imply that anything was perfect. It Implies that the ADC and DAC were accurate within audible limits. This is not only possible, it's not even particularly unusual. DACs used to accurately convert digital audio to an analog signal? I'm not sure there is a lot of work left to be done. Jitter is to our time what THD was to the 70s, except this time the race for better-looking numbers, further and more pointlessly below audible limits is happening in the "high end" instead of Japanese midfi.

Have you ever read Meyer and Moran?

Tim

 

[rbbert]

I do, sometimes, exaggerate for effect, but I'm not aware of having ever gone that far. Examples?

Tim

Your whole recent shtick that unnamed midfi gear that "measures well" would sound better or more accurate than hifi gear used by members here. I would have thought we're well beyond that sort of statement, given the measurable excellence of many members' systems. How many people here do you think use SET triodes as their reference amps??

 

[rbbert]

Why do you believe that no current converters are audibly transparent? On what do you base that? Your own proper blind tests? Are you aware that a major pro audio converter manufacturer did a test where they re-recorded a piece of music ten times in a row, and even after ten generations nobody was able to tell the original from the output? I forget the vendor, and this was a few years ago. But that alone proves to me that one A/D/A pass can be truly transparent. So then the question is as I asked: Why do you believe otherwise?

--Ethan

Here's one interesting article

http://www.benchmarkmedia.com/audio-times-audio-conversion-systems-with-true-analogue-response

Although I haven't conducted blind tests, I have compared some respected ADCs: Korg (DSD) and Sound Devices (PCM), and they don't sound all that similar. Barry Diamente (whom most would consider more objectivist than subjectivist) finds the Metric Halo more "transparent" than any other ADC's he has tried. I imagine Bruce B. could also chime in on his experiences with different ADC's (unless he's already given up on this topic, which wouldn't be surprising).

 

[Gregadd]

Gregaad, I will answer you below folowing ---G...

Some well test instrumented objectivists can measure any electrical audio disturbance, period.

I'm not sure this matters but anything an objectivist can measure has been measured by a subjectivist. The problem is getting the subjectivist to agree that anything that can't be measured is a figment of his imagination---G...any audio electrical signal and any audio disturbance of air is measurable. The imagination of any person, or lets be more specific, the ear/brain interface can not be measured c urrently.

Some subjectivists (including on this forum) can "hear" an actual performance (they claim it is near pefect representation of the real live actual unamplified event) in their system.
Name one who has made that claim. I for one just don't think it is as imperfect(hopelss ?)as you claim.
---G..please, plenty of people on this forum describe "being there" when listening to their systems or some folks systems in particular..now I know that being there simply means that the illusion works for them. Just about any system, that you play in one room in the basement and I listen to upstairs, can be convincing, but put me in the room with the speakers and it sounds like two channel stereo to me. Stereo is not hopeless, but expecting replicaton of a live unamplified event is hopeless to this old boy.

The flaw, IMO is
the reality as far as this man is concerned,
is that fundamentally,
two channel stereo simply can not replicate an actual live event.
No one has made that claim. ---G...I make that claim, hundreds of thousands of audiophiles may disagree, but technically it is impossible for two channel stereo to replicate a live event, at any near distances.....if your speakers are at one end of a football field (with only flat land all around) and you are at the other end of the field, then I guess, yeah, your ears can be fooled, but as normally listened to, in a normal room at normal distances, no way do I hear replication.


The argument therefore to me, is in the degrees that the illusion is presented.....
Agreed. Obvioulsy, I and others claim we are much closer than you care to admit. ---G... Close is fine for 80+ year old technology. Unless you do some serious processing, the type of stereo you might listen to like off an lp or everyday cd can not get closer by swapping out components endlessly, by therefore changing the tones and sounds, cause two channel stereo will never replicate to the degree needed to sound real..just changing degress of the illusion by tone controls (compoenent swapping) will never get but more of the same.


--the objectivists insist that we need zero distortion to get there...no...we need something other than stereo to get us there, two channel stereo can not replicate live sound
.Subjectivist would prefer zero distortion. Since that is impossible we are forced to "pick our poison." ---G...subjectivists want zero distortion equipment? I never knew that, especially since the byword of subjectivits is they dont care about specs, just "sound". AND DONT GET ME WRONG, I have no problems with tone controls cause stereo cant get you there anyway.


--the subjectivits argue that they can be transported there with gear that has distortions...no...if stereo has to be distorted to get you there, it follows its not right to start with.
Again, we realize stereo reproduction is imperfect . We just think that it is far better than you do. ---G...fine, i dont think it is all that great, you do. Its not better than surround or binaural IMO. And binaural is older technology than stereo.


Tom[/QUOTE]



Tom the assmebly of a system is based on many factors. The two most dominant factors are what is commercially available and our desire and our ability to purchase it. I don't think I have the technical credentials to defeat your arguments, nor the financail ability to assmble a system that woukld demonstrate how misleading your statemnts are.

I do know if you stop searching you will never find it. I got on an air plane(I hate to fly) and my opinions about SET amp was forever changed.

My only fear is that otghers will accept your argument and never try. The fact that there is something better than stereo encourages me rather defeat me. Before you accept the argument of me or anyone else listen to the music. Good reviewers and dealers (for that matter some manufacturers) will tell you what muscal selections will allow thier equipment to shine. If you are lucky you can find an audio buddy with a SOTA system that will let you listen.

 

[jkeny]

When nulling Wave files, the resolution is that of the files. So if two 16-but files null to silence, that means any differences are more than 96 dB down. A total null also means both files are identical, proving that any perceived differences are due entirely to faulty perception and not a real physical difference.

No, I'm not talking about the resolution of the files - I'm talking about the resolution of the null i.e when a music Wave file is played through two different devices - when is a sample pair declared to be a null - at what voltage level is a null declared for two samples points , below a mV, below a uV, what?

No, old H-P analyzers play a pure sine wave through the device being measured, then filter out the sine wave. Anything remaining at the output of the device is noise and/or distortion added by the device. So that's different from nulling two wave files. But the concept of nulling still applies, because the analyzer's residual output can easily be analyzed for level and spectral content. The notion that there's more to audio fidelity than is currently known is easily disproved, even with the old-school analyzers. I've seen the residual output of that type of analyzer many times, and all I ever saw was harmonics and noise, as expected.

--Ethan

So you recognise the difference between playing a sine wave & playing a music file through a device? Can you still claim that a null test from 50+ years ago would uncover differences between devices or are you saying that the current method is more sensitive?

 

[Phelonious Ponk]

Your whole recent shtick that unnamed midfi gear that "measures well" would sound better or more accurate than hifi gear used by members here. I would have thought we're well beyond that sort of statement, given the measurable excellence of many members' systems. How many people here do you think use SET triodes as their reference amps??

You must have missed it the last time I answered this. It's on page 14 of this thread. Here you go...

Tim,

IMHO, your passionate summary strongly distorts the many contributions of many members to WBF debates.

It wasn't meant to apply to the many contributions of many members here at WBF, Micro, but it fits many others like a tailored suit. And out there in the rest of the audiophile world, at places like Audio Asylum, it applies much more broadly. The basic phenomenon I'm talking about is the enthusiast who uses equipment that is objectively inferior, insisting it is more natural, life-like, etc., based on some unmeasured goodness is not at all uncommon here. They may be right. We may some day discover that a previously undiscovered layer of resolution in old analog recordings that was being reproduced by vinyl and valves, but somehow altogether filtered out by digital and SS. We could also discover we evolved from lemurs.

You posted similar views several times, and some of us refuted this false argument, presenting you with cases and brands of excellent sound quality by audiophile standards that show state of the art measurements. You always stick to the SET case, that represents the minority of existing audiophile equipment, tube equipment being over-driven and tests carried with headphones. These had not been the typical conditions of our debates.

Take SET out of it, then. And understand that I completely acknowledge that there are many people on this board whose systems are SOTA, way above the capabilities of my own. False argument? Are you really saying no one here is consistently insisting that objectively inferior technology is more revealing, higher resolution, more dynamic, more natural, more life-like? One word, Micro: Vinyl. All of the above has strenuously been insisted upon here; none of it is borne out by the data. And WBF is hardly a good example. This place is hydrogen audio compared to some audiophile communities.

I not only think what I made was not a false argument; I think it was an obvious one. Repeating it was probably as pointless as saying the sky is up today. But it seemed like the answer to Ethan's question.

Tim

I think that addresses the excellence of many members systems here and the SET issue. Regarding unnamed midfi that sounds better than hi-end systems, I didn't say that. "Sound better" is purely subjective. Do you doubt that a lot of midfi measures better than high-end tube (the overwhelming majority, not just SET) pres and amps, better than vinyl and turntables? If you do, I'd suggest a bit of research. If you believe it does but don't care, good for you. It's good to know what you like.

Tim

 

[jkeny]

This is a perfect application of the null test because the comparison is between two time-locked data streams. A null test is vastly more reliable than a listening test, whether sighted or blind. If you extract the Wave files from both the original and copy and they null completely, that proves the two sources are identical and any perceived difference is imagined. Even better, do a bit compare on the two files to see if they hold identical data. I know some people hate when I say this, but it's the truth. Indeed, what other explanation makes sense if identical bits are played back yet it sounds like they're different?

So are you saying that the playback stage need not be considered & that if two files are bit identical they will sound exactly the same. What about playing back through different devices? Are two CDs from the same pressing bit identical - can they sound different when played back using different CD players or is this a placebo? What about the same file played back using two different playback software programs which outputs bit perfect streams - can they sound different?

 

[rbbert]

It doesn't imply that anything was perfect. It Implies that the ADC and DAC were accurate within audible limits. This is not only possible, it's not even particularly unusual. DACs used to accurately convert digital audio to an analog signal?
Tim

I have not heard of a DAC whose measured impulse response is not significantly abnormal, especially at 16/44.1. Strangely enough, there does seem to be some correlation between measured impulse response and subjective sound, although I know you don't acknowledge that. It's not a be all and end all measure, but it certainly indicates some work to be done. In fact, I think that's probably today's most fruitful area of investigation for improving audio.

 

[fas42]

In your opinion. The fact that this can not happen seems to escape you and quite frankly your credibility, Frank.


Growing up with a 3rd generation piano teacher, I can say with complete certainty that I have not. Close and perhaps pleasing but not realistic.

Sorry, that's being purely dogmatic. You state with certainty that something is impossible -- the history of science is littered with the utterances of people at the heads of their professions, and experts in their field, proclaiming that a particular behaviour or "fact" is impossible, only to effectively have egg in their face further down the track ...

Frank

 

[fas42]

Yeah, Tom...had you been here long enough you'd know that Frank goes around the house, shuts off wireless devices, unplugs things like microwaves, then he blu-taks the speakers of his Phillips home theater in a box to concrete blocks, solders speaker wire to terminals and mains cables to jacks and such. As a result, his speakers not only disappear when he has his ear just inches from the tweeter, they image balanced stereo from that listening position and play not only a piano, but a full orchestra, absolutely realistically --- at concert volume!!!. His mojo is so strong that speaker placement is irrelevant, the sweet spot is expanded outside of the listening room, and the worst recordings become life-like. In fact, it all has to do with the electronics in the middle of the chain. The recordings and the transducers are the least important elements. They hardly seem to matter at all.

Why do you question his credibility?

Tim

I have to admit that's an excellent summary, Tim, well done! Of course, the fact that almost no-one else does this means that no-one else, in fact, does "get" it. Weeell, you just have to believe, not because you want to, but because you've had the experience. As just mentioned in another post, the friend I'm lending a hand to is so darn close now it would be pretty scary to many of you here. Trying out a Reference Recording orchestral piece we had big, big sound, tremendous detail and tonality, and very convincing dynamics. Not bad for about $5,000 worth of kit ...

Frank

 

[treitz3]

Sorry, that's being purely dogmatic. You state with certainty that something is impossible....

You can stop right there, Frank. If there is one person on this board [other than you] that challenges that it is possible, then may they speak now. Even one and I would welcome the possibility.

Just one, Frank.

 

[fas42]

You can stop right there, Frank. If there is one person on this board [other than you] that challenges that it is possible, then may they speak now. Even one and I would welcome the possibility.

Just one, Frank.

Roger, Vince, Robert ...????

Frank