[jgazal]

mindbomb, I tried to get someone here to recommend some dimensional sounding binaural recordings of music a while back. No one could do it. They could only point to hair clippers and clapping hands. The one musical album they recommended and I bought sounded like any other album. It was arrayed from left to right through the center of my skull. I think binaural recording is one of those theories that just doesn't work in practice. It's a mental exercise that ultimately doesn't amount to anything.

Headphones present music as a line going through the center of your head. No space. No soundstage.
Stereo speakers put the sound as a flat plane 8 or 10 feet in front of you. It uses the space in the room to create stereo soundstage.
5.1 presents a natural soundstage in front with a horizontal plane extending from front to back. The room is very important and room acoustics can be simulated.
Atmos presents a box of sound with up/down added to 5.1. The room acoustic is very important and more sophisticated room acoustics can be simulated.

It's a progression from one dimensional to three dimensional.

Have you tried with Dr. Choueiri crosstalk cancellation or with an acoustical barrier between the speakers?

Is 5.1 or Atmos really three dimensional?

 

[ironmine]

Headphones present music as a line going through the center of your head. No space. No soundstage.

How do you know what happens in my head? You can only speak about your head. When I listen to music with crossfeed, I have everything: space, soundstage, layering, etc. The brain is very adaptable. You just need to give it some time to get used to it.

Listening to music that is not processed with a crossfeed algorithm sounds unnatural and annoying, especially bass. I am sure you can get used to it (as I said the brain is very adaptable), but why reject this wonderful tool (crossfeed processing) since it's so easily available and helps us overcome one of the main negative attributes of headphones (that being an insufficient crossfeed)?

This thread is completely hi-jacked by crossfeed-haters. Let's come back to basics:

http://www.meier-audio.homepage.t-online.de/crossfeed.htm

 

[bigshot]

Have you tried with Dr. Choueiri crosstalk cancellation or with an acoustical barrier between the speakers?

Is 5.1 or Atmos really three dimensional?

Why would you want to separate the two channels? Is there a reason for that? My goal has always been to try to balance speaker placement, room acoustics and levels to create as perfect a phantom center as possible between each of the five speakers with all of the four other ones. That meshing is what creates the sound field, and makes multichannel not just sound like sound coming from all directions.

My system is 5.1, but I'm told that Atmos allows you to place a sound object anywhere within left right, front back and top bottom parameters. That would be true 3D. 5.1 is more like a 2D plane extending left right and front back. I have a jury rigged placement that sort of improves on that by raising the level of the center speaker and rear speakers a little higher. That kind of creates a raised triangle overhead that works really well with movies, because dialogue is usually up front and ambiences in the rear. It fills my 10 foot projection screen because the center is behind the screen. It works quite well with multichannel music because the vocals are often in the center, and the rears are either discrete parts of the music, or pure ambience. It works well in my room at least.

 

[bigshot]

How do you know what happens in my head?

Are you setting me up for a joke?

I don't hate crosstalk. It can take the curse off having the sound being shoved right up against your ears. I'm sure it improves the sound of headphones. But it doesn't approximate speakers at all. The whole point to a speaker system is how the sound interacts with the space in the room. Soundstage is dependent on the sound being physically in front of you. You don't get that at all with headphones.

 

[ironmine]

Are you setting me up for a joke?

I don't hate crosstalk. It can take the curse off having the sound being shoved right up against your ears. I'm sure it improves the sound of headphones. But it doesn't approximate speakers at all. The whole point to a speaker system is how the sound interacts with the space in the room. Soundstage is dependent on the sound being physically in front of you. You don't get that at all with headphones.

Yes, it does approximate speakers to a certain degree. Many crossfeed plugins have these additional features: simulation of speakers and simulation of room acoustics. You can activate it on or off depending on your preferences. I usually prefer pure crossfeed, I don't like to simulate either speakers or room.

This is an example:

As you see, you can choose speaker presets or build your own speakers. Room presets or design your own room. Read the manual.

 

[ironmine]

I don't hate crosstalk.

I also did not have any unfortunate or violent accidents in my childhood involving crossfeed that would negatively color my perception of crossfeed for the rest of my life. I guess it makes us both lucky, unlike other folks...

 

[bigshot]

Ironmine, do you have a good speaker setup? If so, you're probably familiar with how it sounds. Can you set that plugin to make it sound just like your listening room? How does this deal with directionality? I have a 5.1 system, and a big part of sound location involves the direction my head is facing. I turn my head while I listen and that helps me locate a sound in space. I can clearly hear the difference in location between my center speaker and my rears. Even with 2 channel stereo when I listen to something with very controlled soundstage, like Culshaw's Ring, I find that turning my head makes a big difference to hearing characters moving around on the stage. I don't know how you would accomplish that without a realizer... and I suspect that my eyes turning along with my ears might have some impact on how I imagine where those sounds are located in space as well. Not to mention the kinesthetic chest thump of loud fat bass- headphones can't do that. It just seems like a lot of the sound of speakers isn't there, even with the best processors.

I love my 5.1 speaker system. If I could find a way to reproduce that in headphones, it would be great. I've heard Atmos headphone mixes and binaural recordings. But none of them are able to reproduce the pinpoint sound location of the soundstage of my speaker system. It can sound kind of the same in the response, but not in directionality. The sound is always right there inside or just around my head. It's never ten or fifteen feet in front of me or behind me.

Maybe it's the shape of my ear canals. I don't know... but headphones always sound in my head and speakers always sound a distance away from me.

 

[jgazal]

Each of us advocates - okay, some more emphatically - our own preferences.
@bigshot, I wouldn't say 5.1 is three dimensional.
The way I see it, codecs with height channels or even Ambisonics (which relies on spherical harmonics) do not have channels enough to render high pitch objects in any azimuth or elevation.
Perhaps we could call such codecs "surround (illusion)" similar to "stereo illusion" instead of 3D//?
What I concur is that, without DSP, externalization is very poor in headphones (even with binaural recordings most of us are very susceptible to sound field collapse with head movements). In such aspect, from what we have commercially available, codecs with height speakers go first, 5.1 second, stereo over loudspeakers third.
But as you already said in the post above, with a little bit of DSP magic, headphones can reproduce a sound field equivalent to any of those three.
I still don't know if binaural with xtc can render 3D better than what you called "obscure" high order Ambisonics - HOA. Those are my candidates to be called one day, perhaps, three dimensional.
Anyway, as @gregorio already explained, those may not be suitable for music.

 

[Davesrose]

My system is 5.1, but I'm told that Atmos allows you to place a sound object anywhere within left right, front back and top bottom parameters. That would be true 3D. 5.1 is more like a 2D plane extending left right and front back. I have a jury rigged placement that sort of improves on that by raising the level of the center speaker and rear speakers a little higher. That kind of creates a raised triangle overhead that works really well with movies, because dialogue is usually up front and ambiences in the rear. It fills my 10 foot projection screen because the center is behind the screen. It works quite well with multichannel music because the vocals are often in the center, and the rears are either discrete parts of the music, or pure ambience. It works well in my room at least.

I've got a receiver that handles all the popular 3D surround formats: Atmos, DTS:X, Auro-3D. I like Auro-3D for upmixing 5.1 Dolby Digital. Some of my hi-res BD blu-ray concerts get defaulted to their source DTS-MA channels (and I think it's OK for them to have a 2-D surround plane). I have some UHD discs that are DTS:X, but Atmos has become the most popular format for streaming and overall UHD content. What is interesting is that the *ideal* speaker configurations are slightly different for every 3D format. With Atmos, it's initially designed for either reflected sound off the ceiling or ceiling mounted speakers. DTS:X and Auro-3D can use height speakers (my receiver will also mix Atmos for overhead effects pretty effectively). Auro also recommends having one ceiling speaker directly above you ("voice of god") for what they show as a dome of sound (your 2-D surrounds, then a layer of height speakers, then the top channel for the top dome). I got height speakers because it would be very difficult to install ceiling speakers in my room (that has another floor on top of it). But I find movies are pretty immersive and I can hear overhead effects with various 3D modes. It was worth it to upgrade from 7.1 to 7.1.4 (especially since more content is streaming Atmos).

I love my 5.1 speaker system. If I could find a way to reproduce that in headphones, it would be great. I've heard Atmos headphone mixes and binaural recordings. But none of them are able to reproduce the pinpoint sound location of the soundstage of my speaker system. It can sound kind of the same in the response, but not in directionality. The sound is always right there inside or just around my head. It's never ten or fifteen feet in front of me or behind me.

I've mentioned before that the most realistic headphone surround I've heard is an out of production Sennheiser Dolby Pro Logic processor. I could hear movie effects sweep from front sides to all the way in back. I didn't hear a clear center front dialogue (the way a speaker theater setup is), nor do I think that headphone processing can address the height effects new speaker systems have.

 

[bigshot]

@bigshot, I wouldn't say 5.1 is three dimensional.

I agree. It's like 2 1/2 dimensions at best. It's a flat plane from front to back and left to right. Atmos is true 3D. Three dimensions requires up/down too. I suppose stereo is two dimensions... a flat plane 10 feet in front of the listener. Headphones are one dimensional- a line straight through the ears. DSPs can improve it, but they don't add true dimensionality. They just take the curse off the one dimensionality.

In any case, anyone who is familiar with multichannel mixes for music knows how important it can be, especially in the last two or three years. It not only reproduces the music, it places it within physical space and it can create ambiences that open up the normal ambience of the room. Like any other kind of recording, that provides opportunities to push beyond the limits of just capturing a performance and actually makes it possible to craft a dimensional performance that doesn't mimic any real acoustic. I'd be happy to suggest recordings that take full advantage of this if anyone is interested in investigating it.

What is interesting is that the *ideal* speaker configurations are slightly different for every 3D format.

4.0 is an entirely different approach than modern multichannel. Quad aims for sound coming equally from the four corners of the room, 5.1 aims at creating a coherent front soundstage with separate rear channels. However the newest Atmos recordings when down mixed to 5.1 have been attempting to create a phantom center between front and back- pretty startling. It requires much more precision in speaker placement and calibration, but when it works, it's revelatory, because it creates a relationship between front and back that equals left and right.

With Atmos, it's initially designed for either reflected sound off the ceiling or ceiling mounted speakers. DTS:X and Auro-3D can use height speakers (my receiver will also mix Atmos for overhead effects pretty effectively). Auro also recommends having one ceiling speaker directly above you ("voice of god") for what they show as a dome of sound (your 2-D surrounds, then a layer of height speakers, then the top channel for the top dome). I got height speakers because it would be very difficult to install ceiling speakers in my room (that has another floor on top of it). But I find movies are pretty immersive and I can hear overhead effects with various 3D modes. It was worth it to upgrade from 7.1 to 7.1.4 (especially since more content is streaming Atmos).

Yeah, that's the problem. The vertical dimension requires a particular room setup that most living rooms can't handle. I have a peaked roof with beams, and putting in Atmos would be very difficult to manage. The more channels you add, the better it gets, but the harder it is to implement. You have to strike a compromise for what is possible until you have the opportunity to create a totally optimized space.

I've mentioned before that the most realistic headphone surround I've heard is an out of production Sennheiser Dolby Pro Logic processor. I could hear movie effects sweep from front sides to all the way in back. I didn't hear a clear center front dialogue (the way a speaker theater setup is), nor do I think that headphone processing can address the height effects new speaker systems have.

I don't know for sure, but I think a lot of headphone surround depends on how your particular noggin shapes equate with the average. For some people, it works; but for others it never works. I have struggled to try to get binaural to work for me, but it never locks in. At best, it flickers from three inches in front of my head to three inches behind it. Other people may have no trouble. But everyone can hear the benefits of a well tuned multichannel speaker setup.

 

[gregorio]

[1] How do you know what happens in my head? You can only speak about your head. ...
[2] Listening to music that is not processed with a crossfeed algorithm sounds unnatural and annoying, especially bass.
[3] Yes, it does approximate speakers to a certain degree. Many crossfeed plugins have these additional features: ... This is an example ...
[4] This thread is completely hi-jacked by crossfeed-haters.

1. Agreed.

2. No it doesn't! Don't you really mean that to YOU it "sounds unnatural and annoying, especially bass"? If you do not explicitly specify that is how it sounds to YOU PERSONALLY, then you are making a blanket statement that covers everyone (including me). And the obvious response to that assertion is: "How do you know what happens in my head? You can only speak about your head.". You are therefore effectively contradicting yourself! ... Just to be clear, despite the fact I've stated it a number of times: Listening to music on headphones without crossfeed does not sound natural to me but listening with crossfeed does not sound natural to me either. Furthermore, listening with speakers/monitors does not sound natural to me either and the reason none of the playback scenarios sounds natural to me is because none of it is natural! Stereo itself is an illusion (that isn't natural) and the music recordings themselves are not natural. Almost without exception, music recordings are made by layering multiple different aural locations and/or perspectives, which would only be natural if you had multiple different ears which were simultaneously in different locations.

3. That is NOT an example of a crossfeed plugin! Crossfeed is the act of taking the signal from one channel and "feeding" it (or part of it) to the other channel. The example you gave would only be a crossfeed plugin if you turned off all those "additional features", IE. Turn off the HRTF and "Room Designer" features. Your example is INCORRECT, it is NOT an example of a crossfeed plugin, it is an example of a "Binaural Room Simulator" plugin!

4. Firstly, I wouldn't say I'm a crossfeed hater. It doesn't work for me personally, I personally don't like it and I choose not to use it but I don't hate it or state that everyone else should/must hate it too. Secondly: So, instead of the thread being "hi-jacked by crossfeed haters", you want to hi-jack it as a Binaural Room Simulator fan boy? The thread title is NOT "to-binaural room simulate-or-not ..."!

My system is 5.1, but I'm told that Atmos allows you to place a sound object anywhere within left right, front back and top bottom parameters. That would be true 3D.

That would be true 3D but Atmos does NOT allow you to do that and therefore is NOT true 3D. Atmos allows you to place a sound anywhere in the horizontal plane (left/right, front/back) and can theoretically do that more precisely than 5.1 (although only in a cinema), as it it less reliant on the stereophonic illusion (phantom positioning). Atmos additionally provides ceiling speakers, allowing for height information BUT, with two limitations:

Firstly, the soundfield is effectively between the existing horizontal plane of (5.1) speakers and the higher vertical plane of the ceiling speakers. Using your terminology, Atmos allows you to place a sound object anywhere within left right, front back and top middle parameters! NOT within "top bottom parameters", a sound cannot be positioned below the horizontal plane of the (existing 5.1) speakers. In other words, Atmos theoretically provides a hemispherical soundfield, rather than the spherical soundfield that would be required for "true 3D". I know Atmos (and other similar formats) are marketed as 3D sound formats but that's just marketing, a bit like unlimited data plans that are limited!

Secondly, positioning a sound anywhere within that hemispherical soundfield is arguably only theoretically possible rather that possible in practice. This is because there are many positions within that hemisphere which would effectively rely on a double phantom position; a phantom position between the horizontal plan of speakers, PLUS a phantom position between that horizontal phantom position and a vertical phantom position. In practice, that would be a highly unstable position, even if all the speakers were perfectly phase aligned at a particular listening position, a stereophonic illusion on top of a stereophonic illusion could change (or be entirely destroyed) by a relatively small change in head position relative to any of the speakers. On the other side of the coin, would we actually perceive a change or destruction of this unstable illusion? Our localisation ability in the vertical plane is many times weaker than in the horizontal plane anyway. My educated/experienced guess would be that it would depend on several factors, probably the most important of which is how/if that aural illusion is supported by our vision (IE. Is there a visual cue in the film which indicates the position of the sound and if so, how obvious is it). Another factor would be our position relative to ALL the speakers and another would be an individual's perception. Due to the fact that this localisation may not be perceived as intended by significant portions or the audience and the potential phase issues it could create, these positions within the hemispherical soundfield are typically avoided when creating a mix (that is, typically avoided for static positions, "passbys", sounds passing through those positions are fine).

G

 

[Davesrose]

4.0 is an entirely different approach than modern multichannel. Quad aims for sound coming equally from the four corners of the room, 5.1 aims at creating a coherent front soundstage with separate rear channels. However the newest Atmos recordings when down mixed to 5.1 have been attempting to create a phantom center between front and back- pretty startling. It requires much more precision in speaker placement and calibration, but when it works, it's revelatory, because it creates a relationship between front and back that equals left and right.

I never referred to quad sound. I referred to blu-ray concerts, which often have an option for stereo or 92khz DTS MA 5.1. I’ll keep it the default 5.1 (2D surround plane) as who needs phantom height in a concert setting? I haven’t tried an Atmos receiver with 5.1 or 7.1. It is interesting that DTS and Auro have surround schemes that try to virtually add height speakers. I’ve tried Auro’s mode: doesn’t sound nearly as engaging as having actual height speakers. Having physical speakers will always be ideal. For all of us that don’t have optimal home theater rooms, we just have to find work around.

 

[bigshot]

Yeah, the whole thing about speakers is that it is real physical sound in real physical space. I don't think that can really be synthesized. I've jury rigged my system to add height by raising the center channel above the level of the sides, and raising the level of the rears. This helps pull the soundstage up to fill the screen, which is about 9 feet tall at the top. Without that, the sound seems like it is lower than the action on the screen. I would add height speakers, but they would look ugly in my particular room, and they would probably fire directly into a solid wood beam.

Based on my experience with 5.1, I'm guessing that the benefit of Atmos isn't just making airplanes fly over your head in Top Gun, but it can help create better sound location in the middle of the sound field. It's taken me a lot of work to get my 5.1 system to mesh front to back with a phantom center in the middle of the room. Sound can cross the room diagonally from front left to rear right without a dip in the center. But if I had a half dozen overhead channels, it might be possible for sound to do arcs through the middle of the room, or to locate something in the sound field other than along an x across it. (I'm not sure if I'm describing that clearly...) Another benefit of having overhead channels is creating synthetic room ambiences that are much bigger than your actual room. An organ concert in a cathedral for example, or open air in a field like the opening of the Sound of Music. I can see it allowing much more sophisticated and varied ambiences.

I keep thinking about how I could incorporate Atmos, hiding speakers behind beams... but everything about it would be a pain and I wouldn't have a lot of flexibility to experiment with speaker placement. I still may do it someday. I think it would be a definite notch up in quality of sound field. The reason I mentioned quad is because quad usually doesn't even attempt creating a sound field. It's just individual channels coming from the four corners of the room. A lot of 5.1 is like that too, but the newer immersive mixes are more exciting to me.

can theoretically do that more precisely than 5.1 (although only in a cinema), as it it less reliant on the stereophonic illusion (phantom positioning).

Yes, the more channels you have, the less dependent you are on phantom centers. It's like the center channel in 5.1... it replaces the need for a phantom center up front,. If it is used to be a part of the front image, rather than a separate sound like vocals, you can create a wider soundstage, because you can increase the width between the mains and create phantoms between the center and the left and right. You're right that Atmos's height is basically ear level up, not ear level down. And depending on how many Atmos speakers you are running, there would be dead spots. But it would be light years better than having no vertical dimension at all. Maybe it would be best to call it 2.8 D?

I have an old Cinemascope movie where the lead actor comes out at the beginning and walks in from stage right and crosses to exit stage left talking the whole time. They mixed his dialogue to hand off to the speakers behind the screen to give the voice a pinpoint location. It creates a very interesting effect. And I have recent Atmos music that is mixed for sound to fly diagonally across the room. Even in 5.1 it can work. The Kraftwerk Catalogue blu-ray is in both 3D picture and 3D Atmos sound. I don't do 3D but I've heard it is really astounding on a good system.

 

[Davesrose]

Yeah, the whole thing about speakers is that it is real physical sound in real physical space. I don't think that can really be synthesized. I've jury rigged my system to add height by raising the center channel above the level of the sides, and raising the level of the rears. This helps pull the soundstage up to fill the screen, which is about 9 feet tall at the top. Without that, the sound seems like it is lower than the action on the screen. I would add height speakers, but they would look ugly in my particular room, and they would probably fire directly into a solid wood beam.

Based on my experience with 5.1, I'm guessing that the benefit of Atmos isn't just making airplanes fly over your head in Top Gun, but it can help create better sound location in the middle of the sound field. It's taken me a lot of work to get my 5.1 system to mesh front to back with a phantom center in the middle of the room. Sound can cross the room diagonally from front left to rear right without a dip in the center. But if I had a half dozen overhead channels, it might be possible for sound to do arcs through the middle of the room, or to locate something in the sound field other than along an x across it. (I'm not sure if I'm describing that clearly...) Another benefit of having overhead channels is creating synthetic room ambiences that are much bigger than your actual room. An organ concert in a cathedral for example, or open air in a field like the opening of the Sound of Music. I can see it allowing much more sophisticated and varied ambiences.

I keep thinking about how I could incorporate Atmos, hiding speakers behind beams... but everything about it would be a pain and I wouldn't have a lot of flexibility to experiment with speaker placement. I still may do it someday. I think it would be a definite notch up in quality of sound field. The reason I mentioned quad is because quad usually doesn't even attempt creating a sound field. It's just individual channels coming from the four corners of the room. A lot of 5.1 is like that too, but the newer immersive mixes are more exciting to me.



Yes, the more channels you have, the less dependent you are on phantom centers. It's like the center channel in 5.1... it replaces the need for a phantom center up front,. If it is used to be a part of the front image, rather than a separate sound like vocals, you can create a wider soundstage, because you can increase the width between the mains and create phantoms between the center and the left and right. You're right that Atmos's height is basically ear level up, not ear level down. And depending on how many Atmos speakers you are running, there would be dead spots. But it would be light years better than having no vertical dimension at all. Maybe it would be best to call it 2.8 D?

I have an old Cinemascope movie where the lead actor comes out at the beginning and walks in from stage right and crosses to exit stage left talking the whole time. They mixed his dialogue to hand off to the speakers behind the screen to give the voice a pinpoint location. It creates a very interesting effect. And I have recent Atmos music that is mixed for sound to fly diagonally across the room. Even in 5.1 it can work. The Kraftwerk Catalogue blu-ray is in both 3D picture and 3D Atmos sound. I don't do 3D but I've heard it is really astounding on a good system.

When I upgraded my surround system to HDMI lossless, the way my room was setup, 7.1 seemed like a better setup. I placed the rear speakers a bit higher then ear level, and the surrounds are just to the sides (I thought it better because my viewing area is an open main floor with vaulted ceilings and one of the surrounds being mounted in a bar area against the kitchen area). All the speakers were the same series (tower mains, center, and bipole surrounds). For 5.1 tracks or 7.1, surround effects were good and seemless for having aircraft and arrows having a sense of coming from the screen to the sides and up rear when it hit back surrounds. I started collecting UHD discs before I upgraded to UHD (knowing I should futureproof). One aspect I liked about home movies appearing with Atmos, was that there was more tracks that were 7.1. So I've directly compared Bladerunner 2049 and Last Jedi in 7.1 lossless (often downstreamed from Atmos now) vs now 7.1.4. Certainly scenes like helicopters and remote drones flying overhead are the most convincing aspect of 3D surround. I think my back heights are not quite as effective as the front heights, since my back surrounds are above ear level. However, my height speakers are high up and do point down towards the top of my head. So the combination of front and rear heights do make more of a sense of overhead effects. Also RE sound below you....it's true that "3D surround" is intended as a dome (that depending on speaker config can be slightly below ear level to complete dome above). But it's interesting that the opening scene of Blade Runner (Final Cut, has been remixed to Atmos), I've heard a spinner seem to come from somewhere above center to left bottom. It might be the partial sound mix of sharply panning from height to 2D level, and partly perception of seeing the spinner go off-screen on the bottom.

When it comes to figuring out height speakers, I can give good marks to the height speakers I got. They are SVS Elevations. They're slanted and have the woofer driver in the larger portion of the loudspeaker (which seems like a better design to me: for crossover and focusing tweeter downwards). The mounting bracket allows the speaker to be mounted in any orientation. My 7.1 speakers are a different brand that's pretty much gone now...but they all seem to match well (even if I have room correction off on my receiver).

 

[bigshot]

Your room sounds quite similar to mine.