Surround Sound - How many speakers needed? - Page 10

There's a lot of material that shows that the number of speakers that can be gainfully used in a surround system is not defined by the number of audio channels in the source program. Is that what you had in mind?

Surely that's [only] because "95% of all movies" are [up-or-down] mixed into 5.0+LFE...?! After all, Star Wars wasn't actually 'recorded live in outer space using a 5.1 microphone array' (nor was Hardware Wars despite the claims in that movie's closing credits!) And I have no doubt that had Fox thought they'd make the most net profits in the long run by mixing stems to [all of] 5.1, 7.1, 9.1, and 11.2 even before Star Wars initial release, I'm sure they would have done so!

And concert, etc., 'live' recordings are made with 'as many microphones as necessary' . . . then mixed to 5.1.

Even so, it's perfectly possible to make [at least] 'live' recordings with a "realistic" microphone configuration, as this 15.2 microphone array from an IBC 2008 demo shows [and from which native discrete 5.1, 7.1, 9.2 'high', 9.2 'wide', and 11.2 'wide+high' downmixes could all be created]:



[Images reproduced from SHV DEMOS AT IBC-2008 — BBC, page 25.]

^I can't think of a worse venue for showing off what a 15.2 (".2" microphones ?) surround sound microphone system can do; outdoors atop a tall building away from traffic, aka a poor man's anechoic chamber. With no walls or ceiling to reflect off from a reporter facing the camera, for instance, would be picked up by the front mics only and all those other channels would get nada. OK, sure, maybe once in a while that rear left mic # 13 might pick up a tug boat toot in the distance, or something, but otherwise using such a complex array in a completely non-reverberant environment seems a waste and rather silly.

[Using the great outdoors in lieu of an actual anechoic chamber (which can be big bucks) has been a trick of speaker designers/testers for years. BTW]

[Image and text reproduced from SHV DEMOS AT IBC-2008 — BBC, pages 29/30.]

Looks like a good location for the microphone array to me...?!

I'd let the room dictate. Currently using 5.1 with two subs running off the same .1 sub channel. I cant do rear surrounds in the current room, but its really wide so I think I could benifit from some front wides.
When my Theater room is ready, I'll be able to do 9 channel with rear surrounds and front heights.

Personally, I think more benifit could be made of .2 or .4 channels. I find two midsize subs does much better than one large one, it would be even better if the audio was actually seperate instead of just duplicating the .1 channel. Or .4 could make one sub in each corner... sounds like a bass fiends dream.

If I could afford it and had a room big enough... I suppose I'd have as many speakers as I could... But they'd have to be quality. Theres no sense in having 20 crap speakers if 7 good speakers sound better.

In a normal size room, 5.1 is all that's needed. In a very large loom, 7.1 could be an option. 9.1 or 11.1 is just like valentines day, a profit based proposition.

I agree that 5.1 is best most of the time. Bigger rooms may require more but My personal preference is 5.3 in a 14 ft room. It's not bass if you can't feel it in your chest!

Hi Will,

Thanks much for the considerable reply. Sorry to get you off on a tear. I think we'll need a beer summit to settle this properly.

My original intention was to steer clear of the merit issues altogether, which speakers give best effects, etc. I'm only focusing on the statement that DSX adds no new signals.

It may help if we see the overall text Chris posted:


a) He clearly states that DSX adds reflections.

b) In my read of the last sentence, the use of extracts' does not mean removes from the source, but derives from the source. In other words, DSX applies appropriate frequency response and perceptual processing to the audio in real time to create the new side wall reflections that are being added to the presentation.

Even if one had a recording from a concert hall that included side-wall reflections, I'd be very surprised if any algorithm could even find them, let alone redirect them.

Based on the above, I'm sticking with my original conclusion, that DSX adds new signals to the presentation. I am not making any value judgment, just asserting the point.

Hi, Roger, thanks for the reasoned reply. I'm still not reading it "your" way--esp. as it contradicts other documentation I've seen. But, in any event, it is very ambiguous. IMO, The question that Chris asks is from the perspective, as I read it, of a consumer who's heard mysterious--erroneous--rumors about added reflection info.--not, as I read it, a positive or affirmative statement describing what the process actually does in this regard. But, in any event, the fact remains that the process--whatever it is--is benign. As was your original post--and, no, you didn't wind me up; just reacting to a lot of what I regard as misinformation around here. Perhaps reactionary naysayers will at least give the processing some study and audition time before slamming it blindly (that's all it will take). Then let them slam it, if they must, or can, for real and informed reasons.

Also, I kinda got stranded at my desk in a "tween" mode--I didn't feel well enough (something I ate the night before) to do the physical things I'd scheduled, nor the intellectual work for the coming week, so I pretended that sitting in my desk chair and pontificating on AVS for a couple of hours was constructive, lol. Your post was just the harmless catalyst. (Besides, I knew from past experiences "trading fours" with you, that I was unlikely to get slammed in return! )

Hey, you do know that' I'll do my best to take you up on that beer summit offer, right?!

I first saw the pics of your totally beautiful "Deadwood" theater yesterday--sorry I didn't get a chance to reply and mention this yesterday, before we all climb on the workweek merry-go-round. What an obvious labor of love. I can tell how much thought and crafting went into even the tiniest details (your responses to admirers' postings and pics of the "behind the scenes" crafting were quite intriguing). BTW: Most folks, as you know, I'm sure, hide their diffusors behind fabric. What a cool idea to expose them, and make them part of the overall aesthetic--making these a design asset. Form and function coalesce. The space is both industrial and elegant--"clean" and purposeful, but "rich"--no mean feat; it's terrific looking! With, IMHO, a little bit of a "Danish modern" vibe (in the best sense). Very much complimentary to--and, importantly: stimulative of--my own design sensibilities. (And, not least, the perfect setting for a beer summit!)



Will

P.S. Just re-read your post, and thought it easier to respond to another comment here. (Besides, I like postscripts...can ya tell?) Regarding algorithms' ease of "finding" early reflections and then steering them: this is simple, actually...been done for many years--they're just out-of-phase repetitions of existing signal information that decay in amplitude in predictible ways--and therefore are easily recognized; actually, what would be veritably tough to do--nigh on impossbile at this stage--is figuring out what to add reflections to, on the fly--and, tellingly, when not to do this. And be accurate to the changing soundspaces every time. Extraction/redirection of actual signals has got to be easier (not to mention far better sounding!) than attempting to predict what needs to have simulated reflections-processing added to it, in some ultimately non-natural fake/phasey way, and when. (Dolby Digital has been doing the former for almost two decades, now). Because, clearly, the Wides do not receive reflections-signals constantly (as with ill-regarded DSP ambiance add-ons), but only intermittently, when the soundspace requires. Getting this right, on the fly, with hi-resolution multi-channel soundtracks, given today's processing SOTA, seems impossible. (Note that the original HDMI spec couldn't even permit mere transmission of these high-resolution audio channels!) Even with "sky's the limit" engineering statement pieces/proof of concept projects with unlimited funding. (IMHO, natch!) What Audyssey does with DSX is, in fact, not so revolutionary in terms of DSP execution, at all; it's merely slightly evolutionary in terms of the tools it uses, which have been used by acive steering circuits for the two surround L/R channels in the 5.1 array for decades, now. (Well, call it "1.6 decades," I guess.)

What is significant and actually revolutionary about the DSX process is its psychoacoustic modeling--that it has re-ordered the hierarchy, long dominated by Dolby, in terms of which additional surround channels have the most relative, ranked importance--and where their transducers should be placed in our home theaters--given how our ears/brains function, in terms of recreating a natural, 3D soundspace.

First beer's on you. Make it a stout or double-bock, please!

Will / Roger,

This discussion is very interesting. Considering that this thread will die soon (when the survey is finished), why don't you port the discussion to the Audyssey thread? I'm pretty sure Chris K. (and others) will react.

Mark

Mark,

Nay! This thread will NEVER die!

Why? Because Roger and I have a beer summit at stake, and must make a full public record of our respective diplomatic positions...in order to fill the 'fridge...see, Roger neeeds to reply back (second beer's on me), then I need to reply (third beer's on him), &c., &c. We need to generate at least three cases of fine imported beer this way, I'm thinkin'...and maybe even a bottle of single-malt Scotch, or Remy Martin, just in case. Plus the celebratory champagne for the successful conclusion. (It'll be a several-day summit, you see. Worth doin', worth doin' well, and we have major diplomatic initiatives at stake. This friendly little debate is relatively unimportant, it's just the preliminary sideshow, the means to an end, the opening salvos of a glorious mutual victory, the art-of-the-possible writ large...).

Seriously, that's a good idea, Mark.



Will

P.S. You wouldn't happen to be an accredited member of the press, wouldja, Mark, with a desire to cover this history-making summit? If not, no worries, the rules are lax (we make 'em up, after all--apparently as we go along), and I can whip off some Photoshop credentials in a flash...or, hey, why be formal about it. Just scribble yourself something appropriate but humorous on a bar napkin. I'm sure it'll pass muster (we've got George Clooney as Head of Security, so, no worries). And, you bring the makin's for Eggs Benedict and Bloody Marys (cuz I can see right now that those morning sessions are gonna be rough).

P.P.S. Oh. And, we're obviously gonna need some showgirls. Know anyone with reasonable experience as "Director of Transportation and Entertainment Facilitation?" No? Well, know any serious audio- and/or videophiles with drivers' licenses and Big Black Books? Hey. We're nothin' if not flexible.

Benign or not, delightful or not, stunning or not, is not at issue. Only whether DSX adds synthesized reflections or not. All evidence I've seen, scant though it may be thus far, tells me it does.


Thanks much for the effusive words. Looks like I'm buying.


I think we are using the term 'steering' the same way we do for the various matrix-based surround decoders like PLII or Logic7. If so, the sounds that are directed to a new speaker are removed from the original--like dialog in a 2-ch program played in 5.1. These signals can only be effectively steered if they are coherent. Side wall reflections are by their nature diffuse. I once asked the the guys from Lake if they could remove the room reflections from a recording, and they said it would require de-convolving with the room's actual impulse response. Since I did not hear key words like "sure" or "yup" in there anywhere, I took that as a no.


It is actually not difficult to analyze a soundtrack to separate the correlated from the uncorrelated components. In fact this is what matrix decoders have been doing, rudimentarily though it may be, for decades (since 1983 in the case of Dolby, but others were even earlier). Nor is it difficult to look at a 5.1 soundtrack and concentrate the DSX processing on the L/R inputs, thereby serving up the channels most appropriate for processing on a platter. What with some powerful DSPs and time domain processing, it's more possible than ever to do this well.


Not sure what you mean about Dolby Digital.


Methinks you underestimate the skills and craft of Audyssey.


I'd be a little insulted, if I were Chris.


It's not a new way to the same end, it's a new way to a different end. I see it as a well reasoned opinion or philosophy. No more or less valid than others. As with any set of choices, each person must decide which compromises one wants to adopt or to avoid.

And by the way, Dolby did not define the multi-speaker playing field, it merely facilitated it, and advocated it, across the ecosystem.


We have good beer in Oregon. Seriously. I'll keep the light on.

Hi,

Sorry, Roger, lemme correct that to DS/DPL, not DD. That's what I originally had written, but I changed it, for some reason. And added the reference to 1.6 decades. Dunno why. It was early this morning, go figure. So, we agree to a larger degree, as this changes meaning(s) a lot, from your perspective.

And, yeah, I was listening to the David Hafler "simulated time delay" design (and other matrixed soultions, usually based upon this) in the very early '80s, as a kid.

O.K. Next beer's on me. We're gettin' there!

Will


P.S. A long drive from the mid-Atlantic region...but, you never know...it'd be a glorious one. And I've never driven coast-to-coast. Wanted to! And, I'll do almost anything for a great beer...

If receiver manufacturers want to add on a model or two at the top end with more channels, that is up to them and the market. But I would hate to see the $1500 to $2500 range move in this direction. The costs for the additional channels would have to come from somewhere. So either the costs would increase or quality would go down in other areas. I would still like to see very solid, audiophile quality receivers in this range with 5.1 or at most 7.1 support, with all the dollars going to those channels.

Robb

It seems likely that around 80%+ of AVR owners are running 2.x to 5.x configurations, with another 10% (or so) playing 6.x/7.x . . . leaving a small number (at most perhaps 3 or 4 percent) who might run some system with 8 or more 'amped channels', either now or in the near future...

CEMs can market one 7.1 AVR for all 2.x to 7.x configurations (at some price point) because it avoids their needing separate 2.1, 5.1, and 7.1 SKUs. Whether the CEMs "drift up" to an 8.x/9.x 'amps on board' AVR at certain price points, or stick to 7.1 AVRs plus RCA connects for 'optional' eighth or ninth channels will be determined by the economics of the choices...

We need more video channels too. We've only got 1 or 2 with mono/stereoscopic TV.

Also about audio, wouldn't ambisonic sound be better?
http://en.wikipedia.org/wiki/Ambisonics

The key word in the poll question is "needed". I voted for 5.1 which is what is needed (due mostly to content) but I have a 7.1 setup using FH speakers.

If true discrete content begins to appear (more than the few true "surround back" mixes) then the numbers will grow.

The current matrix ed approaches (DSX and PLIIz) do a good job on most content rendering decent FH effects (in my case) without taking over the sound field. I like what DSX has done for almost every thing we have played.

Pure-ists will want true discrete channels mixed and rendered for what the artist/director intended but until we see "11.x" theaters doubtful there will be any content for quite a while (especially with the focus on 3D....).

which is unlikely due to the cost of sound mixing.

+1

Enough with the brain dead gimmicks, let's get back to building quality products!!!

Of course I can remember seeing the same comments in (what passed at the time for) AV magazines . . . when the [so called] "brain dead gimmicks" were the first (2/1.0) 'digital' Dolby Surround decoders, again later the first digital preamp/processors, etc. . . . and the "quality product builds to which the world needed to return" were 100% analog 2 channel preamps...?!

Quite honestly I don't remember there is such thing as digital Dolby Surround decoders. Dolby Surround is an analog system. Dolby Digital, on the other hand, was called AC-3 and I didn't remember anybody calling it as "gimmicks".

Gimmicks, to me, are additional things companies trying to sell you to produce or "intelligently" reproduce sound channel(s) that is(are) not there to begin with... such as the 7.1, or Soundfield Processing, or DPL IIz

PS: Dolby EX is different because some movies, such as Episode I, was mixed for Dolby EX

My original Dolby Surround decoder is a DSP-100U, for which the block diagram shows the L/R input channels A-to-D'd (44.1K, 16bit) for processing that includes a pair(?) of [specifically identified] YM3413 chips which appear to control the differencing and (15, 20, 30 ms) delay processing. The surround channels are then D-to-A'd . . . and the required Dolby Noise Reduction appears to be applied in the analog domain...?!

that's the only anomaly because of Yamaha's Digital Soundfield Processor. The "gimmick" they're referring to was the Digital Soundfield part. This is because any recording will already have its own reverb/echo built into the mix. Using the DSP, it adds not only echo/room reflections and such to the song but to the reverb/echo that's built into the mix.

Especially in the recording world, adding reverb to another reverb is a no-no.

It's kinda work for the 70mm Cinema Mode (for movies) and Cathedral is my favourite during the early days of audio CD when the entire orchestra or choir usually recorded live in an almost dead room without any additional reverb in the mix.

Nowadays with all the effects put into the mix, the DSP that still works are (based on psychoacoustic logic):
70mm Cinema (for movies)
Adventure (for movies)
Rock Concert (for any live concert in a stadium-setting)
Dance Club (can't remember the name exactly) for any dance music.

But then again, after a while, it gets tiring to my ears. DSP was great for 2-channel application way back when but not so much for multichannel soundtracks with exception of their presence channel (not to be confused with DPL IIz height channels which I absolutely hate)

Nice post. I am appreciating your exchange with Roger and do hope you bring it over to the Audyssey thread.

I am quite sure you were being facetious, but allow me one playfully pedantic and OT comment on your discussion of evolutionary psychoacoustics, for the benefit of those whose knowledge of evolutionary biology may have become confounded by the likes of "The Flintstones" and "1 Million Years BC". Humans were seperated by millions of years of evolution from the dinosaurs and thus, thankfully, never coexisted.

Roger/Will -- the oracle has spoken

at the moment,there is 5.1,7.1 and 9.1 of the home theater on the market.for traditional 5.1 home theater,there is only one best position for listeners.in order to increase the best position for listeners,the factory research the 7.1 and 9.1.so the difference between 5.1 and 7.1,9.1 is just different on the number of the best position.

I think some of the members know the new speaker technology-flat panel technology.for this technology,the sound quality of the flat panel speaker is surrounding the space.so at any spot of your room,you will listen same good sound.so for flat panel home theater,5.1 is enough.

I voted 7.1 for lack of a 6.1 option

Terrific! I hadn't had a chance to edit and plop these convoluted posts over to the Audyssey thread (and now don't have to).

Also terrific, because: now we know, definitively.

Roger, I am properly chastened.

Beer's on me (but you have to be my guide to Oregon's finest, should I make it out there, please).

Mark, thanks, my friend!

And, I should say, that I enjoyed this collegial process--it's among the many attributes that make the AVS Forums unique among forums on any subject, and such a great "club" (Groucho Marks' quip notwithstanding) to be a member of. Thanks for sharing this fun process with me, guys.



Will

P.S. (Ya knew it was coming...the infamous willdao/Mr.Hyde postscript!

I'm actually rather tired of all this civilized, mamby-pamby debate, I must confess; and hugely bitter about my ignominous defeat--although it's well hidden behind my notorious demeanor of placidity and reasoned discourse. Inside, I'm secretly really a twisted troll. So. Now that we're (ahem!) "experts," guys, there's fresh game afoot. Let's all motor our tanks and flame-throwers on over to the Audyssey thread and torch and scorch some poor, confused-neophyte's butt!

(Ouch! Uncle! Kidding! Kidding!)

7.1 system,The surround is very convincing in my 4*6 square meters