Now That Atmos Is Everywhere… Real vs. Phantom Center in a 5.1 Music-Focused Setup

[Newman]

No. At that point It can be believed, taken at face value or ignored.
But, if you want to refute it, it is up to you to prove them different.

eg if the court prosecutor claims that I committed a crime, and I refute it, I don't have to prove him wrong. He has to prove his claim is true.

They said you did it. You say you didn't do it. And THEN it is up to them to refute what you said: they then have to prove you did.

And then there are the folks that say the world is flat (I know one and he ran radio stations, so you never know). It doesn't matter what proof you provide, they still deny it.

I'm not sure which of my statements you are disagreeing with, and which you are agreeing with. I simply outlined the basic scientific principle of burden of proof. Do you not agree with that?

cheers

 

[EJ3]

I too digitised what I had, still collect bluray audio releases from time to time. I just do not want to miss new stuff- and there is so much good music still being made.

But you have that covered with radio - and there is something very relaxing about listening to the radio sometimes.

And, I sometimes find something "new to me" that I want on physical media (whether it is new or old) by listening to the FM band.

 

[EJ3]

I'm not sure which of my statements you are disagreeing with, and which you are agreeing with. I simply outlined the basic scientific principle of burden of proof. Do you not agree with that?

cheers

From what I can tell: I think that we are saying the same thing differently.

 

[BJL]

Yes a little, sorry.
With Atmos I find that a lot of vocals and that which before would be panned "hard center" has now been floated to a point further up vertically.
So now we have that info being shared between the center (real or phantom) and the front height speakers.
Works well either way but probably still best to use a real center for multich systems having a discreet center channel. JMHO

So far as I've noticed, all of the Atmos blu ray audio that I own take advantage of the center channel position, though not always every single track, and not exclusively for vocals. Some examples that use the center in interesting ways: Max Cooper Emergence & Unspoken Words (both instrumental albums); Tony Levin Bringing it Down to the Bass; Mark Knopfler One Deep River; Booka Shade all of their Atmos releases; Yello Touch and Point; many others.

For the recordings mentioned, a phantom center would impair the positional accuracy of the mix. To reproduce the intended mix, a center channel speaker is needed. Of course, if finances or physical room issues doesn't allow a center channel, that is just one of those compromises one may have to make. This may be less important for simple pop/rock recordings.

 

[Matt S]

Max Cooper's production quality - pristine.. much like Jon Hopkins. Those two are experts in the electronic genre, the best IMO.

 

[Sal1950]

So far as I've noticed, all of the Atmos blu ray audio that I own take advantage of the center channel position, though not always every single track, and not exclusively for vocals. Some examples that use the center in interesting ways: Max Cooper Emergence & Unspoken Words (both instrumental albums); Tony Levin Bringing it Down to the Bass; Mark Knopfler One Deep River; Booka Shade all of their Atmos releases; Yello Touch and Point; many others.

For the recordings mentioned, a phantom center would impair the positional accuracy of the mix. To reproduce the intended mix, a center channel speaker is needed. Of course, if finances or physical room issues doesn't allow a center channel, that is just one of those compromises one may have to make. This may be less important for simple pop/rock recordings.

BTW, Some great Atmos mixes mentioned there.

 

[RDoc]

So far as I've noticed, all of the Atmos blu ray audio that I own take advantage of the center channel position, though not always every single track, and not exclusively for vocals. Some examples that use the center in interesting ways: Max Cooper Emergence & Unspoken Words (both instrumental albums); Tony Levin Bringing it Down to the Bass; Mark Knopfler One Deep River; Booka Shade all of their Atmos releases; Yello Touch and Point; many others.

For the recordings mentioned, a phantom center would impair the positional accuracy of the mix. To reproduce the intended mix, a center channel speaker is needed. Of course, if finances or physical room issues doesn't allow a center channel, that is just one of those compromises one may have to make. This may be less important for simple pop/rock recordings.

It's interesting to me you mention this elevation of apparent source in many ATMOS tracks but not all. I've noticed that as well, but thought it was just my imagination. Our bed speakers are all level and at ear level and toed in as needed to cover the listening area. I wonder if they are deliberately mixed to emulate a raised stage?

 

[BJL]

It's interesting to me you mention this elevation of apparent source in many ATMOS tracks but not all. I've noticed that as well, but thought it was just my imagination. Our bed speakers are all level and at ear level and toed in as needed to cover the listening area. I wonder if they are deliberately mixed to emulate a raised stage?

The von Karajan Atmos Beethoven symphonies set on DG does this, though subtly. The orchestra appears slightly elevated giving the impression that one is listening in the concert hall. It's quite lovely. This mix uses all of the Atoms positions, including front wide and top middle. It also uses the center speaker, at a low level, apparently as fill, at least as far as I have listened. Given the choice I would always purchase orchestral music in an Atmos mix. Stereo sounds dull to me, in comparison. The blu ray edition is still available in the US (and I presume elsewhere) as of March 2026.

 

[audiomaestro]

I think there’s an important distinction that depends a lot on listening distance and room type. I’m running a true nearfield setup with Neumann KH 120 II mains and KH 80 surrounds, calibrated with MA-1. I sit at 5–7 ft in a treated room. In this specific geometry, the phantom center from the matched KH 120 II pair is clearly superior to adding a real center channel.

At 5–7 ft, even a relatively small vertical mismatch becomes pretty obvious. If I put a third KH 120 II on top of my TV, the tweeter would be over 15 inches higher than the L/R tweeters. That creates a noticeable upward pull on dialogue and centered sounds, which is way more distracting than the mild comb-filter dip you get from a good phantom center.

The MA-1 calibrated KH 120 II pair gives me perfect timbre match, phase coherence, and time alignment between left and right. The phantom center just locks in and sounds focused. A physical center sitting higher and a bit deeper introduces vertical localization errors and early reflections off my TV.

In bigger living-room or far-field setups (10–15+ ft away, wider seating), a real center often makes more sense because the vertical mismatch is proportionally smaller and off-axis listening is common. But in a nearfield room like mine, the opposite is true. The phantom from two identical, perfectly calibrated monitors ends up cleaner and more accurate.

Floyd Toole’s work is excellent, but most of it comes from typical living rooms with far-field listening. In a treated nearfield room at 5–7 ft, the trade-offs flip. The mild dip from a good phantom is far less problematic than a large vertical mismatch on a physical center.

I was actually planning to add a center after reading this thread, but after doing more research and listening the phantom center is the way.

 

[Newman]

I think there’s an important distinction that depends a lot on listening distance and room type. I’m running a true nearfield setup with Neumann KH 120 II mains and KH 80 surrounds, calibrated with MA-1. I sit at 5–7 ft in a treated room. In this specific geometry, the phantom center from the matched KH 120 II pair is clearly superior to adding a real center channel.

At 5–7 ft, even a relatively small vertical mismatch becomes pretty obvious. If I put a third KH 120 II on top of my TV, the tweeter would be over 15 inches higher than the L/R tweeters. That creates a noticeable upward pull on dialogue and centered sounds, which is way more distracting than the mild comb-filter dip you get from a good phantom center.

The MA-1 calibrated KH 120 II pair gives me perfect timbre match, phase coherence, and time alignment between left and right. The phantom center just locks in and sounds focused. A physical center sitting higher and a bit deeper introduces vertical localization errors and early reflections off my TV.

In bigger living-room or far-field setups (10–15+ ft away, wider seating), a real center often makes more sense because the vertical mismatch is proportionally smaller and off-axis listening is common. But in a nearfield room like mine, the opposite is true. The phantom from two identical, perfectly calibrated monitors ends up cleaner and more accurate.

Floyd Toole’s work is excellent, but most of it comes from typical living rooms with far-field listening. In a treated nearfield room at 5–7 ft, the trade-offs flip. The mild dip from a good phantom is far less problematic than a large vertical mismatch on a physical center.

I was actually planning to add a center after reading this thread, but after doing more research and listening the phantom center is the way.

The TV is the problem, though, right? If it wasn't there, a same-height centre channel would be clearly superior to a phantom centre, even at 5 to 7 ft. At least, that is what the perceptual research is telling us.

You are in essence prioritising video over audio, which is fine, but the resultant audio compromise is what it is.

cheers

 

[krabapple]

So let's just take a moment to think about the implications of this, and why so many "golden ears audiophiles" swear by the apparently increased presence and fidelity of DSD recordings (from reputable sources that have refrained from "mastering" the tracks).

Let's take another moment to think about the many, many 'golden ear audiophiles' who swear the same about the many, many, many SACDs that are not fully DSD from start to finish, i.e., every SACD release of an old analog recording.

Bottom line: the effects of converting any recording source (say, stereo or other two-channel recording) to multichannel--the first thing that is required is to alter the phase of the incoming music tracks to "synthesize" both center channel and surround tracks that were not actually recorded. This means that the "circle of confusion" that Toole talks about so often, as well as Sean Olive, is well embedded in any of these efforts to synthesize surround recordings, and that these phase shifts go a long way toward compromising the fidelity/presence of the recordings--directly the opposite result that the whole effort is striving to achieve through post-mastering efforts.

And yet...Floyd Toole himself is a big fan of upmixing and says he uses it routinely.

 

[krabapple]

For DPLII set soundstage -> Front , Panorama Off, Dimension -> Front, Center Width -> Max

?
Center Width -->Max reverts you entirely to phantom center.

Which I think we have established here, is objectively problematic.

The default for DPL II Center width -- '3', IIRC -- which was the most recommended -- also is not a pure Center setting (which is '0').

This is also a paradox with DSU -- everyone was upset when Center Spread was deactivated (it was later reinstated, after protests from e.g., Gene Della Salla at Audioholics)
but it, too, allows user to deviate from a wholly centered Center (i.e., all common L/R content is steered to the center), to one where 'C' content is shared to a degree with L and R

 

[krabapple]

Well, he seems to be saying that the whole 'bed' for Atmos, that's 7 speakers, needs to be identical and cover 20-20kHz with their own co-located subwoofers. That contravenes any experimental evidence I have seen.

The part I bolded is optimal. But not often practical. Many of us rely on DSP to 'tone match' -- which is kind of flying on a wing and a prayer. (And it's on top of whatever DSP needs to be done due to room acoustics, which is going to complicate things even if speakers are identical, simply due to different positioning)

As for the rest in italics, IIRC mix suites for 5.1 used 5 identical equidistant full range towers (plus one subwoofer for LFE) if they were industry-standards-compliant.

What ATMOS suites should comply to, I can't say. I have the Dolby pdf guide for professionals saved somewhere...

 

[krabapple]

Read some then skipped a lot. Why worry about Atmos with a 5.1 setup?

One interesting reason is : what exactly happens to Side content of a 7.1 'Atmos mix' (or any mix using Side channels) when it's downmixed to 4 or 5 'bed' channels (i.e., 2 front, 2 surround positioned to the rear of the MLP, center optional)?

Conceptually, Side content should be evenly split, on each side, between front and surround -- forming a phantom side image(s) between them.

Is that what happens in practice?

 

[krabapple]

Steven Wilsons incredible talent in both original mastering and the remixing of older material is also what being used to populate the streaming catalogs.

Again: Steven Wilson pretty famously does not master his remixes at all. He prefers them to stay 'unmastered'.
(AFAIK this is true of his own solo music as well.)

If mastering takes place for final release, for either his stereo or surround mixes, it's done by someone else.

Unfortunately at present, lossy compression is being used to stream the large amount of data contained in lossless versions.

More audiophile nervosa. 'Data' doesn't mean 'audible when listening in surround': so says psychoacoustics.

If you're hearing difference between lossy streams and lossless product, it's far more likely an output level difference, or an EQ difference, or a mix difference. None of which are due to lossy data compression per se.

 

[krabapple]

The best way to keep one's hat on is to remember that lossy compression is effectively inaudible above a certain bitrate. Atmos audio at 768 kbps is considered transparent within the industry and academia.

It's not even just a matter of bitrate. As listening channels are added, perception of artifacts goes down.

 

[krabapple]

In this AB testing, four people scored 10/10, so the difference was clearly audible.

Hmm, not necessarily.

Anyway, discussion of test rigor and statistics is tedious, has been done well elsehwere on ASR, and isn't very on point to this thread.

 

[krabapple]

I think there’s an important distinction that depends a lot on listening distance and room type. I’m running a true nearfield setup with Neumann KH 120 II mains and KH 80 surrounds, calibrated with MA-1. I sit at 5–7 ft in a treated room. In this specific geometry, the phantom center from the matched KH 120 II pair is clearly superior to adding a real center channel.

At 5–7 ft, even a relatively small vertical mismatch becomes pretty obvious. If I put a third KH 120 II on top of my TV, the tweeter would be over 15 inches higher than the L/R tweeters. That creates a noticeable upward pull on dialogue and centered sounds, which is way more distracting than the mild comb-filter dip you get from a good phantom center.

? The proper comparison to determine what is 'cleaner and more accurate' is to use an identical center mounted at the same height as the L/R, and of course calibrated for distance and level.

Barring that, have you tried angling your TV-mounted center to fire more directly at you (and also calibrating that)?

In any case the main factor here seem to be the existence of a TV, as it is for me.
One either finds a satisfying workaround using a center speaker (e.g., raise the TV height, angle the center), or 'settles' for phantom.

 

[sfogg]

?
Center Width -->Max reverts you entirely to phantom center.

Which I think we have established here, is objectively problematic.

The default for DPL II Center width -- '3', IIRC -- which was the most recommended -- also is not a pure Center setting (which is '0').

This is also a paradox with DSU -- everyone was upset when Center Spread was deactivated (it was later reinstated, after protests from e.g., Gene Della Salla at Audioholics)
but it, too, allows user to deviate from a wholly centered Center (i.e., all common L/R content is steered to the center), to one where 'C' content is shared to a degree with L and R

Yes, I was giving settings to make PLII about as subtle as it can be in that post. Some people have a really hard time accepting the difference in sound when they loose the comb filtering between L/R when utilizing a center speaker. So for those listeners setting PLII to phantom center wouldn't 'break' them.

That isn't how I would use PLII as I strongly prefer a well done center speaker.

 

[sfogg]

One interesting reason is : what exactly happens to Side content of a 7.1 'Atmos mix' (or any mix using Side channels) when it's downmixed to 4 or 5 'bed' channels (i.e., 2 front, 2 surround positioned to the rear of the MLP, center optional)?

Conceptually, Side content should be evenly split, on each side, between front and surround -- forming a phantom side image(s) between them.

Is that what happens in practice?

I don't think you can get a very good phantom side just by splitting between say left front and left rear. HRTFs and any timing differences will throw off the side image. Especially. if you are trying to have multiple seating positions. This booklet talks about the importance of lateral sound in concert hall and room acoustics quite a bit but especially on pages 4-8.

I'd think a better approach in a 5 speaker system is to put the surrounds at the sides and then try and simulate rears with HRTF processing on the rear info fed to the side speakers. Lexicon's used to do this with a feature called 5 speaker enhance.