[bigshot]

 

[71 dB]

2. No! HRTF stands for Head Related Transfer Function, NOT "spatial information limiting function"! Again, there is no way to limit the spatial information already in the mix and why would you want to, when the recording already contains less spatial information than intended? And, what has any of this got to do with your crossfeed and your assertions about your crossfeed? Your crossfeed does not have HRTFs, it's just a simple delayed crossfeed!

G

Yes. The amount of spatial information is not the issue. The issue is how large ILD it results. Depending on the system we use to listen to a given recording the resulting ILD can be large or small. With headphones ILD is easily excessive. A recording of one impulse of opposite polarity on left and right channel has huge channel difference, but zero spatial information. A recording if a church can have lots of spatial information, but the resulting ILD can be limited if the spatial information is coded that way.

Now HRTF makes sure ILD is not excessive. With headphones we "bypass" HRTF so we want something else to limit ILD. That something is for example crossfeed.

 

[pinnahertz]

Yes, but the mix has often more channel separation than what enters listeners ears, because while listening rooms increase spatial information, ILD is limited by HRTF compared to headhones.

Channel separation is excessive, not spatial information.

Listening speakers limit ILD, because of HRTF.

Confusion with terminology doesn't mean confusion or contradictions in substance.

Is this what we call "colorful language"?

 

[71 dB]

Is this what we call "colorful language"?

It's chromatic distortion free communication.

 

[pinnahertz]

How do you want me to test it? I test all new recording first without crossfeed to see if they belong to the infamous 2 % and if that's not the case, I find out proper crossfeed level. Having heard the non-crossfed version tells me how bad the problem is.

I find crossfeed very successful. I don't know what fails you are talking about. Crossfeed doesn't make everything better (only ~98 % imo), but that's why there's OFF switch and even I used it 2 % of the time. However, without crossfeed 98 % of material sounds worse to my ears. How is that a fail for crossfeed?

I'm sorry I can't make this as a real science because I don't have the resources. Those who have can make the science. I'm confident it will back up my claims.

So, the only thing we have here is opinions!

And you have just admitted this isn't real science. If you have no science to back up your opinions. If the principle of "if you want a different result do something different" holds, then why not try this: Rather than stating your opinion as scientific fact backed up by spatial hearing and math (because by your own admission, that's not true), state your opinions as what they are, opinions, personal preferences, etc.? I have no problem with that at all! In fact, I'll defend your right to state them, even if I disagree! Just so long as they stay framed as your personal preference and opinion, and not a mandatory religeous belief that everyone must adopt or be labeled a heretic.

When you say "This is the way it is, and this is the way it should be, has to be, you are all wrong!" , it puts you in a position of exposure to counter arguments because you condemn the opinions of others, and thereby condemn their right to express them. All you had to do was proclaim your preference and refrain from attacking anyone else by labelling them idiots, spatially deaf, spatially ignorant, or just plain ignorant, challenging the validity of what must certainly total up to nearly 100 years of combined professional experience in opposition to your opinion. Then you fight with us with your own made-up terminology, convoluted logic, and border-line name calling. 44 pages, and this could have all been done in about 2.

YOU chose to start the war, and now have the choice of continue to perpetuate it, or not. You won't even win cross-feed converts by behaving like a zealot, though.

And, while it may not mean much to you, look what you've accomplished. You've gotten me, a cross-feed non-fan, to start using it occasionally! Ok, only about 2% of the time, and only with serious tweaking, but that's a big deal.

You've said you want to learn. How about you figure out why many, including the professionals, don't generally like cross-feed? Or why anyone wouldn't like it? There MUST be a reason (besides they're all spatially-unenlightend idiots)! This will be a study of human sensory perception, not human spatial hearing, and may just turn on a light or two. Perception includes hearing and spatial hearing, but there's a whole lot more going on. Perception can be fooled. Perhaps there's some fooling going on with both polar opposite opinions of cross-feed?

If not then my years in the university have been complete waste of time.

And that statement makes no sense other than being purely emotional response. That's not science either. That's "I've tripped and fallen, my life is a complete waste." Chill, man. Get back on the horse.

 

[71 dB]

So, the only thing we have here is opinions!

And you have just admitted this isn't real science. If you have no science to back up your opinions. If the principle of "if you want a different result do something different" holds, then why not try this: Rather than stating your opinion as scientific fact backed up by spatial hearing and math (because by your own admission, that's not true), state your opinions as what they are, opinions, personal preferences, etc.? I have no problem with that at all! In fact, I'll defend your right to state them, even if I disagree! Just so long as they stay framed as your personal preference and opinion, and not a mandatory religeous belief that everyone must adopt or be labeled a heretic.

My claims can still be correct even if there's lack of scientific methodology behind them.

When you say "This is the way it is, and this is the way it should be, has to be, you are all wrong!" , it puts you in a position of exposure to counter arguments because you condemn the opinions of others, and thereby condemn their right to express them. All you had to do was proclaim your preference and refrain from attacking anyone else by labelling them idiots, spatially deaf, spatially ignorant, or just plain ignorant, challenging the validity of what must certainly total up to nearly 100 years of combined professional experience in opposition to your opinion. Then you fight with us with your own made-up terminology, convoluted logic, and border-line name calling. 44 pages, and this could have all been done in about 2.

I say ILD should be limited to a few decibels at low frequencies, because the ILD of HRTF contralateral/ipsilateral pairs at low frequencies are within a few decibels unless the sound source is VERY near to head. You can't refute HRTF, because those are scientifically measured, so you have to explain why my claim about "mimicing" HRTF in respect of ILD limiting is wrong. If you can explain that, I might learn something from you. Also, I'd be interested to learn why "larger than HRTF" ILD/ITD - values aren't a problem.

YOU chose to start the war, and now have the choice of continue to perpetuate it, or not. You won't even win cross-feed converts by behaving like a zealot, though.

Hopefully you notice I'm trying to be more respectful now. I was so surprised by the opposition to my claims I said nasty things, apparently calling people idiots without realizing it! That's not who I am, so all of this is confusing as hell.

And, while it may not mean much to you, look what you've accomplished. You've gotten me, a cross-feed non-fan, to start using it occasionally! Ok, only about 2% of the time, and only with serious tweaking, but that's a big deal.

That's better than nothing. Appreciate it! I'm not demanding anyone to use it 98 % of the time. That's just how it's for me.

You've said you want to learn. How about you figure out why many, including the professionals, don't generally like cross-feed? Or why anyone wouldn't like it? There MUST be a reason (besides they're all spatially-unenlightend idiots)! This will be a study of human sensory perception, not human spatial hearing, and may just turn on a light or two. Perception includes hearing and spatial hearing, but there's a whole lot more going on. Perception can be fooled. Perhaps there's some fooling going on with both polar opposite opinions of cross-feed?

Even if "crossfeeders" are a minority, some people really "love" crossfeed. We have several members in this thread stating that. So, for some reason some people really like crossfeed, find it more natural, less tiring, "bee free", etc. I believe a lot of people don't like crossfeed because of human nature. It's difficult to "let go" the way you have done things. I was willing to condemn my listening years without crossfeed "wrong" and accept crossfeed as the superior way of listening headphones. I think many find the reduction of effects caused by excessive ILD/ITD loss of spatial detail. I don't. To me effects caused by excessive ILD/ITD are not real information and those effects don't exist when listening to speakers because HRTF limits ILD/ITD. To me headphone listening without crossfeed requires superspatial hearing so that excessive ILD/ITD values make sense. Turning crossfeed ON reduces perceptual loudness, removes effects caused by excessive ILD/ITD making the sound less "impressive" and changes the nature of spatiality. I can understand why some people may have a problem with that, especially when some of the benefits of crossfeed (less tiring) became apparent only after a longer listening session. It has been my believe that explaining people what really happens when you crossfeed would help them to accept crossfeed and hear the benefits better, but these belief have been challenged to extreme levels on this discussion board to my surprise. Go figure.

In my opinion sound engineers shouldn't be offended by crossfeed. I don't think it messes up at all the spatial information painstakingly worked into a mix. In fact, I believe proper crossfeed allows all that spatial information to really come across undistorted in all glory.

 

[pinnahertz]

My claims can still be correct even if there's lack of scientific methodology behind them.

Could be, but you actually have no idea. Evidence and opinion may not be in your favor.

I say ILD should be limited to a few decibels at low frequencies, because the ILD of HRTF contralateral/ipsilateral pairs at low frequencies are within a few decibels unless the sound source is VERY near to head.

Under what conditions? With what material? With what mix? Did you forget about all those variables?

You can't refute HRTF, because those are scientifically measured, so you have to explain why my claim about "mimicing" HRTF in respect of ILD limiting is wrong. If you can explain that, I might learn something from you. Also, I'd be interested to learn why "larger than HRTF" ILD/ITD - values aren't a problem.

OK, but I don't think you'll learn anything. Ready?

Nobody here has "refuted HRTF"!

I'm not claiming that "mimicking HRTF" with ILD an low frequencies is, of itself "wrong". I'm challenging the universal efficacy and application of cross-feed as a "correction" of a problem that may not be subjectively necessary to correct.

"Larger than HRTF" ILD and ITD may or may not be a problem depending on what else is going on in the mix and what the artistic intent is. Because of that, blanket ILD correction stands more chance of being wrong than right, but the key is the subjective opinion of how it sounds. You cannot calculate opinion. Opinion is rooted in perception. You have not factored in perception.

Learn anything? I didn't think so.

Hopefully you notice I'm trying to be more respectful now. I was so surprised by the opposition to my claims I said nasty things, apparently calling people idiots without realizing it! That's not who I am, so all of this is confusing as hell.

Any attempt at respect will be appreciated.

That's better than nothing. Appreciate it! I'm not demanding anyone to use it 98 % of the time. That's just how it's for me.

You say that now, but you did, and have made it quite clear that anyone with another view must be spatially deaf/ignorant/unenlightened, blah..blah..blah. And more than once.

Even if "crossfeeders" are a minority, some people really "love" crossfeed. We have several members in this thread stating that. So, for some reason some people really like crossfeed, find it more natural, less tiring, "bee free", etc.

Not discounting what you said, but please realize you have essentially zero data about cross-feed acceptance and desire. You don't have a statistically valid sample. You cannot draw any valid conclusions from the above.

I believe a lot of people don't like crossfeed because of human nature. It's difficult to "let go" the way you have done things. I was willing to condemn my listening years without crossfeed "wrong" and accept crossfeed as the superior way of listening headphones.

Your belief, stated above, is functioning as a strong bias against your own research. Even if you cannot perform all of the scientific research you'd like to, you could at least employ a basic scientific attitude. However, you have within your power, abilities and budget the capability to do the necessary research, it's just not been your focus. Attributing rejection of cross-feed to "difficulty of letting go" of a way of doing something is a possibility, but certainly not the only one. That, too, could be easily researched.

I think many find the reduction of effects caused by excessive ILD/ITD loss of spatial detail. I don't. To me effects caused by excessive ILD/ITD are not real information and those effects don't exist when listening to speakers because HRTF limits ILD/ITD.

The disparity above would motivate a scientist to find out why it exists.

To me headphone listening without crossfeed requires superspatial hearing so that excessive ILD/ITD values make sense. Turning crossfeed ON reduces perceptual loudness, removes effects caused by excessive ILD/ITD making the sound less "impressive" and changes the nature of spatiality.

You are re-stating your preference. You'll get no validation here. We get your preference, but your bias is so strong you can't even open the possibility of researching the reason why it's not universally shared. It's the problem 3D movie producers have had for years, but now have come to an understanding of it. 3D should, theoretically, be seen universally as better, but it's not. In fact, it's acceptance ranges from a few that seek it out and prefer 3D, all the way to those that abhor it. That's why all 3D movies released today are also released flat. But, today, the reason for this is well known. You have a similar problem, but the reason for the wide variance in acceptance of cross-feed is not known.

I can understand why some people may have a problem with that, especially when some of the benefits of crossfeed (less tiring) became apparent only after a longer listening session. It has been my believe that explaining people what really happens when you crossfeed would help them to accept crossfeed and hear the benefits better, but these belief have been challenged to extreme levels on this discussion board to my surprise. Go figure.

No, what you've stated is a hypothesis. It's also illogical. You can explain 3D movies all day, it won't build audience acceptance. That's been more than proven. You need to stop trying to hypothesize and convince, and start researching why that's not working.

In my opinion sound engineers shouldn't be offended by crossfeed. I don't think it messes up at all the spatial information painstakingly worked into a mix. In fact, I believe proper crossfeed allows all that spatial information to really come across undistorted in all glory.

I accept, but disagree with, your opinion. Here's what you need to research:

Why do some think cross-feed collapses the dimensional and spatial nature of some recordings?
Why, even though you think it is less tiring, others find cross-feed immediately annoying?
Why, when you find it applicable to 98% of all stereo music, others find anywhere from partial agreement to an inverse statistic?

Re-stating opinions and hypotheses without finding proof and testing hypothesis is just wasting bandwidth at this point. You've had several beliefs challenged here. Do you still hold them as absolute, or are you willing to dig deeper into human sensory perception?

And quit arguing?

 

[bigshot]

IBID above

 

[gregorio]

It's good to see that you seem to be coming around, that you've now contradicted and thereby apparently withdrawn a number of your claims; being scientific and some of your claims about spatial information for example. There's still a few problems/misconceptions though:

[1] I test all new recording first without crossfeed to see if they belong to the infamous 2 % ...
[2] I find crossfeed very successful. I don't know what fails you are talking about.

1. It's not an infamous or famous 2%, I'd never even heard of this figure before you came out with it. It's a figure you've invented which correlates with your personal perception/preference for crossfeed, maybe it's a figure which is applicable to some other people, maybe there are some who think the figure should be 100% or 0% or in fact anywhere in between. You do not know, there is little actual, scientific evidence and, we have to be careful because it's not fixed, it varies with individual recordings. As I've already stated, I am not a fan of crossfeed but there are some recordings I've tested which I feel do benefit from crossfeed but my "infamous" figure would be almost the inverse of yours, probably only a few percent which benefit. However, I don't know exactly what music you listen to, maybe if I had your collection of music recordings my figure would increase significantly and maybe if you had my collection your figure would decrease significantly, although the huge disparity between us strongly suggests a significant difference of perception/preference and not solely a difference in our music collections. And of course, you've claimed 98% of ALL recordings benefit from crossfeed, not just 98% of the recordings you own/listen to.

2. That is clearly untrue! What's your "infamous 2%" then? If you don't know of any "fails" why don't you quote crossfeed as being beneficial 100% of the time, instead of only 98%? So you do know what fails pinnahertz is talking about! The differences here are: 1. You're talking about only 2% of recordings fail and we're talking about a very significantly higher percentage and 2. You've stated your 2% as indisputable fact for ALL music and ALL people, which you can't, it's just what you've found with your perception/preference, it's your opinion and NOT fact. I'm stating that your 2% is more like 95% but I'm not stating that 95% figure is an "infamous" fact, applicable to all recordings and all people, just applicable to me and my perception/preferences and 3. You have stated that anyone who disagrees with you is ignorant/an idiot and there's obviously a massive difference on this point between us but fortunately, you're starting to back away from this claim!

[1] Yes. The amount of spatial information is not the issue.
[2] The issue is how large ILD it results. Depending on the system we use to listen to a given recording the resulting ILD can be large or small. With headphones ILD is easily excessive.
[3] A recording of one impulse of opposite polarity on left and right channel has huge channel difference, but zero spatial information.
[3a] A recording if a church can have lots of spatial information, but the resulting ILD can be limited if the spatial information is coded that way.
[4] Now HRTF makes sure ILD is not excessive. With headphones we "bypass" HRTF so we want something else to limit ILD. That something is for example crossfeed.

1. Now that statement is good ... or rather, it's not "good", it's just an improvement. It's an improvement because it withdraws many of your more ridiculous claims about spatial information and indicates that the issue of incorrect terminology is now not so much of an issue, which means we might be able to have a more rational discussion. It's "not good" though, because spatial information is in fact the issue or rather one of the issues!

2. ILD is AN issue with HP presentation and crossfeed does indeed solve that issue. But, ILD is not "The issue", it's only one of the issues of the channel separation which occurs with HP presentation and while crossfeeding solves the issue of ILD, it does not solve and indeed often makes worse, some of those other issues, one of which is spatial information!

3. We may have another incorrect use of terminology here. One impulse of "opposite polarity on left and right channel" would have zero spatial information, zero level difference and the only difference between the channels would be 180deg of phase. Given a perfect stereo system and the listener perfectly positioned between the speakers, the result would be complete silence! Do you mean an impulse panned hard left and THEN panned hard right (or vice versa), what is often referred to as "ping pong"? But then, as panning is one of the aspects of spatial information, you cannot have panning AND zero spatial information. So, do you mean "ping pong" and no other spatial information except for panning (no EQ or volume differences and no reflections/reverb for example)? If I assume this is what you mean, then how many commercial music recordings are there which comply with these conditions? Either none at all or almost none at all! Although admittedly, some early, very basic stereo mixes get somewhat close to these conditions.
3a. A recording made in a church will, virtually without exception, have a great deal of spatial information in terms of reflections and most probably other aspects of reverb. Reverb will reduce ILD but the amount of ILD in the completed mix will depend on how the mix was originally recorded, how the mics are panned during mixing and the exact nature and amount of the reverb.

4. Yes, agreed but there are two issues here: 1. Your crossfeed is NOT a HRTF, it only accounts for part of a HRTF! and 2. When crossfeeding you are not crossfeeding ONLY level differences, you are crossfeeding the entire signal (or in your case, if I've understood correctly, the entire signal below 1kHz). This entire signal contains not only level but obviously also frequency and all the other spatial information (reflections/reverb). Crossfeeding CAN therefore cause undesirable frequency interactions and WILL affect the timing upon which reflections/reverb depends. A problem when we take into account the fact that the reflections/reverb in the final product is a manufactured illusion rather than a pure/real/natural occurrence. Crossfeeding does therefore affect this illusion, although how noticeably depends on the reflections/reverb in the final mix and of course, how sensitive our individual perception is. My experience is similar to pinnahertz, for me the illusion often breaks down significantly, I loose depth, ambience and it all sounds rather 2 dimensional/flat. Maybe others are not so sensitive and either simply don't notice this effect at all or do notice it but not enough to be particularly bothered by it and therefore to them, crossfeeding is either all good or good most of the time.

G

 

[71 dB]

That is clearly untrue! What's your "infamous 2%" then? If you don't know of any "fails" why don't you quote crossfeed as being beneficial 100% of the time, instead of only 98%? So you do know what fails pinnahertz is talking about! The differences here are: 1. You're talking about only 2% of recordings fail and we're talking about a very significantly higher percentage and 2. You've stated your 2% as indisputable fact for ALL music and ALL people, which you can't, it's just what you've found with your perception/preference, it's your opinion and NOT fact. I'm stating that your 2% is more like 95% but I'm not stating that 95% figure is an "infamous" fact, applicable to all recordings and all people, just applicable to me and my perception/preferences and 3. You have stated that anyone who disagrees with you is ignorant/an idiot and there's obviously a massive difference on this point between us but fortunately, you're starting to back away from this claim!

The "2 %" are recordings without excessive channel difference. They don't cause excessive ILD so nothing needs to be done, no crossfeed = proper crossfeed. Crossfeed being successful means it is successful in situation when there is excessive channel difference to be fixed. Umbrellas being successful on rainy days doesn't mean people should use umbrellas all the time. Umbrellas protect from the rain, so you need rain to have a need for umbrellas. To me crossfeed is the "umbrella" for excessive channel difference and it rains excessive channel difference 98 % of the time. Using crossfeed when you don't need it or using too strong crossfeed limits ILD values under the "full potential" and make the sound unnecessorily mono-like. Proper crossfeed limits ILD so that the largest ILD values are the largest that human spatial hearing expects to experience naturally. At low frequences that value for ILD is about 3 dB. So, it max LF ILD of a recording is 3 dB or less you don't perhaps need crossfeed. If max LF ILD = 5 dB, you crossfeed a little. If max LF ILD = 12 dB, you crossfeed a lot. The higher frequencies up to about 1.6 kHz have a say on this too, but that's the general idea. Bass ILD can be limited to near mono, because at those frequencies spatial hearing is based almost completely on ITD or you can have mono bass and have all the spatiality on higher frequencies ( > 200 Hz ).

1. It's not "fail", because OFF is one of the crossfeed levels for situations when excessive channel difference doesn't exist such as binaural recordings.
2. The 2 % is not an indisputable fact. It is what seems to agree with my listening experiences.
3. I take back the idiot part. I don't know why I ever wrote that! I still feel ignorance might play a part for some people. At least it played part for me before 2012.

 

[71 dB]

3. We may have another incorrect use of terminology here. One impulse of "opposite polarity on left and right channel" would have zero spatial information, zero level difference and the only difference between the channels would be 180deg of phase. Given a perfect stereo system and the listener perfectly positioned between the speakers, the result would be complete silence! Do you mean an impulse panned hard left and THEN panned hard right (or vice versa), what is often referred to as "ping pong"? But then, as panning is one of the aspects of spatial information, you cannot have panning AND zero spatial information. So, do you mean "ping pong" and no other spatial information except for panning (no EQ or volume differences and no reflections/reverb for example)? If I assume this is what you mean, then how many commercial music recordings are there which comply with these conditions? Either none at all or almost none at all! Although admittedly, some early, very basic stereo mixes get somewhat close to these conditions.
3a. A recording made in a church will, virtually without exception, have a great deal of spatial information in terms of reflections and most probably other aspects of reverb. Reverb will reduce ILD but the amount of ILD in the completed mix will depend on how the mix was originally recorded, how the mics are panned during mixing and the exact nature and amount of the reverb.

3. Zero difference of absolute values , but ears don't take absolute values when figuring out ILD. The resulting sound with perfect stereo system wouldn't be silence! In anechoic chamber left ear would hear positive impulse from left speaker convoluted with the HRIR associated with that angle of sound + negative impulse convoluted with another HRIR associated with that angle of right speaker. The result is HRIR(from left) - HRIR(from right) which is not zero, because HRIR(from left) is not same as HRIR(from right). The resulting sound is an impulse-like positive spike followed by less impulse-like strongly low-pass filtered negative spike about 0.2-0.3 ms later. Right ear would hear the same, but polarity reversed. In normal living room we'd have also early reflections and reverberation to add lots of spatial information. We would hear kind of 2 versions of the impulse response of the room playing simultaneously. Far from silence.

3a. Exactly! 100 % agreed! I could have written that myself. Now, that is why we have recordings with different amount of ILD and why we need different crossfeed levels to address those differences. Also, what you wrote tells us that we always could choose such mic set ups and positions that in a given acoustics the resulting ILD is optimal or at least within the limits of spatial hearing.

 

[71 dB]

Do you mean an impulse panned hard left and THEN panned hard right (or vice versa), what is often referred to as "ping pong"? But then, as panning is one of the aspects of spatial information, you cannot have panning AND zero spatial information. So, do you mean "ping pong" and no other spatial information except for panning (no EQ or volume differences and no reflections/reverb for example)? If I assume this is what you mean, then how many commercial music recordings are there which comply with these conditions? Either none at all or almost none at all! Although admittedly, some early, very basic stereo mixes get somewhat close to these conditions.

G

I mean you have first mono impulse (same impulse at both channels at same amplitude and timing) and then you reverse right channel. For such signal L+R = 0L and L-R = 2L. That is not ping pong and you don't get such signal with amplitude panoration. Such signal doesn't have spatial information (stereo-sense), because it's not mono (sound is not centered), and sound is not in the right or left either, because we have same absolute amplitude left and right, just in opposite polarity. So in stereo-sense there is no spatial information. In "pro-logic" multichannel-sense there is spatial information, because such signal would be decoded to rear channel, but we are talking about stereophonic sound here.

 

[gregorio]

3.The resulting sound is an impulse-like positive spike followed by less impulse-like strongly low-pass filtered negative spike about 0.2-0.3 ms later. Right ear would hear the same, but polarity reversed. In normal living room we'd have also early reflections and reverberation to add lots of spatial information. We would hear kind of 2 versions of the impulse response of the room playing simultaneously. Far from silence.
3a. Exactly! 100 % agreed! I could have written that myself. Now, that is why we have recordings with different amount of ILD and why we need different crossfeed levels to address those differences. [3b] Also, what you wrote tells us that we always could choose such mic set ups and positions that in a given acoustics the resulting ILD is optimal or at least within the limits of spatial hearing.

3. Have you actually tried this in practise? It takes some messing about because of room acoustics which causes the two speakers to never have exactly the same response but with some messing you can indeed experience silence. I've done this with students, although in a studio, never tried it in a home environment.
3a. I explained that crossfeed does solve the different levels issue but using crossfeed to solve this issue causes other, undesirable issues.
3b. Yes we could, it's not even very difficult. However, in practise "optimal ILD" is not the only concern, it's not even the primary concern. There are other more important concerns!

[1] I mean you have first mono impulse (same impulse at both channels at same amplitude and timing) and then [2] you reverse right channel. For such signal L+R = 0L and L-R = 2L. That is not ping pong and you don't get such signal with amplitude panoration. Such signal doesn't have spatial information (stereo-sense), because it's not mono (sound is not centered), and sound is not in the right or left either, because we have same absolute amplitude left and right, just in opposite polarity.

1. Such a signal DOES have spatial information, the sound will appear at the phantom centre position, this is effectively exactly what panning is! There is no actual centre position in a (2 channel) stereo system only a left and a right. To pan a sound to the centre position results in effectively exactly what you describe, the exact same signal in both the left and right channels at the exact same time with exactly the same amplitude. As one changes the relative level so the sound moves from the centre position and this is called "panning". As one changes the relative timing of the two signals the sound also moves from the centre position and this is called "psychoacoustic panning".
2. When you phase invert one channel relative to the other you get phase cancellation, which is silence, as described previously. So, assuming the signal is identical in both channels, L+R = C (phantom) and L-R = 0 (silence). This is real basic stuff, I can't believe you don't know it, maybe there is some confusion of terms again?

G

 

[pinnahertz]

I mean you have first mono impulse (same impulse at both channels at same amplitude and timing) and then you reverse right channel. For such signal L+R = 0L and L-R = 2L. That is not ping pong and you don't get such signal with amplitude panoration. Such signal doesn't have spatial information (stereo-sense), because it's not mono (sound is not centered), and sound is not in the right or left either, because we have same absolute amplitude left and right, just in opposite polarity. So in stereo-sense there is no spatial information. In "pro-logic" multichannel-sense there is spatial information, because such signal would be decoded to rear channel, but we are talking about stereophonic sound here.

You have chosen a bad example. Impulses cannot and do not exist in practical situation because there’s no way to produce them. The theoretical impulse becomes band limited, phase distorted, by every transducer on earth.

But even if you chose a mono sound signal, noise burst, pretty much anything, the perception of a mono but out of phase signal, presented equally from two speakers, actually is one of very high spatiality! It is perceived as a diffuse and directionless sound (there are test records and CDs that describe it this way) that appears to not come from dead center but all over. Because acoustic reverb also has some of this quality, which you can easily confirm by looking at the content an L-R sum, we associate an out of phase mono signal with a sense of space. When experimenting with acoustic crosstalk cancellation I was surprised that one of the problems was when you cancel acoustic crosstalk with speakers, some recordings presented an unusually high amount of reverb, out of balance direct signal.

What you say about an out of phase signal is wrong, as far as human perception goes, and frankly that’s all that matters.

 

[71 dB]

1. Could be, but you actually have no idea. Evidence and opinion may not be in your favor.

2. Under what conditions? With what material? With what mix? Did you forget about all those variables?

1. I have some idea. If I hadn't I wouldn't be making claims here. Education/knowledge of spatial hearing + listening experiences give me an idea, just not verified scientifically. I have mentioned I have low self-esteem which means I keep my mouth shut about things I know/feel I don't know much. This is something I am very confident about so I make claims online.

2. Under any conditions and with any material/mix.

OK, but I don't think you'll learn anything. Ready?

Nobody here has "refuted HRTF"!

I'm not claiming that "mimicking HRTF" with ILD an low frequencies is, of itself "wrong". I'm challenging the universal efficacy and application of cross-feed as a "correction" of a problem that may not be subjectively necessary to correct.

"Larger than HRTF" ILD and ITD may or may not be a problem depending on what else is going on in the mix and what the artistic intent is. Because of that, blanket ILD correction stands more chance of being wrong than right, but the key is the subjective opinion of how it sounds. You cannot calculate opinion. Opinion is rooted in perception. You have not factored in perception.

Audio science makes little sense if we keep insisting subjectivity. We should try to find objective truth. Having "larger than HRTF" ILD/ITD as your artistic intent is kind of a bad intent, because you can't have "larger than HRTF" values with speakers, only with headphones. So you are doing something that will sound "against your intent" on speakers. Such productions should have a sticker on them saying "For headphones only. Don't use crossfeed." The only recordings with such stickers are binaural recordings and they are especially free of any "larger than HRTF" values for ILD and ITD. Another point is that large ILD is related to sounds very near head/other ear and causes an annoying feeling so that we know to use our hands to swipe the damn bee before it flies into the ear canal! Smaller ILD is more comfortable to listen to.