Analysis of Paper on Measurements of RCA Cables by Kunchur (Video)

[PeteL]

It would be easy to eliminate or confirm some of the biases. Record an instrument not normally associated with height using the "height" methods outlined. Vice-versa record an instrument normally associated with height and place it down low. Use various instruments that have different frequency ranges.

Or record the exact same instrument at different height without moving the microphones, I think that would be the most unbiased test, yes using pre-recorded music is indeed a weakness.

 

[PeteL]

The combined mics provide a very good sense of realism and accurately depict both the lateral position and depth in the recording. But my experience is without actually adding some form of HRTFs to the recording (either filter convolution or dummy head of some sort), any sounds that are behind the mics are simply mirrored to the front of the soundstage, and the height is entirely confined to the tweeter axis (e.g. everything falls in a 2D horizontal plane going into the soundstage and only lateral position and depth can be rendered accurately). Now, of course room reflections could play a part, but this effect is likely to be synthetic.

Thanks for the insight, I am not fully getting your point tough to be honest, you are saying that recording can be 3D, have height, but only trough DSP and not acoustically acccording to your experience correct? By experience you mean what you hear in your stereo system? How can you assess what is artificial and what is acoustic, and is it important? Does the stereo image is 3D or 2D, in the end it's what we want to figure out? And assess if this can be reproduced by a set of 2 speakers, I did not feel wheter acoustic or made up was a big element of the conclusion.

 

[Blumlein 88]

The combined mics provide a very good sense of realism and accurately depict both the lateral position and depth in the recording. But my experience is without actually adding some form of HRTFs to the recording (either filter convolution or dummy head of some sort), any sounds that are behind the mics are simply mirrored to the front of the soundstage, and the height is entirely confined to the tweeter axis (e.g. everything falls in a 2D horizontal plane going into the soundstage and only lateral position and depth can be rendered accurately). Now, of course room reflections could play a part, but this effect is likely to be synthetic.

When using crossed figure 8's the stuff behind the mikes shows up in front on the opposite channel. Have used that recording small groups. They can gather around front and back, and you still get stereo. Have to be careful they don't get too near the side lobes or they can sway around a bit and end up jumping back and forth between channels.

 

[Cars-N-Cans]

Thanks for the insight, I am not fully getting your point tough to be honest, you are saying that recording can be 3D, have height, but only trough DSP and not acoustically acccording to your experience correct? By experience you mean what you hear in your stereo system? How can you assess what is artificial and what is acoustic, and is it important? Does the stereo image is 3D or 2D, in the end it's what we want to figure out? And assess if this can be reproduced by a set of 2 speakers, I did not feel wheter acoustic or made up was a big element of the conclusion.

On a good system, take a fully 3D soundscape in reality, and squash the height down so everything is at the same elevation. That's what I experience. This is likely due to the lack of spectral coloration from the folds in the pinnae that one would normally experience in real life. We only have two channels of hearing, but the brain has learned over time to use the spectral coloration from the ears to identify the elevation, azimuth and distance from ones position. Standard mics don't have ears, so this is absent in the recording, and the brain thus uses the tweeter's elevation above/below the listening position as the reference for the origin of the sound.

 

[PeteL]

On a good system, take a fully 3D soundscape in reality, and squash the height down so everything is at the same elevation. That's what I experience. This is likely due to the lack of spectral coloration from the folds in the pinnae that one would normally experience in real life. We only have two channels of hearing, but the brain has learned over time to use the spectral coloration from the ears to identify the elevation, azimuth and distance from ones position. Standard mics don't have ears, so this is absent in the recording, and the brain thus uses the tweeter's elevation above/below the listening position as the reference for the origin of the sound.

OK, I’ll extrapolate a bit here. So would I be wrong to see your view as, beside Applying a HRTF or some binaural recording head, any elevation perceived would be a construction of the brain, Or you are saying that you don’t notice anything in the elevation axis when listening to your system. It’s a bit different. I hear 3D, but Î’m aware of biases.

 

[Cars-N-Cans]

OK, I’ll extrapolate a bit here. So would I be wrong to see your view as, beside Applying a HRTF or some binaural recording head, any elevation perceived would be a construction of the brain, Or you are saying that you don’t notice anything in the elevation axis when listening to your system. It’s a bit different. I hear 3D, but Î’m aware of biases.

In this case I listened pretty much sans-video. I just had it open in the default view to see what mics are in use. But, on a large screen TV or if you are paying attention to the video the visual cues are a powerful motivator. I have had this provide the needed "missing info" when watching movies on a 5.1 setup. It can be there, but that's likely due to the brain doing what it normally does and combining stimuli to resolve ambiguities.

 

[Cars-N-Cans]

OK, I’ll extrapolate a bit here. So would I be wrong to see your view as, beside Applying a HRTF or some binaural recording head, any elevation perceived would be a construction of the brain, Or you are saying that you don’t notice anything in the elevation axis when listening to your system. It’s a bit different. I hear 3D, but Î’m aware of biases.

Lets also not forget that there are expectations based on experience, too. But when it comes to setting up sound systems I have, over the years, forced myself to screen these out since it can bias ones perception, which is not what you want if you don't want to make assumptions about what you do or don't hear.

 

[DonR]

In this case I listened pretty much sans-video. I just had it open in the default view to see what mics are in use. But, on a large screen TV or if you are paying attention to the video the visual cues are a powerful motivator. I have had this provide the needed "missing info" when watching movies on a 5.1 setup. It can be there, but that's likely due to the brain doing what it normally does and combining stimuli to resolve ambiguities.

A perfect example of the McGurk effect. Visual stimulus overrides auditory every time.

 

[MacCali]

RadioShack rca cables Amir tested are super fire.

Maybe we should order a pair for Kunchur

 

[Cars-N-Cans]

A perfect example of the McGurk effect. Visual stimulus overrides auditory every time.

An interesting observation, and likely hard-coded into us to help in understanding speech in marginal conditions (e.g. a crowded room). Interestingly, if you look past the visual cues in the video below, its possible to here both "Faa" and "Baa" in the second half of the illusion, but not the other way around when the sound correlates to the facial and mouth movements. Shows that these are deeply wired into the brain at a fundamental level.

 

[MacCali]

An interesting observation, and likely hard-coded into us to help in understanding speech in marginal conditions (e.g. a crowded room). Interestingly, if you look past the visual cues in the video below, its possible to here both "Faa" and "Baa" in the second half of the illusion, but not the other way around when the sound correlates to the facial and mouth movements. Shows that these are deeply wired into the brain at a fundamental level.

How does that happen LOL that’s wild.

 

[Cars-N-Cans]

How does that happen LOL that’s wild.

I would say it likely shows how much we really depend on facial expressions and mouth movements for understanding speech. It is crazy how easily the visual cues can override what you actually hear.

 

[MacCali]

I would say it likely shows how much we really depend on facial expressions and mouth movements for understanding speech. It is crazy how easily the visual cues can override what you actually hear.

That’s like a glitch in the matrix. Just the lips moving in a different direction making the manipulation.

I can see it though in the same fashion someone learns to read lips. Oddly I never thought I would ever be able to do that and here we are doing unconsciously

Seems to be something along lines of peripheral logging. When you listen to someone talk you don’t look at that there mouth. Clearly you look at them eye to eye

And what further makes this strange is it’s not a common sound people make yet you are still picking up on it

 

[JSmith]

facial expressions and mouth movements

Some interesting studies and data on this;

What accounts for individual differences in susceptibility to the McGurk effect?

The McGurk effect is a classic audiovisual speech illusion in which discrepant auditory and visual syllables can lead to a fused percept (e.g., an auditory /bɑ/ paired with a visual /gɑ/ often leads to the perception of /dɑ/). ...

www.ncbi.nlm.nih.gov

However, we found that participants who were better able to lipread consonants and extract POA information from the visual modality were more susceptible to the McGurk effect.

Spatio‐temporal distribution of brain activity associated with audio‐visually congruent and incongruent speech and the McGurk Effect

Spatio‐temporal distributions of cortical activity to audio‐visual presentations of meaningless vowel‐consonant‐vowels and the effects of audio‐visual congruence/incongruence, with emphasis on the McGurk effect, ...

www.ncbi.nlm.nih.gov

McGurk perception was associated with decreased activity in primary and secondary auditory cortices and Wernicke's area before 100 msec, increased activity around 100 msec which decreased again around 180 msec.

https://royalsocietypublishing.org/doi/10.1098/rsos.181868

The McGurk effect is an illusion whereby speech sounds are often mis-categorized when the auditory cues in the stimulus conflict with the visual cues from the speaker's face. A recent study claims that ‘skilled musicians are not subject to’ this effect. It is not clear, however, if this is intended to mean that skilled musicians do not experience the McGurk effect at all, or if they just experience it to a lesser magnitude than non-musicians.

While the experiment did not reveal a significant difference between musicians' accuracy in audiovisual ‘McGurk’ stimuli and audio-only stimuli, they nevertheless showed a large numerical effect in the direction of a typical McGurk effect (higher accuracy in audio-only than audiovisual incongruent stimuli).

JSmith

 

[Peluvius]

An interesting observation, and likely hard-coded into us to help in understanding speech in marginal conditions (e.g. a crowded room). Interestingly, if you look past the visual cues in the video below, its possible to here both "Faa" and "Baa" in the second half of the illusion, but not the other way around when the sound correlates to the facial and mouth movements. Shows that these are deeply wired into the brain at a fundamental level.

I love this. Measurements are so critical to make sure we don't get ripped off on audio gear....seriously.

 

[voodooless]

It has been covered in this discussion Voodooles, but I'll answer.

Thanks for taking the time to answer some of the questions Pete, I appreciate that.

Is it so hard saying that you have a different point of view than mine?

It’s not. It’s about understanding your point of view. Isn’t that what these kind of forums are all about? To see what other people think about a subject and find out why that is. I think that is extremely valued, and that is why I keep asking questions. Before we agree to disagree, I need to know why you disagree.

The knowledge I get from this paper, is that perception of heigh and 3d stereo image is real, not imagined and that it can be portrayed and assessed in blind listening.

How does this paper prove that it is not imagined? Can you point me to those parts of the text doing that?

The conclusion can be justified because, in my view it has been assessed by a test that is sufficiently controlled, in my opinion. This blind testing "to the best of my knowledge" and I don't know everything, was properly conducted.

Mkay..

That a correctly set up two‐channel stereo system can in fact portray not only depth and lateral width (azimuth), but also elevation for appropriately recorded material?

I guess, in light of the above, that is not very much a surprise.

Yes. I agree with this conclusion unless someone could explain to me that this test is unconclusive, then I may revise my position. You did not explain why this test is unconclusive.

Of course I did.. reread the my findings.

I don't know, but how about he don't know?

No, not good enough!

Why would he have to know? He describe the system used for testing, with the components used and the fidelity metrics associated.

He does not just describe the system. There is nothing intrinsically wrong with using such a system and describing the components. His claim is that such a system is necessary! Part of the "proof" is his bogus paper destroyed in Amir’s video. Even if that paper were true, that still does not prove it is necessary!

What that says is, I can only tell you it's conclusive on a system of enough fidelity

That is exactly the point where we absolutely differ in opinion.

I can't tell you this test can be conclusive on just any system, that's correct.

Correct

You appear to really want this paper to be describing what system is necessary. He is telling you that on this specific system, it's conclusive.

No, I want to know why the system used is necessary. If you make the claim, you need to back it up.

3D: Width(1), depth(2), height(3)

I know what 3D is. I don't know how any of these parameters translate into what parameters of the system are relevant.

Can you?

You can, using proper experimentation.

Does he conclude to that?

No, he does not conclude that. He just infers it. THAT is one of the major problems.

If he did indirectly this is really a side note that you put a lot of weight to, I may have read fast and maybe some false claimed slipped trough.

I'm sorry, it's not a side note. Because this is exactly what the youtube and review crowd will go with. They will trump this paper as the definitive proof that you need an ultra-high-end system for height perception:

However, the present work is concerned with the 3D sound that unfolds naturally in HEA through a two‐loudspeaker setup and recordings made with two microphones without artificial signal manipulation. The key to sonic dimensionality is to maintain the highest signal integrity at every step in order to preserve the time‐domain fidelity to the subtlest detail.

Funny thing: The recording used is from a 1985 CD (the paper says 1984), and the actual recordings are from the mid-'70s. I bet you they were done on tape. It's totally silly to talk about time-domain fidelity of things like DAC's, amps, or interlinks with a source that is littered with time-domain issues (wow and flutter anyone?).

I don't know. are the factors you are referring to would have canceled the validity of this test? reinforced them? May I know what factors you have in mind? that question is pertinent, I don't unfortunately have the answer.

I did, re-read!

Good points, Yes this natural position of the instrument could have induced a bias.

What, now it's suddenly a good point? I literally brought it up in my post about the paper:

Well duh! A trumpet is a handheld device that you blow into. Usually, people that play the trumpet are standing up. Just by the fact that I hear I trumpet, my mind already knows that the sound is coming from higher up

A few posts down I even had some pointers on how to do an experiment to combat this bias.

 

[PeteL]

Thanks for taking the time to answer some of the questions Pete, I appreciate that.

It’s not. It’s about understanding your point of view. Isn’t that what these kind of forums are all about? To see what other people think about a subject and find out why that is. I think that is extremely valued, and that is why I keep asking questions. Before we agree to disagree, I need to know why you disagree.

How does this paper prove that it is not imagined? Can you point me to those parts of the text doing that?

Mkay..

I guess, in light of the above, that is not very much a surprise.

Of course I did.. reread the my findings.

No, not good enough!

He does not just describe the system. There is nothing intrinsically wrong with using such a system and describing the components. His claim is that such a system is necessary! Part of the "proof" is his bogus paper destroyed in Amir’s video. Even if that paper were true, that still does not prove it is necessary!

That is exactly the point where we absolutely differ in opinion.

Correct

No, I want to know why the system used is necessary. If you make the claim, you need to back it up.

I know what 3D is. I don't know how any of these parameters translate into what parameters of the system are relevant.

You can, using proper experimentation.

No, he does not conclude that. He just infers it. THAT is one of the major problems.

I'm sorry, it's not a side note. Because this is exactly what the youtube and review crowd will go with. They will trump this paper as the definitive proof that you need an ultra-high-end system for height perception:

Funny thing: The recording used is from a 1985 CD (the paper says 1984), and the actual recordings are from the mid-'70s. I bet you they were done on tape. It's totally silly to talk about time-domain fidelity of things like DAC's, amps, or interlinks with a source that is littered with time-domain issues (wow and flutter anyone?).


I did, re-read!

What, now it's suddenly a good point? I literally brought it up in my post about the paper:

A few posts down I even had some pointers on how to do an experiment to combat this bias.

OK Voodooless, apologies for missing your point about "trumpet is standing up", yes just rereading that it is a good point and I missed it. I guess that's the general tone of the whole piece that got me triggered and missed the fine prints.

Now we are still in disagreement with most of your rationale and that is what I pointed out in my comments.
So yes I'll settle for accepting that the experiment don't prove it. OK.

I won't comment on what the review crowd go with, heck trying to argue with the statements of one man is tiring enough. now what the whole "crowd" get from it...
You make it sound like it's a cult with everyone having one single way of thinking. I was commenting on the paper, and on your views not the views of a crowd.

Describing exactly what was the testing conditions are (the system) is important, if he didn't that would have been way more questionable, and still see absolutely nothing wrong with it. Yes this is the main thing we will not agree.

For the record I still do think that a "proper" sound system can deliver a 3D image. Yes you'll ask me what is proper and I'll tell you that I don't know still.
Alright, this paper is not a proof of that. So what I think is just what I think and not what I know. And I did state that for one I didn't think that the paper was perfect or that or it need to be, and I also stated that I can be wrong too.I am interested in knowing more about the mechanisms of it, it's an interesting subject. There's a issue with the music chosen, I'd like to read an other one.

So what do you think? There is no right or wrong, it's a discussion. Do you think a stereo system can deliver a 3D image or strictly depth and width? In the end it's an attempt to answer that. I think a stereo system can.
I like the Ideas about HRTF and binaural recording, and that maybe regular mics can't portray that but if blindly you can reliably perceive height in recorded material, regardless of how it was crafted, coming from only two point source, well, you can or you can't perceive it. I find it interesting, and it is the main question, not what system you need, this is an imperfect attempt at answering that, not conclusive enough.
It's a vast subject that is worth investigating.

 

[voodooless]

OK Voodooless, apologies for missing your point about "trumpet is standing up", yes just rereading that it is a good point and I missed it.

Great. Good that that finally came across. That is really the most important bit of the whole rebuttal, really.

So yes I'll settle for accepting that the experiment don't prove it. OK.

I won't comment on what the review crowd go with, heck trying to argue with the statements of one man is tiring enough. now what the whole "crowd" get from it...
You make it sound like it's a cult with everyone having one single way of thinking. I was commenting on the paper, and on your views not the views of a crowd.

I my view I factor in how other people will interpret what they read. Because the way things are worded in this paper is very deliberate.

Describing exactly what was the testing conditions are (the system) is important, if he didn't that would have been way more questionable, and still see absolutely nothing wrong with it. Yes this is the main thing we will not agree.

It is not. I have no problem with him describing the test conditions, and I already said as much. Far from it. What I object to is his reasoning that these components are in any way necessary for the experiment.

For the record I still do think that a "proper" sound system can deliver a 3D image. Yes you'll ask me what is proper and I'll tell you that I don't know still.

That's fine. The debate is about the method used in the experiment and the conclusions derived from it. These things are clearly invalid.

You don't have to answer the "proper sound system" question obviously. I wasn't expecting that I do however expect it (or something) from him. Because he makes the claim, and even cites his own paper as evidence.

I am interested in knowing more about the mechanisms of it, it's an interesting subject.

Most definitely

So what do you think? There is no right or wrong, it's a discussion. Do you think a stereo system can only deliver a 3D image or strictly depth and width? In the end it's an attempt to answer that. I think a stereo system can.

I don't know. And if it does, I don't know by what mechanism(s). The paper surely didn't bring us closer to an answer. An attempt yes, just a very bad one. Eventually, my position on the matter is of no consequence anyway. Me having a position doesn't bring me any closer to the truth. What is important is to look at what the paper tells us, and that is exactly nothing.

I like the Ideas about HRTF and binaural recording, and that maybe regular mics can't portray that but if blindly you can reliably perceive height in recorded material, regardless of how it was crafted, coming from only two point source, well, you can or you can't perceive it. I find it interesting, and it is the main question, not what system you need, this is an imperfect attempt at answering that, not conclusive enough.
It's a vast subject that is worth investigating.

Fully agree that the type of system needed is not the main question. But in the paper, it is a carefully crafted straw man built by the author to promote his high-end bullshit fetish. He does this in multiple papers, and it's a clear pattern with a clear motive.

 

[PierreV]

It's just a typical case of

Hey, I am a scientist in the queen of sciences.
This means I am more intelligent than 99.9% of the people who will read me.
This is what I believe _a priori_ and therefore this is what my paper will ultimately prove.
I need the right authoritative tone, some obfuscation the poor sods won't catch, and find a way to package whatever I find out in a way that looks meaningful.

It could work and remain unchallenged if Amir didn't deconstruct and monster for his huge echo chamber.

 

[PeteL]

Great. Good that that finally came across. That is really the most important bit of the whole rebuttal, really.



I my view I factor in how other people will interpret what they read. Because the way things are worded in this paper is very deliberate.

It is not. I have no problem with him describing the test conditions, and I already said as much. Far from it. What I object to is his reasoning that these components are in any way necessary for the experiment.

That's fine. The debate is about the method used in the experiment and the conclusions derived from it. These things are clearly invalid.

You don't have to answer the "proper sound system" question obviously. I wasn't expecting that I do however expect it (or something) from him. Because he makes the claim, and even cites his own paper as evidence.

Most definitely

I don't know. And if it does, I don't know by what mechanism(s). The paper surely didn't bring us closer to an answer. An attempt yes, just a very bad one. Eventually, my position on the matter is of no consequence anyway. Me having a position doesn't bring me any closer to the truth. What is important is to look at what the paper tells us, and that is exactly nothing.

Fully agree that the type of system needed is not the main question. But in the paper, it is a carefully crafted straw man built by the author to promote his high-end bullshit fetish. He does this in multiple papers, and it's a clear pattern with a clear motive.

Maybe what we can get from this, if I may dare... I went into this paper convinced that stereo speakers can portray height. That make me biased and prone to not catch some key problems with it. I think that how you've brought it up, and how you conclude, tell me that you went into it clearly questioning his intent, and that makes you also a bit biased on how you analyze the paper. My 2 cents.