Why do we need a center channel in a home theater system?

[Sancus]

IIRC F. Toole also prefers multi-channel setup to pure stereo as it provides the most realistic presentation with all the necessary spacial/reverberation cues lacking in stereo. What I would like to know is whether Mr. Toole prefers stereo upmixed (ie 9 channel Auromatic) over native stereo?

Yes, he does prefer upmixed to native stereo in most cases using Auro3D based on his posts here. Here's another good one. His comments are what lead me to investigate it. Before reading his comments, I was under the impression that upmixing was a gimmick and that stereo was the "correct" way to listen, because that's what I'd heard from many "audiophile" sources.

Needless to say, that attitude is totally wrong and I'm thankful for him calling out the value of surround music and upmixing. It's definitely made a large contribution to my enjoyment of music.

 

[North_Sky]

It arguably is only if you insist on having one.

Yeah, the center channel is very important. That doesn't mean you need a dedicated speaker for it. the center channel in stereo mixes are distributed between the left and right and no one complains about it.

I said in a "home theater system setup".
In a stereo system setup most people don't use a center channel speaker, so the most important speaker is the front Left one, and the front Right one, both equally and preferably within less than 0.5dB level deviation between them when out of the factory...margin. ...That for a quality made pair of loudspeakers.

Frank Sinatra was into 3-channel stereo (with center) music listening.
Frank was ahead of his time.

In 1969 I built a speaker to use as a mono surround channel, right behind the listener.
Just for the fun of it, and it was, and it still is but with more now...four (sides and back stereo), plus four more above, also in stereo pairs. ...So eight surround speakers.
You think I would dismiss the front center channel speaker in that multichannel home theater system setup? Think again.

Déjà vu ...

 

[StevenEleven]

Yes, he does prefer upmixed to native stereo in most cases using Auro3D based on his posts here. Here's another good one. His comments are what lead me to investigate it. Before reading his comments, I was under the impression that upmixing was a gimmick and that stereo was the "correct" way to listen, because that's what I'd heard from many "audiophile" sources.

Needless to say, that attitude is totally wrong and I'm thankful for him calling out the value of surround music and upmixing. It's definitely made a large contribution to my enjoyment of music.

Similarly, he discusses the subject matter at length in Chapter 15 ("Multichannel Audio") of his book. It's a great Chapter, taking you from the days of monophonic audio through the current state of affairs. FWIW, at page 428, he states, "Personally, I enjoy upmixed stereo for many, but not all, programs. . . Switching back to stereo results in a diminished sense of envelopment and a shrinking of the soundstage." Before you get to that explicitly stated preference, though, it is pretty clear that he is favorably disposed to upmixing.

I had gotten used to upmixing most of the time but figured I was some kind of lowlife for doing so. This made me a lot more comfortable in my preferences.

In the book he hits much of the stuff that's being discussed here in a more comprehensive fashion, with citations to and descriptions of published research, etc. Makes for an educational, interesting, and entertaining read.

As a matter of fact a lot of the stuff that people argue over here in this forum is asked & answered in his book, including research as to the questions posed.

 

[Bear123]

Subs are never a waste. Unless your speakers can output with authority down to 20 Hz and below, and if that's the case never mind.

Actually, it doesn't matter if the speakers play down to 20 Hz, 10 Hz, or 5 Hz. Pairing them with subs is still better. Two speakers without subs cannot achieve the same sound quality. This isn't directed necessarily at you, but its very surprising that so many people, especially on this site, don't realize that all speakers should be crossed to a pair of good subs if high fidelity sound quality is the goal. I actually see posts where people *turn off their subs and use a different amp* to listen to music. This is beyond hilarious that anyone would downgrade sound quality to adhere to an old fashioned, outdated, low fidelity implementation(*pure* 2.0 channel)

 

[Kal Rubinson]

In general a center would allow you to space out your frontstage speakers more without losing the center image.

That is true, particularly as you used the work "allow" instead of "require." In a number of tests, I compared the sound of the 3channel DSD tracks with the 2channel DSD tracks from several RCA Living Stereo SACDs. This was not blind but participants were asked what they anticipated the change would be and the consensus was that there would be better center fill, center detail and/or center depth. After listening, almost all participants described the difference as a widening of the soundstage with the 3channel tracks.

Similarly, he discusses the subject matter at length in Chapter 15 ("Multichannel Audio") of his book. It's a great Chapter, taking you from the days of monophonic audio through the current state of affairs. FWIW, at page 428, he states, "Personally, I enjoy upmixed stereo for many, but not all, programs. . . Switching back to stereo results in a diminished sense of envelopment and a shrinking of the soundstage."

There you go.

 

[andymok]

There a reason why a Decca Tree has 3 mics instead of 2...

 

[North_Sky]

The OP might be interested in that older article from 2018 if he hasn't read it yet ..,
• http://archimago.blogspot.com/2018/11/musings-multidimensional-audio-old.html

 

[XaVierDK]

Actually, it doesn't matter if the speakers play down to 20 Hz, 10 Hz, or 5 Hz. Pairing them with subs is still better. Two speakers without subs cannot achieve the same sound quality. This isn't directed necessarily at you, but its very surprising that so many people, especially on this site, don't realize that all speakers should be crossed to a pair of good subs if high fidelity sound quality is the goal. I actually see posts where people *turn off their subs and use a different amp* to listen to music. This is beyond hilarious that anyone would downgrade sound quality to adhere to an old fashioned, outdated, low fidelity implementation(*pure* 2.0 channel)

That's very true, especially when considering room acoustics. Optimal bass rendition requires room-aware placement of speakers, and placing your towers/bookshelves optimally for bass in most cases would ruin the rest of your sound.
With subs these can be placed independently to optimize for room geometry.

 

[SarahSimm]

Yes, he does prefer upmixed to native stereo in most cases using Auro3D based on his posts here. Here's another good one. His comments are what lead me to investigate it. Before reading his comments, I was under the impression that upmixing was a gimmick and that stereo was the "correct" way to listen, because that's what I'd heard from many "audiophile" sources.

Needless to say, that attitude is totally wrong and I'm thankful for him calling out the value of surround music and upmixing. It's definitely made a large contribution to my enjoyment of music.

Similarly, he discusses the subject matter at length in Chapter 15 ("Multichannel Audio") of his book. It's a great Chapter, taking you from the days of monophonic audio through the current state of affairs. FWIW, at page 428, he states, "Personally, I enjoy upmixed stereo for many, but not all, programs. . . Switching back to stereo results in a diminished sense of envelopment and a shrinking of the soundstage." Before you get to that explicitly stated preference, though, it is pretty clear that he is favorably disposed to upmixing.

I had gotten used to upmixing most of the time but figured I was some kind of lowlife for doing so. This made me a lot more comfortable in my preferences.

In the book he hits much of the stuff that's being discussed here in a more comprehensive fashion, with citations to and descriptions of published research, etc. Makes for an educational, interesting, and entertaining read.

As a matter of fact a lot of the stuff that people argue over here in this forum is asked & answered in his book, including research as to the questions posed.

well that's actually a good comment here. All the truth is said.
guess OP did listen to it, huh?

 

[dasdoing]

upmixing stereo to 3.0/3.1/3.2 would be best for keeping center at the center. but not the way film is using the center. center speaker needs to produce panned sound, too. for example, a 75 left pan needs to come out equaly out of left and center speaker. I have not found a upmix matrix doing that.
on the other hand, I don't know how I could fit a center speaker into my system as the screen is in the way

 

[Aerith Gainsborough]

I have experimented with Center speakers but I found them to be superfluous as long as I am the only one watching in a nearfield setup.
My 906s render the image perfectly so I do not notice a difference w or w/o center.

I'd probably still want one if I were to put multiple people on the couch for community watching though.
Imaging is more stable with a real sound source in the middle.

 

[Alexium]

I have noticed when listening to stereo music on a 5.0 system that in terms of left to right pan, I get much better soundstage and imaging when listening in the "Stereo" mode compared to upmixing (either Dolby or DTS). So I went and turned off the center channel speaker in the AVR settings, and now I get the same enhanced soundstage with upmixing, too.

What I've found is that all the sound sources around the center in the original recording are hard-anchored to the center speaker, even if originally they are slightly off to the left, or to the right, or slightly above. I then watched a movie in this configuration and found that there's barely any reduction in the experience, even though for movies there's no problem with having the center channel since the soundtrsacks are originally and properly mixed for it. So, my concern is imaging for upmixed music, what do you think about my experience?

I want to also add that on a couple well-made stereo recording that I've found in my collection, I get the soundstage that's not just between the L and R speakers, it wraps 180 degrees around me, perhaps even more - slightly behind me and off to the side. I've got exactly as good imaging on those particular sounds that come from the very side/back in "Stereo" mode as I did in the upmixed "DD" mode, to the point that I have double-checked on two different occasions that this is indeed pure stereo mode and the surround speakers aren't playing. I find this odd; I wonder if something's wrong with my surround speakers setup since they do play in the upmixed mode but do not improve imaging at all.

 

[JoachimStrobel]

Coming a bit late to this thread .... As an earlier post said, either get three identical LCR speakers or go home. I agree to that 100%. But then there is the screen in the way (assuming that a projector is used). I do not believe in acoustical transparent screens.
That means, that the centre needs a frequency correction. This needs to be done together with all the other speakers and for the full frequency range, together with a good room curve - and that process is the challenge of the centre speaker. So either no centre speaker or identical LCR (+X) speakers with full corrections.

 

[Esa Harma]

Hi,

I think you may also use a center channel to overcome the "Fundamental Flaw in Stereo" as described in Toole's book:

9.1.3 An Important One-Toothed Comb—A Fundamental Flaw in Stereo

The loudspeakers with properly controlled directivity may also create a more natural phantom image due to the higher directivity in the frequency range where our hearing is most sensitive. This may also be characterized as "a better stereo image"?

 

[Spocko]

Coming a bit late to this thread .... As an earlier post said, either get three identical LCR speakers or go home. I agree to that 100%. But then there is the screen in the way (assuming that a projector is used). I do not believe in acoustical transparent screens.
That means, that the centre needs a frequency correction. This needs to be done together with all the other speakers and for the full frequency range, together with a good room curve - and that process is the challenge of the centre speaker. So either no centre speaker or identical LCR (+X) speakers with full corrections.

Or get matching LCR speakers with excellent directivity for off axis consistency up to 30 degrees. This allows you to place the center just below the screen while still timbre matched to the L/R. The Genelec 8351B is just such a speaker - flipped horizontally on its side, its near perfect dispersion characteristics up to around 35 degrees will not require frequency adjustments but of course at $4,000 each, this is the cost of perfection for matching LCR without placing speakers behind the screen.

 

[Vasr]

This discussion is conflating two different considerations:

1. Can a center channel speaker be two speakers creating a phantom center? Technically the answer is yes even for a dedicated phantom center channel. I have always done this and it makes it easier to get accurate height cues and "behind the screen effect" right as long as the two speakers are close to the edge of the screen. It is often easier to place such speakers at the same level at the sides than it is to accommodate a single speaker above or below the screen and get the height cues right for the center sound.

The wall mounted Magnepan MMG-Ws work very well for this purpose for wall mounted large screens. Their "toe-in" angle also keep the center channel stage pinned over a wider listening position.

2. Can the L and R double as that phantom center channel speaker? Depends.

The answers can be different to the two considerations above depending on the room and screen size. They converge as the screen gets bigger and the L and R get closer to the edge of the screen. This is why TVs with built-in speakers work fine as center from a staging perspective if not sound quality.

Center channel does not mean that the sound stays right at dead center! It moves from one edge of the screen to another and beyond. This effect is carefully created by the studio mixing.

If the distance between the L and R is relatively large compared to the screen size, then the effective audio sound stage for the phantom center can be wider than the screen which creates some bad artifacts. A person speaking on left side of the screen visually may appear to be off-screen in the audio stage. Or a person stomping off of the stage to the left or the right may leave the screen stage in audio sooner than visually leaving the stage. This is very disconcerting.

While the above problem exists even with a center channel speaker(s), it is much easier to tune the edge of visible screen to audible sound stage with a separate center to reproduce the intent of recording.

The closer the L and R reproducing C get to the edge of the screen, the easier it is to solve the above problem without a C but that can also make the L and R separation too small or too wide depending on the size of the screen. If you have a 5.x or 7.x than moving the L and R closer can create holes in transitions between surround and front speakers.

If you are only watching talk shows or movies from the 50s and 60s, then you probably don't need a C.

 

[Alexium]

A person speaking on left side of the screen visually may appear to be off-screen in the audio stage. Or a person stomping off of the stage to the left or the right may leave the screen stage in audio sooner than visually leaving the stage. This is very disconcerting.

Can you think of a particular scene(s) that would be good for testing for this problem? Would make for a great tool to tune or check your audio setup!

 

[JoachimStrobel]

Or get matching LCR speakers with excellent directivity for off axis consistency up to 30 degrees. This allows you to place the center just below the screen while still timbre matched to the L/R. The Genelec 8351B is just such a speaker - flipped horizontally on its side, its near perfect dispersion characteristics up to around 35 degrees will not require frequency adjustments but of course at $4,000 each, this is the cost of perfection for matching LCR without placing speakers behind the screen.

After throwing enough money at the problem, it might be possible to construct a speaker that sound horizontally the same as its vertical brothers. Most centre channel voices originate from upright standing actors. A speaker being placed on the floor below the screen might be a bit distracting in that case.
Full frequency room correction is the solution. And while it is possible to get that right with REW and RePhase, a canned product like Dirac offers a troublefree solution.

 

[Vasr]

Can you think of a particular scene(s) that would be good for testing for this problem? Would make for a great tool to tune or check your audio setup!

I wonder if someone has actually made a test video for this. Someone should. There are plenty to test individual speakers but not the transitions that I know of.

I don't have a specific scene in mind and can't recall exactly what I may have used in the past.

Best bet would be using an animated movie with surround sound from the recording (not synthesized locally from stereo). These typically have quite exaggerated sounds corresponding to movements, are simpler in isolated sounds (like foot steps) and typically have plenty of such movements. Any of the recent Pixar or Disney animations should work. Just may have to go through the film and identify where the screen exits and screen positions are in relation to the audio stage.

 

[Alexium]

I wonder if someone has actually made a test video for this.

I've found this:

But would prefer an actual multi-channel movie clip.