Do More Speakers/channels enhance envelopment, or is it just more?

[Matthew J Poes]

I remember when there was a move from 5.1 to 7.1 and found, like many, that in a typical room the extra channels didn't seem to add anything. Later I found scenarios where the extra speakers could enhance envelopment in a meaningful way, but mostly when speaker placement was less than idea for the surrounds, especially long narrow rooms. In general however, I've suggested to people to stick with 5.1 or 7.1 (or 5.1.2 through 7.2.4) and really stayed away from anything more than that.

I'm curious, how many of you have played around a lot with 9+ channels, such as 9.2.4 type setups. If so, what do you think it added? Are there room times you think it adds more to?

One theory I have had is that it may be that in smaller and narrower rooms, the additional speakers might actually enhance the envelopment since wall reflections that might negatively impact speaker imaging and envelopment are mitigated by more ideally placed (de-correlated) direct signals.

Thoughts

 

[Sonnie Parker]

I am still stuck on 5 channel surround for the most part. Never been able to get the rears to help. However, Atmos has been interesting enough that I've kept it and now run 4 height channels. It does seem to be somewhat more enveloping with the 5.2.4 system. Now that I have the heights, I can't make sense of rears for sure now... I just don't see how it can add anything, although I can't say that with certainty since I haven't tried it with the height speakers in play. I may try a rear center speaker one day just to see what happens.

 

[dc2bluelight]

Since the days of mono, it's been known that to produce a perceived change that the majority of listeners would recognize takes an approximate doubling of speakers. Mono to stereo, stereo to Quad/5.1, and so forth, with the practical maximum for the typical home theater landing around 10-12, relating to room size and seat count. However, the specific channel/speaker plan also makes a significant difference. While research had clearly shown that localization is best in the frontal plane, and somewhat extended to the entire frontal hemisphere, indicating that more speakers would be best deployed in that hemisphere where the would do the most good, the brilliant marketing wizards have chosen to add speakers where localization is worst...behind and above. That's a rather sad error! Those developing multichannel film sound recognized very early (as early as 1938) that phantom imaging doesn't work in multi-seat rooms for anyone but those seated directly on the centerline, and those seats at home are few to none. Hence, the most used and important speaker in a theater system, home or commercial, the Center. The boffins at Dolby knew this of course, but instead they completely ignored that problem in the rear when we moved past the short-lived 6.1 plan into the ill-supported 7.1 layout (for the home, not the theater which was different).

Even worse, Atmos (especially the home version with a limited number of height speakers) depends on phantom imaging between a speaker pair for the placement of objects directly overhead, again ignoring the fact that phantom center imaging doesn't work. It's strange! So 5.1.2 or 7.2.4 tends to mis-localize height objects depending on where one is seated, and places even more speakers where they are not as easy to localize.

What I'd actually like to have is more along the lines of the 10.2/12.2 plan suggested in 1999 by Tom Holman and Chris Kyriakakis: LCR, plus Lwide and Rwide (expanding and localizing the front soundstage), then L High, R High complete the front, thus placing channels where they can actually be localized well. Surround in that plan is Ls, Rs, and Surround Back, a single rear surround speaker, taking care of the center-rear phantom image issue, or in the 12.2 plan adds 2 more surrounds and differentiates between diffuse surround and direct surround. I've heard that system several times, and it works WAY better than any home Atmos installation in rendering an impression of an acoustic space. It is challenged if there is a desire to mix a track where an object takes a zig-zag overhead path from the back to the front, of course (that could have been fixed). Unfortunately, 10.2/12.2 was never adopted, so now we're stuck with what we have.

Then we have the practical matter. The average 5.1 system I see in friends homes has at least 1, typically several, misplaced speakers that are where they are for practical or aesthetic reasons. And very few have gone to 7.x. I haven't checked market penetration for Atmos home installations, but I strongly doubt it's more than a percent or so. It's just difficult to do, and expensive not only to get the gear but to wire it. Then there's the little problem that not all films provide a lot of height stimulus. So people do what the sales guys do with in-store demos--crank up the heights so they can be HEARD!

As a result, I typically strongly suggest 5.1, with 2 or more subs (I'm sorry, I still can't get myself to call that 5.2!) because it's doable, affordable, and can be done well. A well done 5.1 system actually will be more pleasing with a far higher WAF than slap-dash systems with higher channel count, once you surmount the more-is-better bias.

At the high end we have Trinnov who has technology that mitigates the placement problem for heights and surrounds by applying sophisticated psychoacoustic processing to virtually position surround and height speakers that can't be put where they should optimally be. But that is a somewhat rarified world, as the processor along costs like a reasonable car, or a nice car, depending on how you configure it.

So I'm at "good 5.1", with dipole/tripole surrounds if possible as a first recommendation. And it gets very granular from there up.

 

[ddude003]

Although I have a pretty standard 7.1 speaker layout, I have found it lacking in something... At one time I had tried putting the rear surrounds up high on the front wall and liked that a little better with some of the height processing the Pre/Pro can do... But to be honest, I think 5.1 actually sounds better... I also notice that with the ESLs things can really sound magical or pretty bad depending on the source material... I do not think they are very forgiving as some speakers can be... Sometimes 3.1 or even 2.1 may sound best to my ears...

I have heard a few Atmos systems in some custom home theatre showrooms and at a few friends home... I was not very impressed with anything so far... And I am not so sure those systems were set up properly...

 

[Mark C Flick]

My short answer, it's just more. I mean I can fully enjoy a film in stereo as long as they are good speakers and there is a sub involved.

 

[AudiocRaver]

It is just more. When I have tried it, the extra channels sound better at first, but over time the inevitable damage to the imaging becomes apparent and I have to shut it off.

 

[JStewart]

Even worse, Atmos (especially the home version with a limited number of height speakers) depends on phantom imaging between a speaker pair for the placement of objects directly overhead, again ignoring the fact that phantom center imaging doesn't work. It's strange! So 5.1.2 or 7.2.4 tends to mis-localize height objects depending on where one is seated, and places even more speakers where they are not as easy to localize.

Great post. Didn't want to re-quote the whole thing here but much of what you suggested for placement is better represented by an Auro3D speaker layout vs Atmos. Any thoughts on this?

Has anyone here heard Auro3D? What did you think?

 

[dc2bluelight]

Great post. Didn't want to re-quote the whole thing here but much of what you suggested for placement is better represented by an Auro3D speaker layout vs Atmos. Any thoughts on this?

There are a few things to like in Auro, but also a few issues. The codec, if you can call it that, is quite simple, just PCM. It's not an object-oriented system, so each speaker is driven by a full-time, full-bandwidth channel. And they did add the VOG speaker, which I think is important, but not a deal-breaker. They add Center High in a few of the higher speaker count plans, but they ignore L and R wide, something pretty impactful, and a regrettable oversight. All Auro really does is add a height "layer". I think it has a bit more potential for replicating natural ambient spaces than Atmos, but less capable of precision object location.

But we get into a real log-jam when we try to implement several immersive systems in one speaker plan. Even something like the Marantz 8805 cannot provide enough connections and signals to handle optimum Atmos and Auro without leaving something out. One configuration leaves the Front Wides off of Atmos, and goes heavily into Top arrays, another includes Front Wides but ditches center top and high. It's a study in compromises, and DTS-X does their compromises in virtualization. When the compromises become insurmountable, then things like the Trinnove Altitude32 start to make sense. More output channels, better speaker locations, fewer compromises. Hey, it's only money.

The real problem is Auro hasn't penetrated the market nearly as well as Atmos, and neither has DTS-X, and that generally spells eventual challenges. And while Atmos mixes could theoretically be easily pre-rendered for Auro, and vice versa, it's hardly the optimal way to mix. We cannot expect many productions to pony up to do all those addition mix variants either. Think of it: 5.1 theatrical, 5.1 home remix, 7.1 home remix, Atmos, Auro, DTS- X? Nope, not happening. It's likely stopping at Atmos and skipping 7.1, if that. Yeah, there are a bunch of Auro releases, but the general marketing principle is the guy first into the category owns it, at least in name and concept. They're the Kleenex, and thus have a huge lead. Everyone else is playing catch-up, especially DTS. Sometimes two competing products can co-exist for a while, but eventually there's usually a winner. It doesn't matter ultimately that Auro may be better with spacial replication and Atmos may be better with precise object location.

And we really are talking about a tiny market fraction here. Dedicated home theater rooms make up less than 1% of installed systems, and ones in existence may not be so easily retrofitted with top or height speakers. I'm not sure any of this matters all that much.

 

[JStewart]

Some excellent points dc2bluelight. Thanks much.

 

[AudiocRaver]

When I was at Sonnie's a year ago, we watched several ATMOS releases on his system. Not much ever came from the height speakers, except helicopters. EVERY movie had a helicopter in it somewhere, fluttering overhead. Maybe the biggest change that will occur in the movie business from ATMOS acceptance will be more helicopter rentals to impress ATMOS viewers - and help them justify the extra hardware purchase.

Mixed feelings about more speakers across the front vs the back. A properly set up stereo pair of Fronts can produce stunning imaging for the center seat. If there is no center seat, shame on the designer, she/he does not deserve great localization (kidding, but only partly). Also, the way a typical home theater is used, the owner is the only one who cares &/or has the listening skills to appreciate that imaging anyway.

However, when talking about more ideal circumstances and the theoretical limitations, it makes sense that the existence of more front speaker positions would yield better localization for more seats. Why wouldn't the same apply to the rears, perhaps to be appreciated and enjoyed less and to have less practical use in a cinema mix, but in theoretical terms, that would seem the best way to improve rear image precision, or image precision in any direction from a listener position. Where the human capability is weaker, more assistance is called for. I admit that my perspective is more front-center-seat-centric, because that one seat is the only one that I personally care about (two channel guy).

 

[dc2bluelight]

.

However, when talking about more ideal circumstances and the theoretical limitations, it makes sense that the existence of more front speaker positions would yield better localization for more seats. Why wouldn't the same apply to the rears, perhaps to be appreciated and enjoyed less and to have less practical use in a cinema mix, but in theoretical terms, that would seem the best way to improve rear image precision, or image precision in any direction from a listener position. Where the human capability is weaker, more assistance is called for.

I can see why you'd think that, but more speakers is not assistance in localization because we don't have that ability. You can't localize well behind you, it's very easily fooled. So adding even more vectors of what is already vague doesn't really help in the say way that adding front vectors does.

I admit that my perspective is more front-center-seat-centric, because that one seat is the only one that I personally care about (two channel guy).

It's not just you, or two-channel guys. It's humans. We are designed with forward focussed senses, visual, audible, even tactile. It figures, doesn't it? Life is in front of us. We look forward, move forward, deal with things in front of us. We have enough sensory input in other directions that we recognize stimulus and turn to face it, thus moving our highest sensory acuity to point to the source. It follows, then, that directional hearing is optimized for frontal stimulus, with vision augmenting the experience (or sound augmenting vision if you prefer).

I had a directional mess-up today. I stood at the food store checkout, working through the maddening credit card touch pad, when the unit directly behind me at the next aisle sounded it's transaction-completed honk. Because it was somewhere behind me, and I was facing my own touch pad expecting that sound, I pulled my card...early...and had to start over. I could swear it was my touch pad! But wrong! As I started again, I heard the unit behind me honk again and observed how easily that vector could be completely miss-localized. That was real-life stimulus, and it failed completely. No reason to expect virtualized vectors from a surround speaker array to work any better than life.

There was a move to a single rear surround (6.1) just before we got 7.1, and many AVRs can still be set up that way. It's an improvement over just a pair of dipole diffuse-field surrounds, and was actually requested by mixers for theatrical systems first. And remember the old THX 7.1 configuration of the mid 2000s? They put the rear surrounds side by side, close together, and "virtualized" sounds outside of them, but it also gave them a hard-center rear. Even THX gave that one up, and don't get me started on THX!

But a full array, equivalent to the front high, wide, and center high would be largely wasted, except a few top-speaker applications, like helicopters, jungle canopy, etc.