To crossfeed or not to crossfeed? That is the question...

[71 dB]

Please cite a reference (other than your own).

No, I don't. I'm not here to do what you tell me to do. If you don't take my opinions serioustly then don't. I'm beginning to think you are a tormentor. I have every right to come up with my own new terminology. For normal people explaining the meaning is enough. Not to you apparently.

You are one opinion, and you have never once referenced anything that shows your opinion is widely held, or even held by anyone but you. If crossfeed were so good, so essential, such a key function, then why hasn't it been standard on even 1% of all music players since the original Sony Soundabout (Walkman) of 1979? It certainly could have been done and very low cost. But it wasn't, hasn't, and isn't still. "Loudness Compensation" had a far better market penetration, and that didn't work well either!

That's the damn problem! Products aren't as good as they could be, because customers are ignorant and manufacturers offer what customers want and customers become even more ignorant and the spiral into stupidity is ready. Crossfeed should have been EVERYWHERE from day one. Even I was totally ignorant of crossfeed until 2012 when I was 41! People need to be educated!

You are delusional if you think most recordings have adequate ILD for headphones. They are mixed for speakers, not headphones!

You can't reduce room modes with pre-filter, and I'm sure you know that. You can only reduce the results of room modes. Room modes can only be reduced with physical means.

Does a room mode exist in silence? It is a philosophical question… …anyway, result is what matters.

And exactly how many stereo recordings were released in 1958? What percentage of the total releases ever does that make up? Wouldn't be reasonable to expect early stereo mixes to have a few issues until we learned how to handle the new medium? You've "cherry-picked" an example, which is thus meaningless.

If you think they were stunning binaural recordings you are delusional. The thing is it's not only 1958. King Crimson's 70's stuff has insane separation too. Tangerine Dream too. No matter what it is I need crossfeed. Old new whatever. That's how it is. Not cherry picking at all. This stuff was made for speakers. Miles Davis, Cato Barbieri, Herbie Hancock. Rose Royce. Carly Simon. Always too much ILD for headphones. Of my 1500 CDs maybe 3 dozen work best without crossfeed.

We have these "strange fights" because your science doesn't apply to reality, it's specifically targeting contrived conditions.

Seems to apply to reality pretty well, because it is created based on observations of reality.

However, my objects are, and have been:

1. Promoting headphone cross-feed as if it were a "compensation" for some form of distortion, and as if it were a universal solution. Such is not the case. The application of cross-feed is highly generalized at best, inappropriate at least, and never compensates properly for any speaker-mix condition because you have no idea what the intentions were in the first place, nor the precise monitoring conditions used in the mix. You advance it as if it were a complimentary equalizer (like RIAA) when it is at very best a coarse approximation based on uneducated assumptions.

2. You authenticate your "science" with your own passionate opinion, but offer no other statistical listener preference for crossfeed.

3. Your opinions are (still!) stated as immutable fact. And when others express their opinions you view them as personal attacks.

4. Your "science" is tightly targeted at a narrow set of conditions, and ignores the facts that actual mixes are as much art as science. There's no "correcting" for "art".

1. Yes, normal crossfeed is an approximation of the acoustic crossfeed that happens with speakers. Acoustic crossfeeder isn't "perfect" either unless you listen to the recording in the same studio it was mixed in. We have to do with less than perfect solutions in real life. As such, crossfeed is amazingly good in reducing ILD/ITD and making headphone sound more natural and less fatiqueing. Crossfeed doesn't give 100 % what was intended in the studio, but it gives close-enough version of it that is pleasant and enjoyable. That is what counts.

2. Ask Andy Linkner (Andolink). I built him a crossfeeder in 2014. Also, if you read discussion boards about crossfeed you'll see how many use crossfeed "all the time" just like me. So, your attempt to make me look a singular fool fails.

3. My "facts" may not be 100 % correct (understanding is refined every day), but I have put effort in this so much, that if I was wrong it would be really strange. It would be like saying Einstein didn't understand relativity. I take your posts as personal attacks because instead of demostrating my faults you try to discredit me. You have managed to correct me once (vinyl stylus movements and L+R, L-R). I really was mistaken that time and remembered the directions wrong. Mostly you just point out that your opinions differ from mine. I don't know about you, but I try to build my opinions on scientifically sound premises and finetune them if needed.

4. People correct "art" with acoustic crossfeed all the time and nobody cares. Room, speakers and listening position is a complex acoustic system, much more unpredictable than headphone crossfeed.

 

[71 dB]

1. No it does not. ILD is not the sole proximity cue! Your HRTF model may work in an anechoic space, but doesn't account for a real room of any size or reflective nature. In other words, your HRTF alone doesn't model spatial hearing in real life.

2. Everyone please note the emphasized text. It is exactly what it says: opinion. That means there will be other conflicting opinions! You want fact? Collect statistics, don't quote your own opinion! The bulk of modern music producers and engineers disagree with you.

3. No, not splitting hairs at all. In fact, you are correct that modes create 10-15 dB peaks, and dips as deep as 30 dB or more. You are incorrect in that reverberation below 200Hz causes modes to become denser. Simply not true, and if you'd measure a few small rooms, you'd know that. Those deep dips are not correctable because room EQ systems must limit gain to only a few dB (Audyssey's gain limit is 9dB, for example) because of what that kind of gain does to amplifier power requirements and speaker power handling. If you'd bother to do a few real-room measurements you'd see radical dips that are very, very location specific. In fact, that's why room EQ can only be properly done by averaging many measurement points. However, ears sit in single positions, and therefore are subject to some rather deep frequency specific notches.

4. You really should get out and measure like two dozen rooms and see what reality is like. No, those are not pathological situations, they are typical of single-sub rooms. Yes, we do try to fix them with treatment and multiple subs, but if you're working with two-speaker stereo, you cannot move your chair much! Sub-200Hz dips occupy real estate, and your chair can only be moved along the center-line. If you move a speaker, you'll also move the center line, but you can't move a speaker enough to mitigate room mode dips at the optimum LP! Sorry, expericence shows the grim reality. And that's just another reason why multiple-sub, multichannel audio wins hands down.

5. But most bass is mixed 0dB channel difference (center), and with good reason. There is absolutely no performance or perspective advantage to mixing bass 3dB off center in stereo.

1. Of course not, but it's the cue that most of the time goes wrong in headphone listening so it's the one that we want to fix with crossfeed.

2. You are the one who wanted me to use that word. Mono bass is easy. You just make your bass tracks mono. Some pop music has mono bass, some has large ILD so it's up to the producer what they do. I tend to have some ILD on my bass when making music. Depends on how lively or calm bass I want. That is my opinion. You have yours.

3. Standing waves become denser with the frequency and in typical room so dense that the modes form reverberation around 150-200 Hz. Also, at higher frequencies the peaks an dips aren't that strong, because the acoustic losses increase.

4. Yes, the reality is bad. I agree. Of course multichannel wins.

5. Most bass? That's not my experience. Some producers such as Dr. "raper" Luke do favor mono bass, but not all producers. In room you do have some ILD at bass. You even used room modes to demonstrate greater than 3 dB ILD, so I can see advantages to mixing bass at 3 dB to have some sense of acoustic environment. In free field about 5 dB ILD at bass is all you get for sounds not very close to head. In a room you have diffuse sound field, so the reasonable level of ILD is between 0 and 5 dB and that's when 3 dB show up. You can have a little less or a little more, but that's your "default" level. I am really done with explaining this. If you continue not getting my point, I have no choice but to consider you dumb.

Cross-feed changes stereo separation, that's it. Whether or not that's an improvement (a "fix") is entirely subjective, and varies with every recording from improvement all the way to detriment.

Have I said otherwise? It's the same kind of improvement that acoustic crossfeed does. It also changes stereo separation so that a ping pong stereo recording with huge separation becomes something with much smaller separation in our ears. The biology of our hearing suggests that reducing separation to natural levels should be subjectively an improvement, but of course individuals can hold twisted conceptions, because we are human beings, not machines.

You have presented no data to support the premise that cross-feed is either universally desired, or perceived as an improvement at all, much less that it is over the (claimed) majority of recordings. None. Only one person (your) opinion.

Crossfeed is not universally desired because people are too ignorant to understand it or even know about it. People are pretty mindless consumers brainwashed by global companies to buy their products. It's easier to Fidget Spinners than crossfeeders. What problem does Fidget Spinner fix? People like you make sure crossfeed will never become popular. There is no hope. Thanks a lot.

 

[71 dB]

1. I do know that, but you've never backed that number up with data. It's just your opinion, that's a statistic of 1. Do you realize what the margin of error on that is?

2. That's not a "lot", that's One. How many stereo recordings were released in 1958? What percentage of the total of all stereo recordings is that?

1. I read that number online somewhere and have no reason to disagree so it's a statistic of 2. Your own numbers are a statistic of 1. I win. Seriously, this is silly, but you want to play this game. It doesn't matter how many people agree with me or you. 7 billion people can be wrong and in many things is. Sometimes you are right alone before rest of the world catches up.

2. What difference would it make to list all of them? All people except you know that the stereo separation was often huge back then, because stereo was new and the stereo effect was maximized in a naive manner just like all new technology at first. Do you think they thought about headphones in 1958? No, they did not. All they thought about was to boost sales with stereo sound. Not much different from high-res downloads today.

 

[jgazal]

No, I don't. I'm not here to do what you tell me to do. If you don't take my opinions serioustly then don't. I'm beginning to think you are a tormentor. I have every right to come up with my own new terminology. For normal people explaining the meaning is enough. Not to you apparently.

He does that because we are in sound science forum and usually most descriptors were already agreed by previous researchers so that the newcomers don’t need to reinvent the wheel.

Mr. pinnaherz maybe. Don't know about Professor Choueiri, who is talking about speaker audio and how to get "headphone binaural" sound with speakers using "anti-crossfeed", crosstalk canceling. That's completely different from what I'm talking about.

You are delusional if you think most recordings have adequate ILD for headphones. They are mixed for speakers, not headphones!

1. Yes, normal crossfeed is an approximation of the acoustic crossfeed that happens with speakers. Acoustic crossfeeder isn't "perfect" either unless you listen to the recording in the same studio it was mixed in. We have to do with less than perfect solutions in real life. As such, crossfeed is amazingly good in reducing ILD/ITD and making headphone sound more natural and less fatiqueing. Crossfeed doesn't give 100 % what was intended in the studio, but it gives close-enough version of it that is pleasant and enjoyable. That is what counts.

I am failing to see how why crosstalk is bad for speakers and good for headphones.

With speakers you add your own HRTF and the real time filtering effect of your head movement. Professor Choueiri goes further and eliminates crosstalk.

With headphones you loose the head and torso component of your HRTF and add some filtering from the headphones itself.

But adding crossfeed does not bring back your HRTF or real time filtering effect of your head movement. And most of the time ILD for each mixed instrument/track are already blended in the master.

Have I said otherwise? It's the same kind of improvement that acoustic crossfeed does. It also changes stereo separation so that a ping pong stereo recording with huge separation becomes something with much smaller separation in our ears. The biology of our hearing suggests that reducing separation to natural levels should be subjectively an improvement, but of course individuals can hold twisted conceptions, because we are human beings, not machines.

Crossfeed is not universally desired because people are too ignorant to understand it or even know about it. (...) People like you make sure crossfeed will never become popular. There is no hope. Thanks a lot.

All people except you know that the stereo separation was often huge back then, because stereo was new and the stereo effect was maximized in a naive manner just like all new technology at first. Do you think they thought about headphones in 1958? No, they did not. All they thought about was to boost sales with stereo sound. Not much different from high-res downloads today.

I see your point with “ping pong stereo recording” (do you really believe that 98% of musical recordings available nowadays have such strong stereo separation?).

But perhaps your emotional involvement with your previous efforts with crossfeed, a hard work that deserves our highest respect, may have given you a bias to extrapolate its supposed improvement to 98% of stereo recordings.

I see that your desire is to further increase the crossfeed even with modern mixes, but I still fail to see why that would be beneficial instead of neutral or even detrimental.

So I guess pinnhertz is not fighting you our personally attacking you, but just trying to do the same thing you believe you are doing, which is educating consumers.

The beauty is that I cannot know for sure who is right, because there are too many variables. But every time you both argue I see a new subtlety I was failing to see previously. So don’t feel bad with people disagreeing.

 

[RRod]

Very interesting question!

Do you mean:

1. 5 speakers in a row of 2 meters parallel to the coronal plane of the dummy head.
2. Five sweeps/chirps from 20hz to 20khz are played one after another in each speaker and recorded by the dummy head (binaural master reference audio file).
3. Five sweeps/chirps from 20hz to 20khz are played one after another in each speaker and recorded by the an ORTF microphone pattern (ORTF master reference audio file).
4. The two farther speakers in the same row of 2 meters parallel to the coronal plane of the dummy head now play the “binaural master reference audio file” and is against recorded by the dummy head (playback acoustic crosstalk corrupted audio file A).
5. The two farther speakers in the same row of 2 meters parallel to the coronal plane of the dummy head now play the “ORTF reference audio file” and is against recorded by the dummy head (playback acoustic crosstalk corrupted audio file B).
6. Compare playback acoustic crosstalk corrupted audio files A and B?

Dr. Choueiri might have done it, but I never saw any paper about such experiment. Theoretically they already know the ILD/ITD/spectral cues from the dummy head itself (the HRTF is certainly in the research database of HRTF) and recording engineers also know ILD/ITD from an ORTF pattern (might be useful to place a foam disc between the mics).

I just can’t find graphics in the internet to compare such chains/paths.

Sure, that's more ascetic than the example in my mind but would be better for generalizable results. But such comparisons against several 'common' mic-ing (ortf, decca) and playback (2ch, headphone, 5.1) setups would be very interesting.

 

[pinnahertz]

No, I don't. I'm not here to do what you tell me to do. If you don't take my opinions serioustly then don't. I'm beginning to think you are a tormentor. I have every right to come up with my own new terminology. For normal people explaining the meaning is enough. Not to you apparently.

You said, "[1] Spatial deafness is simply a subgenre of Auditory Adaptation." Where's the backup for that statement? None? OK, then "spatial deafness" isn't a real term. You made it up.

That's the damn problem! Products aren't as good as they could be, because customers are ignorant and manufacturers offer what customers want and customers become even more ignorant and the spiral into stupidity is ready. Crossfeed should have been EVERYWHERE from day one. Even I was totally ignorant of crossfeed until 2012 when I was 41! People need to be educated!

Marketing 101: if it seems better, and costs nothing, it's a "feature" and raises value, you included it differentiate your product and sell more. Cross-feed is a very old idea, but it's never surfaced in products at all.

You are delusional if you think most recordings have adequate ILD for headphones. They are mixed for speakers, not headphones!

Yeah. Me and, um...how many others? I think the statement works both ways.

Does a room mode exist in silence? It is a philosophical question… …anyway, result is what matters.

No. Silly question.

1. If you think they were stunning binaural recordings you are delusional. 2. The thing is it's not only 1958. King Crimson's 70's stuff has insane separation too. Tangerine Dream too. No matter what it is I need crossfeed. Old new whatever. That's how it is. Not cherry picking at all. This stuff was made for speakers. Miles Davis, Cato Barbieri, Herbie Hancock. Rose Royce. Carly Simon. Always too much ILD for headphones. Of my 1500 CDs maybe 3 dozen work best without crossfeed.

1. I don't think they are binaural recordings at all.

2. That is your opinion! You leave no room for anyone else's opinion, and project yours on everyone. I would check my attitude and see if it allows for and respects others opinions. This is a big key to "our problem".

Seems to apply to reality pretty well, because it is created based on observations of reality.

Clearly someone else sees it differently. Does your humility allow for that difference?

1. Yes, normal crossfeed is an approximation of the acoustic crossfeed that happens with speakers.

"Approximation"...this is a very broad and liberal use of that term, if we are still talking about your flavor of cross-feed.

Acoustic crossfeeder isn't "perfect" either unless you listen to the recording in the same studio it was mixed in.

...at which point you don't need cross-feed.

We have to do with less than perfect solutions in real life.

Provided there actually is a problem for that solution.

As such, crossfeed is amazingly good in reducing ILD/ITD and making headphone sound more natural and less fatiqueing. Crossfeed doesn't give 100 % what was intended in the studio, but it gives close-enough version of it that is pleasant and enjoyable. That is what counts.

Your opinion, no evidence yours is global, and mine is different. Can you allow for that?

2. Ask Andy Linkner (Andolink). I built him a crossfeeder in 2014. Also, if you read discussion boards about crossfeed you'll see how many use crossfeed "all the time" just like me. So, your attempt to make me look a singular fool fails.

One example doesn't have any statistical validity.

3. My "facts" may not be 100 % correct (understanding is refined every day), but I have put effort in this so much, that if I was wrong it would be really strange. It would be like saying Einstein didn't understand relativity. I take your posts as personal attacks because instead of demostrating my faults you try to discredit me. You have managed to correct me once (vinyl stylus movements and L+R, L-R). I really was mistaken that time and remembered the directions wrong. Mostly you just point out that your opinions differ from mine. I don't know about you, but I try to build my opinions on scientifically sound premises and finetune them if needed.

You have put forth a lot of effort in solving a problem without any put forth in evaluating the global perception that it even is a problem in the first place.

4. People correct "art" with acoustic crossfeed all the time and nobody cares. Room, speakers and listening position is a complex acoustic system, much more unpredictable than headphone crossfeed.

Headphone perspective is relatively stable across all headphones, and known even to the content creators. There is also the possibility that it is what they intend, already a recognized and accounted for compromise. Can you recognized that someone else (the creators) may already have considered this and deemed it not a problem?

 

[71 dB]

1. He does that because we are in sound science forum and usually most descriptors were already agreed by previous researchers so that the newcomers don’t need to reinvent the wheel.

2. I am failing to see how why crosstalk is bad for speakers and good for headphones.

3. With speakers you add your own HRTF and the real time filtering effect of your head movement. Professor Choueiri goes further and eliminates crosstalk.

4. With headphones you loose the head and torso component of your HRTF and add some filtering from the headphones itself.

5. But adding crossfeed does not bring back your HRTF or real time filtering effect of your head movement. And most of the time ILD for each mixed instrument/track are already blended in the master.

1. The use of crossfeed in headphone listening hasn't been studied much. So, rather than reinventing the wheel, this is inventing the wheel.
2. Stereophonic sound can be divided into three classes: Normal stereo (for loudspeakers), binaural (for headphones) and omnistereophonic (compromize between normal stereo and binaural stereo). Normal stereo on headphones needs crossfeed. Binaural on speakers needs crosstalk canceling. Omnistereophonic sound works as it is on both speakers and headphones.
3. Yes.
4. Yes.
5. Crossfeed is a coarse approximation on HRFT. It models the fact that ILD is small at low frequencies and increases with frequency + it has ITD information. The coarse nature means that it's not "wrong" HRTF, because it doesn't even try to be real HRTF. It just scales ILD to natural levels. ILD is mixed for speakers and are most of the time too large for phones.

1. I see your point with “ping pong stereo recording” (do you really believe that 98% of musical recordings available nowadays have such strong stereo separation?).

2. But perhaps your emotional involvement with your previous efforts with crossfeed, a hard work that deserves our highest respect, may have given you a bias to extrapolate its supposed improvement to 98% of stereo recordings.

3. I see that your desire is to further increase the crossfeed even with modern mixes, but I still fail to see why that would be beneficial instead of neutral or even detrimental.

4. So I guess pinnhertz is not fighting you our personally attacking you, but just trying to do the same thing you believe you are doing, which is educating consumers.

5. The beauty is that I cannot know for sure who is right, because there are too many variables. But every time you both argue I see a new subtlety I was failing to see previously. So don’t feel bad with people disagreeing.

1. Don't be silly. Nobody does ping pong in 2017! It's extreme. A modern pop song requires much weaker crossfeed than a 1958 ping pong recording. This is how I define spatial distortion SD:

A ping pong recording has very strong spatial distortion, maybe 80 % or so. You need very strong crossfeed at level -1 dB or so. A modern pop song requires maybe crossfeed at level -7 dB meaning the spatial distortion is "only" 20 %. The hearing threshold of spatial distortion is about 5-10 %, so below -11 dB crossfeeding becomes useless and the recording is practically free of spatial distortion. Red means severe channel separation and green mild. So, from 1958 to 2017 we have come from red to yellow maybe in average. However, because recordings are mixed for speakers, not many recordings are in the green area and downmixed multichannel movie soundtracks contain A LOT of separation, because rear channels are encoded as channel difference.

2. No. My efforts and hard work come from the realization that almost all recordings benefit from crossfeed.
3. As long as the ILD is unnaturally large crossfeed is beneficial.
4. So what is the message of pinnahertz? Don't do anything?
5. I hope this is beneficial to people reading this board.

 

[bigshot]

If crossfeed were so good, so essential, such a key function, then why hasn't it been standard on even 1% of all music players since the original Sony Soundabout (Walkman) of 1979? It certainly could have been done and very low cost. But it wasn't, hasn't, and isn't still.

I've always wondered why equalization is only available on A/V receivers. It's probably the most important form of signal processing you can use, yet if it exists at all, it's in a primitive bass and treble dial.... and high end amps don't even have that! I guess you can't compare market penetration and usefulness.

If I listened to headphones more, I would like to have a good cross feed. I probably wouldn't use it all the time, but when I needed it, it would be handy. EQ I use all the time.

By the way, if you think no one does ping pong stereo any more, you don't listen to hip hop music!

 

[71 dB]

You said, "[1] Spatial deafness is simply a subgenre of Auditory Adaptation." Where's the backup for that statement? None? OK, then "spatial deafness" isn't a real term. You made it up.

All terms are made up, including Auditory Adaptation. I can make up a term such as RedC as long as I tell others it means "red car." You brought the term Auditory Adaptation here. I would have been happy with spatial deafness only, because that's the part of Auditory Adaptation relevant to crossfeed. You don't need to prove that A = B, if you define A = B by definition! I don't need to prove RedC is a subset of all cars, because my definition of RedC makes the claim true.

Seriously man, this is so pointless. Why can't you just accept my definitions and terms as they are?

Marketing 101: if it seems better, and costs nothing, it's a "feature" and raises value, you included it differentiate your product and sell more. Cross-feed is a very old idea, but it's never surfaced in products at all.

Huh? SPL monitor. Corda Jazz. Dolby Phone. Never surfaced in products? Crossfeed is more relevant than ever, because more people listen to headphones and the sound quality of headphones has improved so that things such as spatial distortion becomes one of the largest problems.

2. That is your opinion! You leave no room for anyone else's opinion, and project yours on everyone. I would check my attitude and see if it allows for and respects others opinions. This is a big key to "our problem".

Tracks can be analysed to see that the ILD-information is objectively unnatural in most cases. My opinions are based on objective facts. The level of proper crossfeed level can be argued and even I may change my opionion about it from day to day, but your opinions are "too" far from the reality as I see it. All I can think of is that you suffer from severe spatial deafness. That would explain why only a few recordings in your opinion benefit from crossfeed.

My attitude might seem arrogant, but this is how I am. I am very shy when I feel I don't know much and very confident when I feel I know what I am talking about.

Clearly someone else sees it differently. Does your humility allow for that difference?

No, I don't think so. Not in THIS issue.

"Approximation"...this is a very broad and liberal use of that term, if we are still talking about your flavor of cross-feed.

Appoximation of ILD and ITD information, not spectral cues. You need something else to have spectral cues.

...at which point you don't need cross-feed.

Except the studio is an acoustic crossfeeder too. Your both ears hear all the monitors. That's crossfeed.

Provided there actually is a problem for that solution.

Unfortunately there is: Spatial distortion.

Your opinion, no evidence yours is global, and mine is different. Can you allow for that?

You don't need my approval to your opinions. You can have any opinions you want. However, if your opinions aren't rooted to reality you may find it difficult to justify them to someone who knows his/her stuff.

One example doesn't have any statistical validity.

How many do you want? A thousand? A million?

You have put forth a lot of effort in solving a problem without any put forth in evaluating the global perception that it even is a problem in the first place.

Six years ago I didn't know it is a problem, but people get wiser. I had no clue about the potential of headphone listening, very few have. People are blind to problems they don't understand. For years I didn't "connect the dots", but in 2012 I did. I wasn't spatially ignorant anymore. It was about the time! Experiences like that make one humble.

Headphone perspective is relatively stable across all headphones, and known even to the content creators. There is also the possibility that it is what they intend, already a recognized and accounted for compromise. Can you recognized that someone else (the creators) may already have considered this and deemed it not a problem?

Why is it that by default everything other people do is "probably" correct, but my crossfeeding is by default iffy? I don't think your average music producer has as deep understanding of human hearing and acoustics as someone who has studied that stuff in university. Understanding how to create catchy radio hits is their field of expertise. It does vary. Some producers demonstrate better understanding of spatiality than others, but somehow I like to use crossfeed with almost everything.

 

[71 dB]

By the way, if you think no one does ping pong stereo any more, you don't listen to hip hop music!

I don't listen to hip hop so I didn't know...

 

[castleofargh]

eheh, I like the name ping pong stereo. it really describes it perfectly.

 

[pinnahertz]

1. The use of crossfeed in headphone listening hasn't been studied much. So, rather than reinventing the wheel, this is inventing the wheel.
2. Stereophonic sound can be divided into three classes: Normal stereo (for loudspeakers), binaural (for headphones) and omnistereophonic (compromize between normal stereo and binaural stereo). Normal stereo on headphones needs crossfeed.

Unsubstantiated statement, at very least the "need" is subjective and varies both with listener preference and specific recording.

Binaural on speakers needs crosstalk canceling.

Yes, but the acoustic crosstalk cancellation is specific to the room, speaker and listener position. It's certainly not ever just an ON vs OFF situation.

Omnistereophonic sound works as it is on both speakers and headphones.

...or it may just be an acceptable compromise for both speakers and headphones.

5. Crossfeed is a coarse approximation on HRFT. It models the fact that ILD is small at low frequencies and increases with frequency + it has ITD information. The coarse nature means that it's not "wrong" HRTF, because it doesn't even try to be real HRTF.

It's not "Right" either then.

It just scales ILD to natural levels. ILD is mixed for speakers and are most of the time too large for phones.

Here's where cross-feed collapses: it's not "right" for anything, exactly, and the need varies widely. It cannot "scale ILD to natural levels" unless the natural level is known...and it isn't.

1. Don't be silly. Nobody does ping pong in 2017! It's extreme. A modern pop song requires much weaker crossfeed than a 1958 ping pong recording. This is how I define spatial distortion SD:

Tend to agree here in general, though not in specific.

[/quote]Opinion. Not fact at all!

A ping pong recording has very strong spatial distortion, maybe 80 % or so. You need very strong crossfeed at level -1 dB or so. A modern pop song requires maybe crossfeed at level -7 dB meaning the spatial distortion is "only" 20 %. The hearing threshold of spatial distortion is about 5-10 %, so below -11 dB crossfeeding becomes useless and the recording is practically free of spatial distortion. Red means severe channel separation and green mild. So, from 1958 to 2017 we have come from red to yellow maybe in average. However, because recordings are mixed for speakers, not many recordings are in the green area and downmixed multichannel movie soundtracks contain A LOT of separation, because rear channels are encoded as channel difference.

Opinion, opinion, opinion!!!!! IT depends on the goal of the recording and the preference of the listener. The chart, and above statements are unproven, opinion only.

2. No. My efforts and hard work come from the realization that almost all recordings benefit from crossfeed.

That is opinion, and not agreed with by everyone. There is no data to support this.

3. As long as the ILD is unnaturally large crossfeed is beneficial.

4. So what is the message of pinnahertz? Don't do anything?

We're on post 267 and you still don't know? Re-read. I'm not suggesting cross-feed cannot be a benefit, I'm stating that your opinions, presented as facts, cannot be facts. My personal preference strongly conflicts with yours. Use it if it works, don't if it doesn't. I object most emphatically to your definitive statements about the necessity of cross-feed!

5. I hope this is beneficial to people reading this board.

I strongly doubt that.[/QUOTE]

 

[71 dB]

eheh, I like the name ping pong stereo. it really describes it perfectly.

Yes, I like the name too, put not the stereo image itself. It doesn't work well even with speakers!

 

[pinnahertz]

All terms are made up, including Auditory Adaptation. I can make up a term such as RedC as long as I tell others it means "red car." You brought the term Auditory Adaptation here. I would have been happy with spatial deafness only, because that's the part of Auditory Adaptation relevant to crossfeed. You don't need to prove that A = B, if you define A = B by definition! I don't need to prove RedC is a subset of all cars, because my definition of RedC makes the claim true.

If you look up Auditory Adaptation you will find a clear and universally accepted definition. If you look up Spatial Deafness, you will find you are misusing the term.

Seriously man, this is so pointless. Why can't you just accept my definitions and terms as they are?

Because they are understood only by you. That makes them useless.

Huh? SPL monitor. Corda Jazz. Dolby Phone. Never surfaced in products?

What are those? How much of the market share do they have? How about iPhone? iPod? Sony products? Sanza? Anything main stream at all?

Crossfeed is more relevant than ever, because more people listen to headphones and the sound quality of headphones has improved so that things such as spatial distortion becomes one of the largest problems.

...or not. Opinion again.

Tracks can be analysed to see that the ILD-information is objectively unnatural in most cases.

Have you now defined "objectively unnatural" in an artistic medium? What do you use to objectively measure "unnatural"? Perhaps you should look up the definition of "objective".

My opinions are based on objective facts. The level of proper crossfeed level can be argued and even I may change my opionion about it from day to day, but your opinions are "too" far from the reality as I see it.

Your opinions are not based on anything except your opinions. If it's an objective fact it won't change with your opinion. Have you tested your theories on a large sample of listeners? Have you determined general preferences? No, and No! You don't have a shred of objectivity here.

All I can think of is that you suffer from severe spatial deafness. That would explain why only a few recordings in your opinion benefit from crossfeed.

SOOO typical. "I hear it and you don't so you must be defective/deaf/stupid/<fill in any derogatory term here>. Your attitude of superiority and arrogance is indeed monumental.

And you would think that because if you believe it yourself, then it is absolutely right and all other views are wrong. You don't see just a tiny bit wrong with that?

Well, here's news: I'm not spatially deaf, not at all! I can easily and correctly localize sounds in a 360 degree sphere, and properly judge distance. I've made binaural recordings, I've researched acoustic crosstalk cancellation decades before you were listening to headphones, I've mixed in 5.1/7.1. I've designed and built many high-end surround systems and calibrated theaters. Yeah, I'm spatially deaf as a post. Right. I also don't think cross-feed is "necessary" in most cases, but is beneficial in a few. Do you know the difference between a physical defect and a difference of opinion?? Or, do I correctly assume than anyone who differs with your opinion must, by definition, be defective?

My attitude might seem arrogant, but this is how I am. I am very shy when I feel I don't know much and very confident when I feel I know what I am talking about.

Yes, it seems arrogant, to the extreme. I'm afraid you don't know what you don't know, which is actually true of everyone. Realize that's true.

No, I don't think so. Not in THIS issue.

Not a surprise given the apparent lack of humility.

Except the studio is an acoustic crossfeeder too. Your both ears hear all the monitors. That's crossfeed.

But that's already taken into account when the mix is created, it's already been compensated for.

Unfortunately there is: Spatial distortion.

I don't completely disagree that your (made up pet) term is real in some recordings and that cross-feed can mitigate it. I disagree with your universal application of it as the universal fix. You have no factual backup for that at all. Restating your strong opinion as fact doesn't constitute factual data or research.

You don't need my approval to your opinions. You can have any opinions you want. However, if your opinions aren't rooted to reality you may find it difficult to justify them to someone who knows his/her stuff.

I could actually just copy/paste that as a reply, but you may not get it because you haven't yet. Your opinions are repeatedly stated as immutable fact, but you have no actual test results or research to back them up. You have your theory, and analysis of what you believe is a "problem", but don't have any data to back up the fact that it is generally viewed as a problem. In short, you have strong opinions you state as fact. I have a problem with that.

How many do you want? A thousand? A million?

I'd like you to take your wild claims and scale them to reality. That's going to take more than one cherry-picked example from an era when the stereo medium wasn't fully understood.

Six years ago I didn't know it is a problem, but people get wiser. I had no clue about the potential of headphone listening, very few have. People are blind to problems they don't understand. For years I didn't "connect the dots", but in 2012 I did. I wasn't spatially ignorant anymore. It was about the time! Experiences like that make one humble.

6 years! Well, I've been researching acoustic crosstalk, headphone cross-feed, psychoacoustics, the reproduction of 3D sound, blah blah blah....it doesn't matter...since 1980. Does that help you respect my opinion? No! At this point it's pointless for us to throw our backgrounds at each other. I don't really care if you've been at it for 60 years or 600 years, stating opinion as fact with zero backup is propaganda. Arrogant self-righteousness and the blatant disrespect of others opinions is abhorrent in a scientific community, or any community.

Why is it that by default everything other people do is "probably" correct, but my crossfeeding is by default iffy?

I didn't say everything other people do is "probably" correct. I'm saying your cross-feeding iffy because of lack of supporting data. You have not done the research.

I don't think your average music producer has as deep understanding of human hearing and acoustics as someone who has studied that stuff in university. Understanding how to create catchy radio hits is their field of expertise. It does vary. Some producers demonstrate better understanding of spatiality than others, but somehow I like to use crossfeed with almost everything.

You are discussing three mutually exclusive properties: a producer's ability to make something the market demands, the understanding of human hearing and acoustics, and your preference for cross-feed. I do not see that they must all be equally present for a successful recording.

 

[pinnahertz]

I've always wondered why equalization is only available on A/V receivers. It's probably the most important form of signal processing you can use, yet if it exists at all, it's in a primitive bass and treble dial.... and high end amps don't even have that! I guess you can't compare market penetration and usefulness.

If I listened to headphones more, I would like to have a good cross feed. I probably wouldn't use it all the time, but when I needed it, it would be handy. EQ I use all the time.

Auto EQ is only on A/V products for several reasons, while every single system would benefit from EQ, it's not found on stereo systems for many reasons. One reason is, with stereo you're only dealing with two, usually identical, speakers. Two is: two-channel stereo only actually works right in one tiny LP, so it is generally accepted as mostly out of calibration because few will actually make routine effort to sit in that LP. And possible Three: EQ is viewed by the two-channel audiophool group as categorically bad...even though the entire speaker/room system is already applying drastic EQ. B

Auto EQ made multichannel system setup much easier, and much more accurate, and compensated for difficult position issues with delay and different speaker types with EQ. This seemed to happen at the same time that the problems with early noise/RTA based systems (that didn't hit the target well) were mitigated by sophisticated FFT-based systems.

Today you can buy a miniDSP product that does two channels, pretty much automatically.