Why do some Objectivists fear Psychoacousitics?

[Robh3606]

In short, those who seek realism in the reproduction of music, have little or no interest in objectivists or blind testing.

Wow your kidding right. I mean you can't be serious. So blindly stumbling trying to tweak this or that is the road to nirvana. Don't think so. Ever try building your own speakers?? You can take a measurement and see exactly what you have no guessing just the facts right there in front of you. To do it any other way is just plain dumb and in this day and age completely unacceptable for a manufacturer or for a home builder.

Rob

 

[amirm]

They are not interested in any of the concerns of blind testing.

What is your guiding light if I may ask when it comes to speakers? Is there no true north? Just a sea of literally thousands of products to listen to and hope one fits what you think is high fidelity sound?

And that audio science needs to be that lost in the woods or else is not valid?

I like to think after decades of audio research and building stuff randomly that we have figured out the general direction a speaker design needs to take. To believe otherwise is to believe in us, the design and research community, to be awfully stupid, lazy, and far worse than any other field of science. Is that what you really think the reality is?

 

[amirm]

Here is a strange but I promise you related question to the topic at hand . Why do think dogs tilt their heads this way?

 

[JackD201]

Wow your kidding right. I mean you can't be serious. So blindly stumbling trying to tweak this or that is the road to nirvana. Don't think so. Ever try building your own speakers?? You can take a measurement and see exactly what you have no guessing just the facts right there in front of you. To do it any other way is just plain dumb and in this day and age completely unacceptable for a manufacturer or for a home builder.

Rob

I agree. When one buys a speaker using only his ears, the objective science has simply been delegated to the designer/manufacturer. Like any tool, its level of utility lies in the person that uses it.

The DIY horn community is very vibrant here. Using basically the same parts it's amazing how large the variances in SQ can be. 9 out of 10 times it's the guys with engineering know how, whether formally trained or passionate enough to do the studying on their own that do well. Odds far better than plain old luck.

 

[JackD201]

Here is a strange but I promise you related question to the topic at hand . Why do think dogs tilt their heads this way?

They are watching Young Frankenstein and each one is looking at a different Marty Feldman eye?

 

[jkeny]

So, this guys hearing I suppose did not improve in time, however, he must have become more critical a listener, which supposedly causes fatigue and rules out blind tests? Or, oh my, did someone (like Harmons training course) teach him what things to listen for? In any case, now does he subscribe that he can hear differences between any two dacs we should put up to listen to? I doubt it. He has trained himself to be a more critical listener, and hence in the past was not a very critical listener is all I can gather from that example. Just sayin. And frankly, in all my time as an audiophile, I have never heard another audiophile say every component he listens to, every system, every room sounds the same, its kind of like a kid saying he can identify sweet from sour, not really much to crow about really.

I really don't see how he can blame blind tests for his inability to critically listen and hear "those all important and so what differences". He certainly had ample opportunity to develop his opinion based on lack of knowledge of what to listen for, but the blame lies on him for not asking what or how others listened for IMO, if it was all that important to him in the first place. I would have just thought, OK he cant hear differences, so maybe he is not as critical a listener as he might have though he was, that's all.

All too often, if audible differences aren't obvious in a quick A/B comparison then people consider that there is no difference worth talking about or investigating - mostly they actually think there is no audible difference, full stop. And that's the end of the story for most & they will defend the POV that there is no difference between DACs/amplifiers/whatever. Some defend it gently, some defend it aggressively. What was interesting about this guy was that instead of the usual attitude, which snear at & derides those who relate their experience of heard differences, he retained an open mind & it drove him to investigate further.

He is unusual in this regard, IMO - usually a closed-minded attitude is encountered, reinforced by measurements, the mathematical perfection of signal theory & digital audio, & null results from blind tests.

The point I'm making & he reinforces is that it requires trained users as listeners in blind tests, not the usual hobbyist. BUT, hobbyists submit themselves to these half-baked blind tests, hear no audible differences & are reinforced in their bias that no difference exists - its a self-fulfilling prophecy. We would be far better off without these types of blind tests but unfortunately, they are promoted by objectivists who, when encountering an anecdotal claim invariably look for evidence from a DBT but I never see them querying a null result from a DBT/ABX test.

Arny's ABX test lasted for 15 years of null results without any examination of it's validity & only when recent positive results were reported was the Foobar ABX test examined & changed. Changed, in the main, to prevent gaming the test - still no consideration to include controls for false negatives.

 

[TBG]

Basically, I personally don't think we know how to build the best speakers. All of our mechanisms for vibrating air have shortcomings. All cabinet or frames vibrate. Most speakers are able to get down to 60 or 50 Hz at most at 3 db down. Most continue to think that there is no need to reproduce sounds above 20k Hz, despite evidence that somehow the mind is mindful of sounds well above that. How? Perhaps other than movement of the ear drums. As I said, we need the perfect reverse of the microphone.

And yes, long ago I did build my own speakers as well as amps and preamps. Yes, I could have declared an EE degree, but didn't.

But what I really mean is that customers don't buy wine because of how it is made or the measures of it. They don't buy a good many things based on measurement of these products. They buy what they like. And many are only interested in reproducing "good sounding" or "listenable" music, not realism. I have recently realized this in comparing cables.

Fundamentally, I think most audio and especially speakers are an art and not a science. I have participated in enough blind and double blind tests and even sighted A/B comparisons to know that I think humans fail when told it is a test rather than just taking a listen.

So go about your efforts, but don't get sanctimonious about your interests, most don't put any value in what you do.

 

[jkeny]

Logically, if the recording contained all the information needed to create a stunning image of the live performance, then truly accurate reproduction and the war to make as many zeros after the decimal in IMD and THD tests and the least amount of spurs in the spectrum tests would win out.

But, the reality is stereo is quite incapable of reproducing the real thing, and folks all hear differently and since they are spending their own money they want the thing to sound good to them.

No reproduction system is capable of reproducing the real thing - so it's a case of maximising the system up to it's limitations. I'm not sure we have fully discovered it's limitations yet. I don't agree "folks all hear differently" - we are all governed by the same auditory processing mechanisms & apart from some anomalies we generally all hear in the same way.

I don't know anyone who can pass a moderate hearing test that could be played music from different speakers and not hear differences. Hearing differences is not tantamount to being able to describe accuracy to the original event since the original event is someones idea of what they think you would like to hear, in plain old stereo.

If two channel audio was a replication system that could bring the event to your room, then pure accuracy to the signal would likely be the measure of reference, but since it can't the measure of reference is each persons ears and internal references deciding what sounds good at that particular moment.

But accuracy is the premise on which the replay side of the audio chain is founded by objectivists . Seeing as we don't know the relative importance of the elements in the signal that our auditory processing system uses, accuracy is the sledgehammer answer - the thinking being, if you can recreate the input signal with accuracy then you don't really have to know the auditory processing rules. There are a number of problems with this approach - one you've identified already - possibly, the limitations of 2 channel reproduction but a bigger problem is that there is no system that is completely accurate & so we are in the situation of what elements to measure & what level of accuracy is acceptable. The problem here is that our measurements don't allow us to predict what we will hear because they are not the full set of measurements & the JNDs established for hearing are lacking because they have traditionally used simplistic signals. Maybe it's time objectivists looked at developing new measurements & JNDs?

Two channel audio is an entertainment system trying to trigger in our heads some semblance of what we want to hear and like to hear based on uniqueness that each of us has. However, those Harmon tests do show that taking the speaker as a sound reproducer, folks tend to like one that is more accurate. Its a fact for the way they did their test and test conditions.

While I do not align myself fully with either camp, the part of me in the objectivists camp is not afraid of psychoacoustics, however, I think the reality is that the other side of the camp is more afraid to find out just how trustless their hearing is. They have staked all they know, over and over on forums, about how their ears are superior. they are, to them. but to some of us, with some knowledge of the audio art, we know better. Looks like Amir has been one to mention how you can begin to doubt yourself when the differences are so small as you begin to start hearing things....I have done this in designs where I have made a change only to find out that the lead fell off the component and there I was thinking, hummm, this sounds better. Humility is one thing those with "ear trust" could use more of IMO. However, only some tout their experience as some sort of trump over others. Thankfully most are simply stating their opinion.

actually, two channel audio is a system simply to allow two speakers to generate images (in our heads) between them, read the patent.

The underlines statement is, to me, is the crux of the matter - IMO, quick A/B testing tries (unsuccessfully) to deal with this. It attempts to do so by ignoring the majority of & most important part of the workings of our auditory perception - it attempts to force this system into performing like an instrument to perceive audible differences. This only works for certain types of differences & is fraught with difficulties - the very difficulties that you are suggesting - the large number of factors that influence the results of this test. Such A/B testing will likely work with speakers & other devices that have certain types of differences between them that the test subjects have been trained to hear but look at Harmon's test - they also have to use one speaker because stereo is a confounder. In other words, they eliminate all the binaural information that we normally hear in our stereo systems. Does this not show much this type of instant A/B testing is fraught with difficulty? If clearly identifying differences between speakers requires the test environment to be changed so far from our normal listening, then what is the efficacy of these A/B tests at uncovering more subtle differences in a stereo listening setting in normal rooms?

Yes hearing is a guessing game where the processing involves continuous parsing of the audio signals to make auditory sense of them. It is inherently unreliable. IMO the best way of dealing with recognising this is using long term natural listening as a means to evaluate a particular device. I don't believe that our preference in an A/B test will necessarily correlate with our preference over long term listening because all the perceptual factors ignored by quick A/B listening then come into play.

So, I think you have taken the polar wrong view of what is the core issue

 

[amirm]

They are watching Young Frankenstein and each one is looking at a different Marty Feldman eye?

Bingo. You got it . But there is another answer related to this thread.

They are tilting their head sideways so that each ear is at a different elevation. That way, they can detect the height of the sound source. The sound will arrive at different times to each ear due to differing distance, and the cognitive part of their brain analyzes that and figure out how high the sound source is. Psychoacoustics in its simplest but very useful form.

Likewise, we have learned a lot about listener preference for good sound reproduction out of a loudspeaker. Take reflections. The ones from the sides help to stretch the speaker location to those positions. Listening tests show that we very much prefer that. Now take reflections that come from the front or back. They do not have this benefit. Floor reflections don't either and can color the sound when they combine with the direct sound of the speaker. But listening tests show that this happens at frequencies above 500 Hz. Which is a lucky break because a thick carpet and padding is able to absorb that. Whereas if we wanted to get rid of reflections to much lower frequencies we may have needed to 4-6 inch thick absorbers!

Those side reflections work best when they are tonally similar to the direct sound. In that case, the brain learns that they are just shadows of the direct sound and no longer treats them as harmful "echos." Take a good speaker like this and put it in a room. At first you hear the room+speaker. But after a few minutes, a lot of the room sound melts away as the brain realizes that the echos/reflections are of no informative value and filters them out. This is why a speaker tends to have its own sound almost regardless of which room you place it in.

Similarly your loved ones sound the same no matter where they are in your house. The acoustic characteristics vary hugely yet the brain filters them all out. My wife sounds the same to me whether we are in our master bathroom with its hard surfaces or the living room with carpets and furniture.

So as you see, a simple thing like which direction a reflection comes from has been analyzes and the results highly instructive in how we create our acoustic environment and design our listening space. In the last 20 years, we have learned a lot. Speaker companies like KEF, Paradigm, PSB, Harman (Revel, JBL, Infinity) use a lot of this science to design their speakers. Many others do the same but don't say it.

As the thread says, we seem to fear psychoacoustics. Heard on the radio yesterday of a great line by Martin Luther King Jr. That men fear each other because they don't know each other. Likewise, once we learn the science, it makes sense. And if you experience it such as sitting in the blind listening tests, it really sinks in. That is what happened to me 5 years ago when I got immersed in the science of audio reproduction. I didn't want to believe much of it. Literally hundreds of technical papers and in-person education later, the science became overwhelmingly convincing.

To be sure, we don't know 100% of this field, and 100% of listeners won't agree with each other, but there is a ton, ton of convincing data we need to take into account before we spend our money on expensive speaker and room treatments. A couple of hours in a demo room at a high-end audio salon or CES is not it I am afraid. Lest you want to know less than our faithful dogs know about this field.

 

[JackD201]

I see you are referencing the Haas time window Amir. I always cringe when I hear the "all first reflection points are bad" mantra. Really.

 

[jkeny]

Here is a video demo of this localisation (ITD sensing) in action

 

[jkeny]

Yes, Amir, Greisinger did a lot of research into what constitutes our perception of audio engagement in auditorium's - reflections being the major part of this. BTW, this video is from 11/11/14 just in case anyone thinks it's not current info- see here

 

[Robh3606]

Here is a video demo of this localisation (ITD sensing) in action

Hello Jkeny

Now that is truly amazing hearing through the snow and doing a blind spot on pounce!

Hello TBG

So go about your efforts, but don't get sanctimonious about your interests, most don't put any value in what you do

It doesn't matter what others value what matters is the results.

And yes, long ago I did build my own speakers as well as amps and preamps. Yes, I could have declared an EE degree, but didn't.

As someone with that background I am surprised you are not more open to current research.

Fundamentally, I think most audio and especially speakers are an art and not a science.

I disagree and think it's a bit of both with the science giving you a leg up and making it easier to accomplish your goal, a good sounding speaker

Rob

 

[jkeny]

Amir, it will be interesting to see your impressions of the SR Atmosphere when you get hold of it because, if I'm right about how it operates, it sounds like it directly addresses a certain peculiarity of our audio perceptions - comodulation masking release (CMR).

This would not be amenable to quick A/B testing, as the SR guy said to you

 

[TBG]

I find that an extraordinary post from an electrical engineer.
Keith.

That is because I didn't declare EE as one of my two majors, choosing physics and political science instead. I didn't think we knew enough to do quality circuits or the best speakers. I went on to getting my doctorate in political science although I had more graduate courses in psychology.

 

[amirm]

Amir, it will be interesting to see your impressions of the SR Atmosphere when you get hold of it because, if I'm right about how it operates, it sounds like it directly addresses a certain peculiarity of our audio perceptions - comodulation masking release (CMR).

This would not be amenable to quick A/B testing, as the SR guy said to you

There has been no follow up since the original contact so this evaluation may not happen after all. When he demoed it, the effect was pretty quick in that he stopped music, turned it on or off and played again. If it is that quickly audible, then it should do so in AB too.

 

[esldude]

snippage....


The underlines statement is, to me, is the crux of the matter - IMO, quick A/B testing tries (unsuccessfully) to deal with this. It attempts to do so by ignoring the majority of & most important part of the workings of our auditory perception - it attempts to force this system into performing like an instrument to perceive audible differences. This only works for certain types of differences & is fraught with difficulties - the very difficulties that you are suggesting - the large number of factors that influence the results of this test. Such A/B testing will likely work with speakers & other devices that have certain types of differences between them that the test subjects have been trained to hear but look at Harmon's test - they also have to use one speaker because stereo is a confounder. In other words, they eliminate all the binaural information that we normally hear in our stereo systems. Does this not show much this type of instant A/B testing is fraught with difficulty? If clearly identifying differences between speakers requires the test environment to be changed so far from our normal listening, then what is the efficacy of these A/B tests at uncovering more subtle differences in a stereo listening setting in normal rooms?

Yes hearing is a guessing game where the processing involves continuous parsing of the audio signals to make auditory sense of them. It is inherently unreliable. IMO the best way of dealing with recognising this is using long term natural listening as a means to evaluate a particular device. I don't believe that our preference in an A/B test will necessarily correlate with our preference over long term listening because all the perceptual factors ignored by quick A/B listening then come into play.

So, I think you have taken the polar wrong view of what is the core issue

Actually I think you have the the wrong polar view. If stereo is a confounder hindering accurate reliable results (which it appears it might be vs mono), the answer isn't listen longer with those factors. The answer is mono reduces the confounding factors so that a more highly discerning result is possible. So if you determine what makes for the preferred speaker in mono, you can be pretty sure it is good for stereo. If the interaction of two channels unveiled additional discernment, rather than confounding discernment, then results of stereo testing would be the better test. Again, the answer isn't listen longer, and Harman's results do not indicate stereo is more discerning.

Everything known about hearing indicates as the signal complexity increases the discernment of small differences is reduced. More complex signals can and often will have greater total information content. One shouldn't confuse that with better acuity of small differences with a more complex signal.

 

[jkeny]

There has been no follow up since the original contact so this evaluation may not happen after all. When he demoed it, the effect was pretty quick in that he stopped music, turned it on or off and played again. If it is that quickly audible, then it should do so in AB too.

It's a shame if it doesn't happen. I understand what you mean - I just sensed that it wasn't amenable to quick A/B style testing? I'm just surmising here but it may be more noticeable by stopping the auditory stream first & then rebuilding the perceptual auditory from scratch. Switching it on/off while playing may not be as noticeable because it's not like a distortion or amplitude difference coming or going - it's more an effect to the overall perception & we gradually adjust to the new perception if switched on/off?

Anyway, if you don't get one to audition, you won't be able to try various scenarios!

 

[jkeny]

Actually I think you have the the wrong polar view. If stereo is a confounder hindering accurate reliable results (which it appears it might be vs mono), the answer isn't listen longer with those factors. The answer is mono reduces the confounding factors so that a more highly discerning result is possible. So if you determine what makes for the preferred speaker in mono, you can be pretty sure it is good for stereo. If the interaction of two channels unveiled additional discernment, rather than confounding discernment, then results of stereo testing would be the better test. Again, the answer isn't listen longer, and Harman's results do not indicate stereo is more discerning.

But we only have one source, O'Toole/Harmon stating that mono listening preference correlates very strongly with stereo listening in normal rooms? Has this research been been repeated & verified as accurate by other researchers?

I also asked about Harmon's minimising of room boundary reflections in their tests - if they also do this their test is becoming more & more divorced from the conditions of real world use.

So, again this highlights what can be wrong with overly simplifying the test conditions (in order to make it easier to analyse) - it becomes so divorced from the actual usage configuration that extrapolating the results become meaningless - it's the case with a lot of measurements (over simplistic test signals) & audio perceptual tests for JNDs using simple clicks, tones. Failure to recognise this shortcoming is what the thread title is highlighting

Everything known about hearing indicates as the signal complexity increases the discernment of small differences is reduced. More complex signals can and often will have greater total information content. One shouldn't confuse that with better acuity of small differences with a more complex signal.

See my points above but also why then are all audio tests for small differences not therefore using mono signals. If you consider stereo such a confounder why does this not apply to DAC, amplifer, etc testing - I often see volume matching being cited for DBTs but not use of mono?

 

[esldude]

But we only have one source, O'Toole/Harmon stating that mono listening preference correlates very strongly with stereo listening in normal rooms? Has this research been been repeated & verified as accurate?

I also asked about Harmon's minimising of room boundary reflections in their test - if they also do this their test is becoming more & more unlike the conditions of real world use.

So, again this may be what's wrong with overly simplifying the test conditions (to make it easier to analyse) - it becomes so divorced from the actual usage configuration that extrapolating the results become meaningless - it's the case with a lot of measurements (over simplistic test signals) & audio perceptual tests for JNDs using simple clicks, tones. Failure to recognise this shortcoming is what the thread title is highlighting

See my points above but also why then are all audio tests for small differences not therefore using mono signals. If you consider stereo such a confounder why does this not apply to DAC, amplifer, etc testing - I often see volume matching being cited for DBTs but not use of mono?

The difference is Harman didn't do mono tests to simplify the test analysis. They did mono tests because they worked better in comparison to stereo tests. You could say it makes the human listening easier. But we pretty well would have expected that.

I would imagine one reason other DBT isn't done with mono testing is because of the complaints you would raise. And for some purposes academic research is done in mono. It depends on what is being tested. In terms of casual testing done by hobbiests you probably would have a hard time convincing them to compare say hiresolution music to redbook by doing it in mono (I don't know if that would work better). But hobbiests in high end audio are often difficult to convince about a good many absurd beliefs they hold.

It would be good to have others replicate Harman's work to see if it holds up or not. But Harman came to these ideas through plenty of work on their part. It wasn't like they set out with predetermined ideas of what was best. It is what their work and testing taught them. Further while they talk about it some it hasn't been any big part of their marketing. As in ad copy. They market their speakers much like anyone else. It would appear they believe their testing has let them build better speakers, and given everything in total they believe better speakers at a given price point will earn them more marketshare. So they are marketing the speakers not the testing benefits for design purposes.

Using simple tests, learning how things work, what is important piece by piece is an old time honored successful idea. Jumping whole hog into something and insisting you have to get it all in one go in totality or not at all has a history of slow or no progress with lots of confusion about what is important along with confusion about what works and what doesn't. When an area of life transitions from the old way to the newer way based upon investigation it is a repetitive story to see those decry the testing because the results disagree with previously held though foggy ideas about something. Though the early period of progress can sometimes be two steps forward and one back, even that is tremendously more rapid and effective than what has gone before.