[jgazal]

2. Don't be silly.

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

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

I am not being silly.

I really have empathy for you.

So let me ease your pain, so that I can ease mine also.

So let's go toguether, but promise you are going to be patient with me.

Read what you had previously written:

Cancellation of loudspeaker crosstalk as a concept is familiar to me. I studied acoustics in the university and worked in the acoustics lab for almost a decade. However, I am not sure why you talk about loudspeaker cross-talk cancellation in a thread about cross-feed in headphone listening. Personally I am not that worried about loudspeaker cross-talk. It is a "natural" acoustic phenomenon that doesn't create unnatural signals to my ears. Making the listening room more absorbent and using more directional loudspeakers one can reduce cross-talk, if that is an issue. (...)

I think you confuse cross-talk and cross-feed in some places.

Is it loudspeaker crosstalk or crossfeed?

I thought feed was something that you do deliberately and not something that occurs without human intervention.

Now read what Professor Choueiri wrote:

10 How does BACCH™ 3D Sound work?

Imagine a musician who stands on the extreme right of the stage of a concert hall and plays a single note. A listener sitting in the audience in front of stage center perceives the sound source to be at the correct location because his brain can quickly process certain audio cues received by the ears. The sound is heard by the right ear first and after a short time delay (called ITD) is heard by the left ear. Furthermore there is a difference in sound level between the two ears (called ILD) due to the sound having travelled a little longer to reach the right ear, and the presence of the listener’s head in the way. The ILD and ITD are the two most important types of cues for locating sound in 3D and are to a good extent preserved by most stereophonic technique11.

When the stereo recording is played through the two loudspeakers of a standard stereo system, the ILD and ITD cues are largely corrupted because of an important and fundamental problem: the sound recorded on the left channel, which is intended only for the left ear, is heard by both ears. The same applies to the sound on the right channel. Consequently, an audiophile listening to that recording on standard stereo system will not correctly perceive the musician to be standing on the extreme right of the stage but rather at the location of the right speaker.

Consequently the perceived soundstage is mostly confined to an essentially flat and relatively limited region between the two loudspeakers irrespective of the quality and cost of the hardware in the standard stereo system - the 3D image is greatly compromised12.

In order to insure the correct transmission of the ILD and ITD cues to the brain of the audiophile, the sound from the left loudspeaker to the right ear, and that reaching the left ear from the right loudspeaker (called “crosstalk”) should be cancelled.

The technique of crosstalk cancellation (XTC) has been known for some time and can be applied by filtering the recorded sound through an XTC filter before feeding it to the speaker. This can easily be done digitally. However, until recently, XTC filters have had a detrimental effect on the sound as they inherently add a strong spectral coloration to the processed signal (i.e. they severely change the tonal character of the sound). This is why XTC had not been widely adopted by stereo manufacturers and audiophiles. (See the detailed discussion on XTC-induced sound coloration in the technical paper).

BACCH™ 3D Sound is based on a breakthrough in XTC filter design, that allows producing optimized XTC filters, called BACCH13 filters, that add no coloration to the sound for a listener in the sweet spot. Not only do BACCH filters purify the sound from crosstalk, but they also purify it from aberrations by the playback hardware in both the frequency and time domains.

The result is a 3D soundstage with a striking level of spatial and tonal fidelity never experienced before by audiophiles.
_______________

11 They are most accurately preserved if the recording is made with a dummy head (see Q&A 13).

12 Aside from greatly compromising the 3D image, standard stereo (and even more, surround sound), inherently suffers from the problem of comb filtering, which significantly alters the tonal content of sound, and which is due to the interference of sound waves emanating from more than one speaker.

13 BACCH stands for “Band-Assembled Crosstalk Cancellation Hierarchy” - a name that represents the mathematical filter design method and pays tribute to the great composer with a similar sounding name.

Do you still think the "Spatial distortion" you want to address by adding electronic/digital crossfeed at headphone playback is worse than the fundamental problem Professor Choueiri wants to solve with acoustic crosstalk cancellation at speakers playback?

I don’t.

If you do, then this is the second time Dr. Choueiri is in disagreement with you.

Do you still think that "the studio is an acoustic crossfeeder too" instead of what Dr. Choueiri designates as "crosstalk", an "important and fundamental problem"?

If you do, then this is the third time Dr. Choueiri is in disagreement with you.

That’s why I believe pinnahertz insisted that you need to be precise with terms describing physical phenomena.

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

I would only say "natural levels" provided that you clearly state that your reference is stereophonic recordings played with loudspeakers without crosstalk cancelation.

I would say "standard levels with currently mainstream dsp-less playback environments".

And why one would need to state that?

Because dsp-less mastering and playback environments with the "fundamental problem" of "crosstalk" are not the state of the art anymore.

6. Unfortunately there is: Spatial distortion.

I would not say "Spatial distortion" at all.

If you have stereophonic recording convolved with a binaural room impulse response (two measured speakers and two looking angles) with interpolation in real time to account for head movements and equalization to neutralize the headphone filtering, would you still add crossfeed?

If you want to monitor how your mix is going to sound in currently mainstream dsp-less playback environments, then certainly yes.

If you want to monitor how your mix is going to sound in state of the art playback environments in which your listener has the same headphone rig as you have or speakers with a crosstalk cancellation device (for instance Professor Choueiri bacch-dsp or an phased array of beaforming transducers), then adding crossfeed in the “magic number” of “98%” of your recordings is certainly decreasing "spatial and tonal fidelity".

So now read this post again:

I don’t think he is trying to fight.

I believe he is trying to figure out what is in your opinion the percentage of recordings with unnatural ILD.

So please, do you mind to describe which recordings from 1958 are you referring to and how they were mixed?

To have a numerical perspective, just forget the algorithms that choose music according to music preference and past choices and tell us: if anybody plays 100 musics choosen randomly, how many does have in your opinion unnatural ILD?

You write about crossfeed as if they (recordings with unnatural ILD) were the majority.

To put into perspective, this is Professor Choueiri opinion, which seems to be the opposite, in other words, the minority of recordings:

13 Is the 3D realism of BACCH™ 3D Sound the same with all types of stereo recordings?

The stereophonic recording technique that is most accurate at spatially representing an acoustic sound field is, incontestably, the so-called “binaural” recording method15, which uses a dummy head with high-quality microphone in its ear 16. Until the recent advent of BACCH™ 3D Sound, the only way for an audiophile to experience the spectacular 3D realism of binaural audio was through headphones. Many such recordings exist commercially, and more have recently been made thanks to the iPod revolution.

BACCH™ 3D Sound shines at reproducing binaural recordings through two loudspeakers and gives an uncannily accurate 3D reproduction that is far more stable and realistic than that obtained by playing binaural recordings through headphones17.

All other stereophonic recordings fall on a spectrum ranging from recordings that highly preserve natural ILD and ITD cues (these include most well-made recordings of “acoustic music” such as most classical and jazz music recordings) to recordings that contain artificially constructed sounds with extreme and unnatural ILD and ITD cues (such as the pan-potted sounds on recordings from the early days of stereo). For stereo recordings that are at or near the first end of this spectrum, BACCH™ 3D Sound offers the same uncanny 3D realism as for binaural recordings18. At the other end of the spectrum, the sound image would be an artificial one and the presence of extreme ILD and ITD values would, not surprisingly, lead to often spectacular sound images perceived to be located in extreme right or left stage, very near the ears of the listener or even sometimes inside of his head (whereas with standard stereo the same extreme recording would yield a mostly flat image restricted to a portion of the vertical plane between the two loudspeakers).

Many of well-made popular music recordings over the past two decades have been recorded and mastered by engineers who understand natural sound localization and construct mostly natural-like stereo images, albeit artificially, using realistic ILD and ITD values. Such recordings would give a rich and highly enjoyable 3D soundstage when reproduced through BACCH™ 3D Sound.

——————
15 The accuracy is due to the fact that binaural audio preserves not only the correct ILD and ITD cues discussed in Q&A10, but also contains so-called “spectral cues,” which are the effects the torso, head and ears have on the frequency response and which the brain uses, in addition to ILD and ITD cues, to locate sound, especially at higher frequencies.

16 The spatial accuracy of dummy head recording is only surpassed by recordings made with microphones placed in the listener’s own ears - alas, a rare commodity that would have benefits upon playback for only that listener.

17 This is because binaural playback through headphones or earphones is very prone to head internalization of sound (which means that the sound is perceived to be inside the head) and requires, in order to avoid this problem, an excellent match between the geometric features of the head of the listener and those of the dummy head with which the recording wasmade (this problem has been recently surmounted by the Smyth headphones technology http://www.smyth-research.com/). Pure Stereo does not suffer from this problem as the sound isplayed back though loudspeakers far from the listener’s ears.

18 The 3D realism is the same although the ability of reproducing a sound source at a location that accurately corresponds to the original location is relatively decreased due to the absence of spectral cues.

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

I believe he agrees that stereophonic with "unnatural ILD" (more than one would find if an real instrument were played at the position you want to pan it) may benefit, stating clearly that the reference for "natural ILD levels" is the real musical event and a dummy head and not the currently mainstream mastering room dsp-less environment.

The reference used to be the crosstalk corrupted mastering environment. Now dsp allows an such ambitious reference as the real musical event thus ILD experienced with acoustic crosstalk cannot be considered "natural".

I know @pinnahertz is going to disagree since he still uses the mastering room environment as the reference for fidelity, but I am just saying that, creative intent aside, the state of the art dsp allows the mastering enviroment to be potentially faithful to the spatiality of the music event.

7. 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.

I am afraid you still do not agree with Professor Choueiri that acoustic crosstalk speakers playback is worse than the lack of crossfeed in headphone playback. And I am certain that Professor Choueiri is humbler than you.

I know it is hard to give up something that was so dear to you since 2012, but just let it go. We need to.

That is why I guess, I believe, I have the opinion that deliberatly adding, digitally or eletronically, crossfeed at playback may be beneficial only to "ping pong recordings", recordings with "strong stereo separation", recordings with "unnatural ILD".

P.s.: By dsp-less I mean mastering environments in which dsp in not restricted to panning and synthetic reverberation, but goes further to cancel crosstalk in speakers and externalize virtual speakers on headphones.

 

[bigshot]

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

It works great for Sly and the Family Stone!

 

[pinnahertz]

I

I would only say "natural levels" provided that you clearly state that your reference is stereophonic recordings played with loudspeakers without crosstalk cancelation.

I would say "standard levels with currently mainstream dsp-less playback environments".

And why one would need to state that?

Because dsp-less mastering and playback environments with the "fundamental problem" of "crosstalk" are not the state of the art anymore.

Technically correct, but nearly all mixing is still done without DSP correcting acoustic crosstalk of the monitor speakers in the mix environment.

I would not say "Spatial distortion" at all.

If you have stereophonic recording convolved with a binaural room impulse response (two measured speakers and two looking angles) with interpolation in real time to account for head movements and equalization to neutralize the headphone filtering, would you still add crossfeed?

No, that would pretty much do it right there.

If you want to monitor how your mix is going to sound in currently mainstream dsp-less playback environments, then certainly yes.

But nobody does that. You could go nuts calling up one virtual room/system after another. You can't make mix or mastering decisions that way.

If you want to monitor how your mix is going to sound in state of the art playback environments in which your listener has the same headphone rig as you have or speakers with a crosstalk cancellation device (for instance Professor Choueiri bacch-dsp or an phased array of beaforming transducers), then adding crossfeed in the “magic number” of “98%” of your recordings is certainly decreasing "spatial and tonal fidelity".

So now read this post again:





I believe he agrees that stereophonic with "unnatural ILD" (more than one would find if an real instrument were played at the position you want to pan it), stating clearly that the reference for "natural ILD levels" is the real musical event and a dummy head and not the currently mainstream mastering room dsp-less environment.

The reference used to be the crosstalk corrupted mastering environment. Now dsp allows an such ambitious reference as the real musical event thus ILD experienced with acoustic crosstalk cannot be considered "natural".

I know @pinnahertz is going to disagree since he still uses the mastering room environment as the reference for fidelity, but I am just saying that, creative intent aside, the state of the art dsp allows the mastering enviroment to be potentially faithful to the spatiality of the music event.

Actually, you've sort of made the point for me. The question of where/when to "correct" depends entirely on the knowledge of what was done during the creation process, knowledge of the difference between that environment and the play environment, AND, ultimately, the intent of the creator vs the intent of the listener. If it matters to the listener that he hear the content as it was intended, then obviously he needs to replicate the creative environment as nearly as practical. However, the listener may not give a crap about what the creator intended and may think his preference is superior anyway, as in the case of 71dB. Or, the listener may not perceive anything is amiss and joyfully listen to his Bose Lifestool system. Whatever, we all know that virtually nobody will actually attempt to replicate the mix environment, so it's not expected. Music is mixed on studio speakers, checked on other speakers and headphones, and whatever adjustments may help are made. But thinking you can somehow correct for a relatively random set of conditions is simply delusional. What's actually being done is the exercise of free choice and preference.

But "Spatial Distortion", whatever that is, is actually a relatively minor problem in playback. There are such HUGE other problems, I find it strange anyone would lock into that one. Frequency response of play systems is a very big problem, particularly with typical headphones and speakers. It's all over the place, with excursions from any hope of match falling into the 20dB range sometimes. And FR affects our mix, perspective, balance...pretty much the whole presentation. And yet if you go and demo headphones, the ones that actually come at all close to good and smooth are very, very few. Faced with that, and the lack of any clear subjective preference for cross-feed in general, it seems like beating a dead horse.

I am afraid you still do not agree with Professor Choueiri that acoustic crosstalk speakers playback is worse than the lack of crossfeed in headphone playback. And I am certain that Professor Choueiri is humbler than you.

Behind the arrogant stand a long line of the more humble.

I know it is hard to give up something that was so dear to you since 2012, but just let it go. We need to.

That is why I guess, I believe, I have the opinion that deliberatly adding, digitally or eletronically, crossfeed at playback may be beneficial only to "ping pong recordings", recordings with "strong stereo separation", recordings with "unnatural ILD".

P.s.: By dsp-less I mean mastering environments in which dsp in not restricted to panning and synthetic reverberation, but goes further to cancel crosstalk in speakers and externalize virtual speakers on headphones.

Nearly every mixing and mastering environment uses minimal DSP in the monitor chain, perhaps some precision EQ, until you get to multichannel rooms. None use DSP crosstalk cancellation, and nobody makes mix decisions on virtualized speakers in headphones.

 

[jgazal]

Technically correct, but nearly all mixing is still done without DSP correcting acoustic crosstalk of the monitor speakers in the mix environment.

Agree.

AND, ultimately, the intent of the creator vs the intent of the listener.

Agree that intent of the creator must be respected.

But "Spatial Distortion", whatever that is, is actually a relatively minor problem in playback. There are such HUGE other problems, I find it strange anyone would lock into that one. Frequency response of play systems is a very big problem, particularly with typical headphones and speakers. It's all over the place, with excursions from any hope of match falling into the 20dB range sometimes. And FR affects our mix, perspective, balance...pretty much the whole presentation. And yet if you go and demo headphones, the ones that actually come at all close to good and smooth are very, very few. Faced with that, and the lack of any clear subjective preference for cross-feed in general, it seems like beating a dead horse.

I agree that reference can’t be something that is not in mainstream use.

I was just trying to pull the argument to its limit.

But anyway consider what Smyth and Choueiri says:

19 How does BACCH™ 3D Sound correct problems in audio playback?

While the foremost goal of BACCH™ 3D Sound is 3D audio imaging, the BACCH filters at the heart of BACCH™ 3D Sound have the additional advantage of correcting, in both the time and frequency domains, many non-idealities in the playback chain, including loudspeaker coloration and resonances, listening room modes, spatial comb filtering, balance differences between channels, etc…

This capability, often called “room correction” (even though it corrects more than just the effects of the listening room), is inherent to the nature of BACCH filters, whose design is made under the requirement that a perfect impulse response fed into, say, the left channel (only) of the playback system, should appear as close to a perfect impulse response at the left ear of a listener sitting in the sweet spot.

Therefore a BACCH™ 3D Sound processor functions also as a state-of-the-art and customized digital room correction unit.

In the following podcast interview (in English), Stephen Smyth explains concepts of acoustics, psychoacoustics and the features and compromises of the Realiser A16, like bass management, PRIR measurement, personalization of BRIRs, etc.

He also goes further and describes the lack of absolute neutral reference for headphones and the convinience of virtualizing a room with state of the art acoustics, for instance “A New Laboratory for Evaluating Multichannel Audio Components and Systems at R&D Group, Harman International Industries Inc.” with your own PRIR and HPEQ for counter-filtering your own headphones (@Tyll Hertsens, a method that personalizes room/pinnae and pinnae/headphones idiosyncratic coupling/filtering and keeps the acoustic basis for Harman Listening Target Curve).

Still very ideal scenarios. But who knows what future is going to be like with easier ways to acquire personalized HRTF instead of just binaural room impulse responses...

I agree that what really needs to be preserved from an real recorded event are mainly (a) the acoustics relationships between the hemispheres that are divided by the human sagittal plane (crosstalk cancellation with speakers or crossfeed free externalization), (b) spectral cues (personal HRTF/HRIR/BRIR or in a lower performance tier a generic HRTF/HRIR/BRIR) and (c) bass response (for instance, the A16 allows direct bass, phase delayed bass and even equalization in the time domain). But once those three factors are controlled to maintain localization fidelity, the overall frequency response and seat perspective not necessarily need to be referenced to the real event.

Then it will be up to the creator to deliberately define the art, by creative intent, not only by carefully altering some of the frequency bands of real recorded events (dummy heads, baffled spaced microphones, ambisonics, eigenmikes), but also in synthetic recreations (stereophonic mixing or binaural synthesis with synthetic reverberations).

By “carefully” I mean choosing the recording room, where to place the musician and instrument, where to place the microphone etc. Anyway, if your aim is to keep elevation precision, I would be extra careful when altering some specific frequency bands with post equalization either of recordings made with microphone patterns that encode spatial info (dummy heads, sound field microphones or eigenmikes) or with synthetic mixing.

Correct for such “relatively random set of conditions” may seem “delusional”, but if we do not do it, virtual reality won’t work. And I believe consumers want it to work.

Edited to rephrase a confusing paragraph and to complete the idea of creative liberty of the producer...

 

[jgazal]

But "Spatial Distortion", whatever that is, is actually a relatively minor problem in playback.

By the way, I recently created a PRIR for stereo sources that simulates perfect crosstalk cancelation. To create it, I measured just the center speaker, and fed both the left and right channel to that speaker, but the left ear only hears the left channel because I muted the mic for the right ear when it played the sweep tones for the left channel, and the right ear only hears the right channel because I muted the mic for the left ear when it played the sweep tones for the right channel. The result is a 180-degree sound field, and sounds in the center come from the simulated center speaker directly in front you, not from a phantom center between two speakers, so they do not have comb-filtering artifacts as they would from a phantom center.

Binaural recordings sound amazing with this PRIR and head tracking.

How do you mute the opposite microphone?

To mute it I unplug the left or right microphone from the Y-junction between sweeps. I set the "post silence" to 8 seconds beforehand to give me enough time. To make it easier I plan to hook up an A/B switch.

I actually got the idea from a comment by Timothy Link in this Stereophile article about Dr. Choueiri's BACCH.
http://www.stereophile.com/content/bacch-sp-3d-sound-experience

You can also add a rear speaker to the PRIR for the left and right surround channels to achieve a full 360-degree circle like PanAmbiophonics, and additional speakers for hall ambience.

Can you describe that sensation of envelopment improvement you heard between the first PRIR and the "Hafler" PRIR?

Does the front soundstage keep believable in both PRIRs as you turn your head?

Using the first PRIR, central sounds seem to be in front of you, and they move properly as you turn your head. However, far-left and far-right sounds stay about where they were. That is, they sound about the same as they did without a PRIR, and they don't move as you turn your head. In other words, far-left sounds stay stuck to your left ear, and far-right sounds stay stuck to your right ear. It's possible to shift the far-left and far-right sounds towards the front by using the Realiser's mix block, which can add a bit of the left signal to the front speaker for the right ear, and a bit of the right signal to the front speaker for the left ear.

Using the Hafler PRIR, there seems to be a greater sense of space and ambience for all sounds. If the recording was matrix-encoded, some sounds extend beyond the far-left and far-right and wrap around you. Initially I noticed that far-left and far-right sounds moved too much when I turned my head, but after I increased the front speaker level to be 3 dB higher than the rear speaker level, they moved properly.

Why is this happening?

Would it be a minor problem if you wanted to match sound localization and image from a virtual reality headset?

I am just hoping the musical industry to take a free ride in vr development.

 

[jgazal]

(...) None use DSP crosstalk cancellation, and nobody makes mix decisions on virtualized speakers in headphones.

“Nobody” is such an strong word: https://www.soundonsound.com/reviews/smyth-research-realiser-a8 (published July 2013).

Though I agree that possible very few make mix decisions on virtualized speakers in headphones and a PRIR set to avoid crossfeed.

I say very few because Professor Choueiri and Chesky Records recording engineer may be the only two human beings doing that.

 

[71 dB]

What are the scientifically proven facts of crossfeed that are not opinions? Human hearing has been studied for a long time. A lot of research done. I apply those "facts" to audio reproduction and to my ears it works. It works for Andy Linkner. It works for my friends.

 

[71 dB]

The question of where/when to "correct" depends entirely on the knowledge of what was done during the creation process, knowledge of the difference between that environment and the play environment, AND, ultimately, the intent of the creator vs the intent of the listener.

Some aspect are more relevant than others. A creator can't expect his/her creation to be listened to in an exact same environment. A creator needs understanding of what kind of variation you have within typical listening environments and optimize his/her creation to work as well as possible for such "set" of environments. People don't consume music in studios. They consume it in at their living room and other everyday places, but almost never in a studio.

The intents of creators are heavily affected by commercial demands which are often rather irrational such as loudness war. So, we should not blindly worship all intends of the creators. We can use our own head.

However, the listener may not give a crap about what the creator intended and may think his preference is superior anyway, as in the case of 71dB.

Creators have 3 options:

- mix for speakers
- mix for headphones
- mix for both and minimaze the downsides of such compromize

The default is speakers. As headphone listening has become more popular, maybe it's 60-70 % for speakers and 30-40 % headphones for typical pop, but that is not a balanced compromize between the two and leaves some spatial distortion to be crossfed away for headphones. I'm sure most creators agree with me about proper crossfeed. It makes the sound more natural and less tiring, so why wouldn't they? As I have said before, spatial distortion hardly is the intent of any creator.

But "Spatial Distortion", whatever that is, is actually a relatively minor problem in playback. There are such HUGE other problems, I find it strange anyone would lock into that one. Frequency response of play systems is a very big problem, particularly with typical headphones and speakers. It's all over the place, with excursions from any hope of match falling into the 20dB range sometimes. And FR affects our mix, perspective, balance...pretty much the whole presentation. And yet if you go and demo headphones, the ones that actually come at all close to good and smooth are very, very few. Faced with that, and the lack of any clear subjective preference for cross-feed in general, it seems like beating a dead horse.

What HUGE problems? Listening with speakers suffer from acoustics problems and less than optimum placement of speakers and listening position. These problems are significant, but also hard to fix, althou not impossible if you have money and an understanding wife. Fortunately, there is nothing unnatural about bad acoustics. Boomy bass is not unnatural, it just sounds bad. Fuzzy stereo image is not unnatural, the ILD an ITD information is natural presentation of fuzziness. Frequency response errors hurt fidelity of course, but are totally natural. Spatial distortion however is unnatural and that makes is a HUGE problem. Fortunately it can be fixed with crossfeed pretty easily.

 

[pinnahertz]

What are the scientifically proven facts of crossfeed that are not opinions? Human hearing has been studied for a long time. A lot of research done. I apply those "facts" to audio reproduction and to my ears it works. It works for Andy Linkner. It works for my friends.

You have not established:

1. A statistically significant portion of headphone listeners don't like the perspective

2. A statistically significant portion of headphone listeners prefer cross-feed...at all, much less 98% of all recordings.

You have established:

1. You like it so much you can't live without it

2. Your friends like it (could be any number from 2 up, but we don't know how it was tested, demo selections, biases...etc.)

Now, none of that is a problem until you go and proclaim it as THE SOLUTION, and mandatory on 98% of all recordings ever made, proclaim the listening masses as ignorant, and accuse any with an opposing view as being somehow inept or impaired, and proclaim your own view as the only correct one.

 

[bigshot]

A creator can't expect his/her creation to be listened to in an exact same environment. A creator needs understanding of what kind of variation you have within typical listening environments and optimize his/her creation to work as well as possible for such "set" of environments.

That's only true insofar that the sound in the studio isn't compromised. Creators usually focus on the main mix on the big monitors. They might do a pass at the end listening on small speakers, but that is mostly just to make sure nothing is blasting out. I've never seen anyone actually mix to anything but the main monitors in a mixing stage. Making the sound fit for various purposes is the job of the mastering engineer, and the creator usually has very little part in that.

As for the necessity of cross feed... I think DSPs will become more important as time goes by. They're already standard in multichannel AVRs, and I think DSPs for two channel will eventually be implemented as plugins for cell phones and DAPs for headphone listening as well. I don't think the fact that cross feed hasn't been standard up to now is any reason to think they won't be standard in the future. It just takes more sophisticated processing schemes designed specifically for how people listen to music today.

 

[pinnahertz]

Some aspect are more relevant than others. A creator can't expect his/her creation to be listened to in an exact same environment. A creator needs understanding of what kind of variation you have within typical listening environments and optimize his/her creation to work as well as possible for such "set" of environments. People don't consume music in studios. They consume it in at their living room and other everyday places, but almost never in a studio.

Those who still have short term memory will recall that I have already stated that creators mix assuming few if any will listen to their work in a similar environment.

The intents of creators are heavily affected by commercial demands which are often rather irrational such as loudness war. So, we should not blindly worship all intends of the creators. We can use our own head.

Creators have 3 options:

- mix for speakers
- mix for headphones
- mix for both and minimaze the downsides of such compromize

There are other options, like the one actually used: mix on speakers, check on headphones. And by that I mean, assure the intent of the mix doesn't change on headphones, and accept the new perspective as valid (since not everyone agrees that "spatial distortion" is bad or even a problem).

The default is speakers. As headphone listening has become more popular, maybe it's 60-70 % for speakers and 30-40 % headphones for typical pop, but that is not a balanced compromize between the two and leaves some spatial distortion to be crossfed away for headphones.

Where on earth do you get your statistics?

1. I'm sure most creators agree with me about proper crossfeed. 2. It makes the sound more natural and less tiring, so why wouldn't they? 3. As I have said before, spatial distortion hardly is the intent of any creator.

1. Why...because you are the God of Cross-Feed and unless they agree with you they are all imbeciles and should immediately be devoted to destruction? You've already found one who disagrees...quite strongly, so you have two data points: yours and mine. Not even enough to extrapolate a trend!
2. OMG!!!! OPINION! OPINION! OPINION! OPINION! Here's another: It flattens depth, removes ambiences, sucks the life out of the mix, alters mix balances...
3. How would you know? Oh yeah, the God of Cross-Feed knows what everyone intends. How incredibly disrespectful of the very creators on whom your listening depends.

OPINION!

What HUGE problems?

Frequency response would be the biggest one.

Listening with speakers suffer from acoustics problems and less than optimum placement of speakers and listening position. These problems are significant, but also hard to fix, althou not impossible if you have money and an understanding wife.

A symmetrical room layout with an LP in the "sweet spot" is not hard to do, it just isn't a priority. It doesn't take a lot of money, or even that much understanding from your spouse. It does take an interest and a priority. The problem is it takes money for good speakers, and room acoustic treatment is highly disruptive to decor and layout. All that's true, but basic system response is still the huge problem.

Fortunately, there is nothing unnatural about bad acoustics. Boomy bass is not unnatural, it just sounds bad.

Boomy bass is a distortion of the mix! Thin bass is also a distortion! Over-emphasized upper mid is a distortion!

Fuzzy stereo image is not unnatural, the ILD an ITD information is natural presentation of fuzziness.

But that doesn't represent the original...so it's a distortion!
Frequency response errors hurt fidelity of course, but are totally natural.
[/quote]This is absolutely and completely wrong. Response errors are huge, and affect the mix balance to a far, far greater extent that anything else. Spend a week mixing a song, then play it on a consumer system...see how you like your work being remastered by some random response curve!

Spatial distortion however is unnatural and that makes is a HUGE problem.

OPINION! YOURS! "Spatial distortion" has been accepted for decades, where frequency response has been worked on for decades. The industry in general has prioritized response and fully ignored spatial distortion! Why? The reality is, the headphone perspective is fully accepted by most listeners, and has become an alternate version of "natural". That's why there is no cross-feed on any main stream music player. Yet every respectable AVR has room EQ. There are many ways of applying headphone EQ.

Fortunately it can be fixed with crossfeed pretty easily.

The application of the "fix" is subjective, generalized, and requiring variation with every track. This is not true of response equalization.

You have defined the priority and magnitude of problems according to your own preference. I see no point in presenting a different opinion since our resident egomaniac has made up his mind.

Yesterday, as a sanity check (readers will understand my need for this!) I listened to an early 1970s album all the way through using cross-feed. Initially I was surprised because the first track heard with cross-feed presented a guitar moved out of it's normal pan position into one elevated up about 45 degrees. Interesting, but not right. I listened to the entire album, and within a song or two, the cross-feed version no longer sounded "wrong", in fact, it seemed like the band was less in my head, more in front of me. Then, I listened again to the entire album without cross-feed. Wow! I was drawn into the music, the ambience surrounded me, I heard mix details missed before, and the music touched me far more deeply. I enjoyed the entire experience more.

The album was "Who's Next" by The Who.

Now all of that is just my opinion, and I don't expect anyone to share it or agree with it. It's just my preference. But at least I gave it a shot, I opened the possibility that cross-feed might improve the experience. I felt it didn't in that case. However, I'm not done either. I'll be listening to more and varied recordings with cross-feed. Please know, this is not something new I'm doing now. I've experimented with cross-feed for decades, various forms. I keep giving it a chance. I know it has merit, and I'll see if I can find where that is. And, where possible, I ask others opinions (not yours...I have yours pretty much in hand).

I welcome others to do the same kind of comparison. And I'll let you know when I find a track that cross-feed improves.

 

[Zapp_Fan]

As someone who is currently working on feature sets for headphones, I can tell you that it's true, the reason crossfeed isn't implemented on headphones or DAPs is because most people don't know what it is, and so don't ask for it. Whether it's really an essential feature for listening or not, I decline to make a definitive statement

I've done a bit of market research on this very question, and the underlying problem is that it's pretty difficult to get the average person interested in better stereo imaging - unless they are physically listening to an A/B comparison in person. Telling people that the stereo separation is better in marketing copy simply does not work on the typical consumer, at least in my limited but direct experience. If it costs money to implement but the consumer doesn't care, you don't implement it. People have a good sense of what they'd be willing to spend on more SPL, more bass, or a general notion of clearer, better sound, but claims about stereo image is not something that makes the wallets come out, for whatever reason.

All that said, I noticed that the companion app for the Sony WH‑1000XM2 wireless cans includes a spatialization feature, it does some kind of simulation of "club" and "arena" reverb type sounds. Unfortunately, I didn't personally find the effect very pleasing, nor was it something you'd recognize as crossfeed, so much as a special effect. They did, however, implement some pretty convincing HRTFs that offer a front/back/left/right positioning of the sound, so a quality crossfeed scheme built into headphones is certainly within reach today.

But, this does point the way to widespread adoption of crossfeed. Headphone BT chipsets are getting powerful enough to run a relatively sophisticated crossfeed DSP, and I expect companion apps will become more common for headphones over time, so I would say the prospects for crossfeed in the mass market are probably better than they've ever been - the reason being there is no marginal cost for implementing it as a DSP effect, you just need to pay your engineer to implement it once in the firmware.

 

[gregorio]

This thread seems to have become extremely polarised, with some pretty strange assertions which rather typically in audiophile discussions are based on personal opinion/perspective, incorrect facts and important facts which have been simply omitted/ignored/overlooked.

Crossfeeders don't serve coffee. They don't fix everything in the sound. They fix things related to excessive stereo separation and that's it. Why do I even need to say this?

I don't know why you think you need to say that, maybe because some of us realise/understand that crossfeeding does NOT fix things related to excessive stereo image! When we listen to speakers yes, we have the direct dry sound which in the case of a sound panned right is heard first in the right ear and then in the left ear at a lower level. With HPs we would only hear the sound in the right ear and therefore crossfeeding some of it may seem like a good solution and obviously some people feel (strongly apparently) that it does. However, there are a number of reasons why it doesn't:
1. There is not the expected time delay between left and right ears or the expected masking effect of the skull.
2. It's extremely rare for there to be only a direct dry sound on a recording. Sure, in the early days of stereo there was relatively little which could be done about this but even in the 60's ways around that were being commonly/ubiquitously employed; echo chambers, plates and other sympathetic resonators and even reverb units (acoustic/analogue devices such as the EMT units). Recording and playback technology was improving and therefore recordings didn't need to be as dry as previously. The ping/pong effect never really died completely, it just became less obvious over time, using technology such as stereo delays and reverbs for example. Reverb units been stereo for many years, in fact the EMT had a stereo return unit in the early 60's, so just crossfeeding a mix is going to damage, in some cases quite severely, the often delicate timing of left/right reflections.
3. Reverb again but this time it's the reverb of the listening environment, both of them. There was a period in the late '70's and part way through the '80's when control room design fashion was for quite dead/dry acoustics, quite dry multi-track recordings and then add a bunch of reverb and delays artificially. It was fairly quickly realised that did not result in good translation to the consumer and so, for around the last 30 years or so music studios have erred far more towards diffusion than absorption. The concept of mixes being created in relatively dead/dry controls rooms is pretty much an audiophile myth. Most decent music control/mixing rooms have RTs (reverb times) broadly similar to average to the average sitting room. The application of reverb and delay effects is therefore based on the interaction of those effects with the hopefully neutral reverb of the control room. Obviously, consumers don't typically have appropriate diffusers in their listening environments and don't have neutral room reflections/reverb but nevertheless, what they do have is still an important interaction with the reflections/reverb on the recording. Hopefully it's obvious that crossfeeding does not accomplish this same feat, if anything, it can again just make matters worse.

Ultimately, every playback scenario has it's strengths a weaknesses. Speakers are better because the music has been mixed and mastered for them but are badly compromised by relatively poor listening environments. Headphones cut out the worst of poor consumer listening environments but also throws out significant bits of the baby with the bath water. Crossfeeding "fixes" some those issues but makes others worse. For many people it's a case of which disadvantages/weaknesses bother them the least. When I listen on HPs I know I'm going to get an artificially wide stereo image, a recording with less reverb and depth than intended, if I want it to sound more like it does through a good speaker system/environment then I listen on a good speaker system/environment. I've heard some recordings on HPs which benefited from crossfeed, where crossfeed's deficiencies were less objectionable, and I've heard many others where I found the deficiencies too objectionable. Personally, I therefore avoid it, at least with just straight cans I know I'm getting the sound which the engineers in the studio checked and didn't think was too objectionable. That's just my personal preference though, what personally bothers me the least. It's obvious you've got pretty strong feelings on the matter, that you obviously aren't much bothered by the deficiencies of crossfeeding, while being impressed by the issues it does solve. That's up to you of course but also of course that's only your personal preference. You seem to be getting a bit confused and carried away, apparently claiming as fact that it would be better for almost everyone and almost every recording.

Since subwoofers typically output only lowest bass, only in larger rooms we have problems with modes and small room behaves more like a pressure chamber and it's easier to find a good placement for the sub.

I'm not sure how you came to this conclusion, actually the reverse is true. As room size increases, the fundamental room mode frequencies and levels get lower as the distance between the main reflective surfaces increases. In large rooms, most cinemas for example, the lowest room modes are below 20Hz and therefore of no real concern. If small rooms were better, we'd all be mixing in them because it'd be cheaper. Generally, the smaller the room, the worse the acoustic issues and the more difficult they are to effectively treat.

G

 

[bigshot]

My room is about 20 by 20 with a 20 foot high peaked roof and a 10 by 10 foot L off the rear on the right. It's a spit in the wind guesstimate, but I would bet the room mode frequency in my room is down below 30Hz based on my attempts to find the resonant frequency. A half octave at the very bottom isn't enough for me to really worry about, so I just tame it through EQ and furniture placement to lessen reflections and I don't worry about eliminating it through more extreme means. I do try to make sure that the sub has a clear shot at all the listening positions and I put it as close to the center of the front wall as I can, but not exactly in the center. (I find that symmetry sometimes causes problems.) It works for me.

Is there a way to estimate the resonant frequency based on room dimensions? I'm lousy at math, but I'd be interested to see if my guesstimate is in the proper ballpark.

 

[bigshot]

Behind the arrogant stand a long line of the more humble.

At Sound Science that line forms to the rear! The humble may be blessed, but they don’t get much air time around here!