How Many Subwoofers Is
Enough?
Floyd Toole and Todd Welti may
surprise you with their answer.
In previous columns on subwoofer placement, we’ve
discussed the conventional wisdom of adding the ".1" to 5.1-, 6.1-, and
7.1-channel surround sound. That wisdom pretty much comes down to "keep trying until
it sounds right," which is a variation on Duke Ellington’s advice: "If it
sounds good, it is good."
You can get good results that way -- after all, most of us
end up pretty happy with the sounds of our systems -- but it ain’t easy. And, in a
lot of rooms, it ain’t even all that good -- not really. Not measurably.
The problem isn’t the subwoofer. The problem is the
room and where you -- and the guy next to you, and the subwoofer -- are in the
room. No wonder we A/V junkies are all a bit crazy. If we didn’t start out that way,
trying to balance all those variables would drive us nuts.
So it came as a relief to discover that even the most
knowledgeable experts were as frustrated as I am by this situation. I’m referring to
Dr. Floyd E. Toole, Harman International’s vice president of acoustical engineering
and one of the most respected people currently working in the audio industry. When I heard
that Dr. Toole had some interesting ideas on the subject of subwoofers, I contacted him.
It turns out that Toole and Harman are about to set the
industry on its, er, ear. Before we spoke, however, Toole suggested I read a white paper
published by his colleague Todd Welti, "Subwoofers:
Optimum Number and Locations," an extremely detailed analysis of Welti’s
search for some answers to the hit-or-miss nature of subwoofer placement.
It’s worth reading, if only for fun: Welti has
considered the placement of 5000 subwoofers in a room. He asked, "Can a sufficiently
large number of subwoofers cancel out all room modes?" His conclusion:
"Theoretically, yes. Practically, NO!"
That’s when I knew this interview was going to be a
hoot.
Wes Phillips: You guys seem to have had quite a
bit of fun with this white paper.
Floyd Toole: And that’s not the end of it,
either. We’ve done more work since that one -- it was just the opening shot, I guess.
We started this line of inquiry because trying to provide decent bass for more than one
listener with a single subwoofer was seemingly a hopeless cause -- well, it is a
hopeless cause. You can’t do it.
However, if there is only one listener -- or only
one listener who matters -- you can provide good results with a single subwoofer, as long
as you realize that, in the end, you’re going to have to equalize everything. And
we’re not talking about the one-third-octave graphic-equalizer BS that the industry
has purveyed for decades.
That said, no matter how many subwoofers and how many
listeners we’re talking about, equalization should be the final step to make it sound
right. A single subwoofer can entertain a single listener with equalization -- good
sound is possible. But once you have more than one listener, then you need multiple
subwoofers.
At that point, the problem can be divided into two
separable categories: one where the room is a regular rectangular shape, and the other
where the room is not perfectly rectangular or symmetrical, which is the sort of room most
of us live in.
The problem Todd solved in the white paper we published
concerned perfectly rectangular rooms with fairly "normal" middle-of-the-room
seating. For that, there are some fairly standardized arrangements of subwoofers that work
-- or at least reduce the variations from seat-to-seat performance. Of course, once you
reduce the seat-to-seat variations, any equalization you apply will apply, more or less,
to all of the seats.
WP: Todd, is it possible to synopsize the
conclusions of your white paper?
Todd Welti: It was surprising to me that I was able
to come to as distinct a conclusion as I did. When I first started working on it, I
assumed that I would come up with a solution for a theoretical room but it wouldn’t
work for real rooms. But I was pleasantly surprised to find that it worked in real rooms
as well. From reports I’ve been getting from the field, it not only works in our
room, but most other rooms as well.
The conclusion I came to was that two subwoofers give you
about 90% of the performance that is possible, and that four take you about as far as you
can reasonably expect to go. Anything more than four is not going to get you much in the
general sense -- and these are general conclusions.
WP: So the 5000-subwoofer solution just
didn’t work out.
FT: Acoustic wallpaper, as I call it.
TW: In terms of recognition, it sure got a lot of
people’s attention, but no, four subwoofers seems to do the trick -- and we came up
with three configurations out of about 100 that we looked at that are the cream of the
crop. Actually, there’s a fourth solution -- one that Floyd came up with -- that
worked even better than the others, but it involves bringing the subwoofers out from the
walls, which I generally didn’t consider.
FT: It’s not very practical, unless you could
put them in the ceiling. It uses four subwoofers located at the 25% points from all four
walls.
TW: In other words, you shrink the whole room by 25%
and put the subwoofers at the corners of that virtual room. You get incredible
performance, but that’s just not practical in most rooms. But if you use two or four
subwoofers in the corners or the wall midpoints, you can get pretty good performance --
and that’s about as succinct as I can get.
WP: That’s pretty succinct. But you’ve
continued to do research, and now you’re trying to crack the conundrum of the
irregularly shaped room. How’s that going?
FT: Oh, it’s done -- but it hasn’t
resulted in a product yet, which is the frustrating part. Harman is working on fine-tuning
the algorithm and the user interface. It requires some serious computer calculations.
The way I like to describe what this makes possible is that
the customer or installer can basically just walk around a room and identify locations
where a subwoofer could be positioned, bearing in mind all of the practical constraints of
doors, windows, and visual aesthetics. Once you’ve identified some number of possible
locations, you then identify where the listeners’ heads are going to be, and you take
one subwoofer and you go in turn to each of these locations and you take measurements,
establishing a technical database of transfer functions that will be stored in the setup
computer you are using.
Once the computer has all that data, you sit down and ask
the optimizing program to find for you the best locations for some number of subwoofers.
In doing so, it will add the time delay, level adjustment, and one band of parametric
equalization. Of course, that means you need some additional electronics in the signal
paths of each of the subwoofers that you end up with.
Now the question is, how many subwoofers and which
locations will prove to be optimal? You can ask the optimization program to find the best
solution for these listening locations in this room using, say, two subwoofers -- or three
or four or five -- depending on how many you feel you can afford. We find that plateaus
pretty quickly at three or four as the best you’re likely to get. The program will
calculate what processing you need for each of the channels for each of those locations.
Then you apply global equalization to make it sound good for all of the listeners.
WP: Is this a technology we’re likely to see
in products that will be available any time soon?
FT: Yes. It’s not something you’re likely
to see in mass-market products any time in the foreseeable future, but it is very
seriously being worked on in our higher-end offerings from Lexicon, Revel, Mark Levinson,
and JBL Synthesis, of course. They’ll have access to it first -- and it still
requires human intervention.
Getting good bass in a room has always been pretty much a
trial-and-error business. Some systems work pretty well and some don’t. But now, for
the first time, we actually have a handle on the problem -- we’re stunned to discover
how well it works. I have this system at home, and without it I was getting 28dB
variations among my five principal seats at low frequencies. Now, with it, that’s
down to about ±3dB. It’s a staggering, amazing difference! You can literally walk
around the room listening to your favorite bass riff and it just basically doesn’t
change. The room just disappears -- or at least what we recognize as the negative
attributes of the room disappear.
It’s going to change things around the world. American
rooms are bigger and more forgiving of deep bass than the rooms in a lot of other
countries -- and people outside of America tend to shy away from systems that deliver
intense bass signals because of all the room-interaction problems they cause. They just
don’t sound good in hard, small rooms. But if we can solve those interaction
problems, it’s going to change everything for an awful lot of listeners.
WP: That’s a prospect that’s going to
make a lot of music -- and movie -- lovers very happy.
...Wes Phillips
wes@onhometheater.com
|
