>Home
>What Is Ambiophonics
>Introduction
>MUST READ!
>Related Articles
>Testimonials
>Bio
>Email
>Photo Page
[ Technical Papers ]
^Updated 11/02/03^
Ambiophonics
2nd Edition
Introduction
Preface
Chapter
1
Chapter
2
Chapter
3
Chapter
4
Chapter
5
Chapter
6
Chapter
7
Chapter 8
Chapter
9
Appendix
A
Appendix B
Figures
>Figure 1
>Figure 2
>Figure 3
>Figure 4
>Figure 5
|
Ambiophonics,
2nd Edition
Replacing Stereophonics to Achieve
Concert-Hall Realism
By Ralph Glasgal
Chapter 6
Ambiophonic Loudspeakers - Ambiopoles, Ambiostats, and Surrstats
In an Ambiophonic system there are two
different functions the loudspeakers must perform. The first one is to generate the
combfiltering and non-phantom-image free front stage and the second one is to reproduce
the surrounding concert hall ambience. I call the pair of speakers (Ambiostats) that
generate the front stage an Ambiopole and the other speakers that provide early
reflections and reverberation tails, Surrstats. While I will describe the ideal
loudspeaker for each Ambiophonic purpose, the ultimate choice for audiophiles will, as
always in stereo or home theater, be determined by their, budget, space, and what they
already own that can be adapted to this purpose.
We wish to apply the rules of good concert-hall design to the choice of
home concert-hall loudspeaker characteristics and speaker placement. Let us assume that we
have available the high quality software-generated hall ambience signals described in
detail in Chapter 8. Let us also assume that our listening room is treated well enough to
eliminate the bulk of the counterproductive listening room reflections using absorption
panels and hopefully an electronic room correction system as described in Chapter 5.
Furthermore, let us also assume that we will be using the software or barrier Ambiopole
arrangement for the left and right front channel speakers which are separated by a ten
degree angle directly in front of the listener or listeners as described in more detail in
Chapter 7.
There is one general characteristic that applies to all the loudspeakers
used in a domestic concert hall: all speakers should be as focused as possible, so as to
reduce the number and level of stray listening room reflections. This is desirable because
no practical room treatment is ever fully absorptive, so the less there is to absorb in
the first place the better the result. In a multi-speaker surround system there is also
the possibility of significant reflections from the surround speaker cabinets, diaphragms
or frames so room treatment is even more desirable in such systems.
The 5.1 Home Theater Conundrum
Since so many of my readers are devotees of video home theater and its
ad-hoc arrangement of two surround speakers placed at the rear sides, I think it would be
best to first discuss the shortcomings of this arrangement before proceeding to describe
something more realistic and scientifically based. The home theater movie people recommend
two dipole speakers placed on edge so that the acoustic null such speakers produce is
facing the listening position. Dipole speakers are speakers that radiate sound equally
from opposite sides. Additionally, these sounds are of opposite polarity and so cancel
where they collide in a room. Some dissenters argue that monopole, that is direct, single
polarity, radiators, are better. Either type of rear surround speakers may be reasonable
some of the time for movie and video sound reproduction. However, where classical music,
jazz, etc. is concerned the Dolby/THX crowd has ignored some serious and seemingly
insoluble acoustic problems common to both of these 5.1 rear surround speaker types.
The use of dipole speakers assumes that the listening room is quite live,
because otherwise, the dipoles would be relatively inaudible. But a live home theater room
means that the direct sound from the front speakers will be reflected, in spades, from all
these nearby surfaces, especially since there are three of them up front emitting direct
sound. To add insult to degradation, their reflections cannot be thoroughly eliminated by
room treatment with absorbers or diffusers, if the dipoles are to function properly. In
movies, these spurious early reflections only slightly impede our ability to locate dialog
and sound effects because of the precedence effect and because the brain has no
preconceived notions of the acoustic spaces the rapidly changing scenes are supposed to be
set in. In contrast, in classical music reproduction, these early direct sound home
theater wall reflections produce cues indicating the hall is small while the recording and
the brain say the hall must be large. The brain usually resolves this contradiction by
deciding that the music is canned. This is one of the many reasons realism, as opposed to
mere localization, in both stereo or multi-channel 5.1 music reproduction, is such an
elusive goal.
But if one forgoes dipole surrounds and uses directional rear speakers in
conjunction with room treatment, the front stereo or 5.1 three-speaker stage improves, but
music reproduction still sounds unrealistic. This effect in 5.1 is due to the fact that
all the rear half-hall ambient sound is coming from two discrete speaker locations where
as in a real hall the sound comes equal in power (but not in detail) from all directions
(diffuse field). Even if the recording is made so perfectly that no direct stage sound is
emanating from these rear speakers there is still no concert hall in the world that
delivers all its early reflections and reverberant tails from two small side spots. The
resultant pinna angle error added to the abnormally low interaural cross correlation
factor, signals the brain that something is rotten in the state of Dolmark and again the
result for 5.1 classical music reproduction is disappointment for any experienced
audiophile or concert goer. Already this point is being conceded by such as Lexicon,
Yamaha, and Holman who now talk about 7.1 or 12.2. Even conceding that monopoles with room
treatment are better for music than dipoles without, the problem remains that making such
multi-channel recordings or trying to extract hall ambience to feed more than two
surrounds (without including erroneous proscenium direct sound or frontal early
reflections) is easier said than done. The answer is full surround convolution as
discussed in chapter 8.
The Front Speakers
In an Ambiophonic system the front speakers should be placed almost
directly in front of the listener with each speaker aimed at the listening area. (See next
chapter). For best results the front main speaker pair, the Ambiopoles should be as
directional as possible. In, theory the ideal speaker for this purpose would behave like a
flashlight, with a sound beam emanating from a single point at ear level and the rest of
the room in deep shadow. The more focused an Ambiopole is, the less sound absorption
treatment the room requires and the more effective the software or mechanical crosstalk
elimination is.
The front speakers used should be capable of reaching concert-hall volume,
the normal speaker selection criteria of good frequency response, low distortion,
reasonable time coherence and affordable price naturally still apply. Since the use of a
Speaker/Room correction DSP can correct most speaker response anomalies, one can choose
the front speakers based primarily on their radiation patterns. An Ambiophonic speaker,
designed by Soundlab, called an Ambiostat, can be used in pairs to form a virtually
perfect Ambiopole. This speaker is a six-foot by three-foot vertical panel that is
slightly curved in the horizontal direction. Such a concave panel focuses sound quite
nicely centered about eight feet from the speaker. The focusing prevents sound from being
sprayed where it is not wanted and makes the job of combing elimination that much easier.
If you are not too critical, this focusing can obviate the need for any other crosstalk
elimination, but then the size of the best listening area becomes smaller.
Front Early-Reflection Loudspeakers
At least one pair of the early-reflection speakers should be placed about
the critical plus or minus 55-degree angle to the listening position. This angle is where
the ear is most sensitive to such spatial cues. Of course if many frontal surround
speakers are available then they should be spaced in whatever way is most convenient or
specified by the hall convolver. The ideal speaker for this purpose is one that radiates
to the listener from as large an area as possible. Large electrostatic or ribbon
loudspeakers are excellent in this application since they can be aimed and so do not cause
unwanted room reflections especially from ceilings and floors. If they are dipoles don't
forget to put sound absorbing material behind them.
A useful property of such large-area full-range sound radiators is that
they provide significant diffusion without invoking spurious room reflections, as would
physical diffusion panels. Ideally, one wants all surround speakers in an Ambiophonic
system to cover as wide a horizontal arc as possible. This corresponds to the situation in
a real concert hall, where the predominant early reflections arrive from slightly
different side directions (but still hopefully centered around 55 degrees) because the
originating sound sources are spread out on the stage and have various angles of incidence
and, therefore, reflection. In the home environment, the computer reconstructed early
reflections are the same for all the right-channel instruments, the same for all the left
channel instruments, and the same for all the center instruments. This moderate lack of
precise spreading of the apparent early reflections would seem to detract from the
concert-hall ideal. But just as the perfect Philharmonic Hall has yet to be built, so our
home room may be real but not 100% ideal.
By mounting speakers on their sides or by leaning tall speakers, at say a
45-degree angle, the ambient signals arrive at the listening position with a greater
diversity of direction and delay. As discussed below, the reverberant field needs to be as
diffuse as possible. Therefore, to the extent that either recorded reverberation or
recreated reverb is present at these front-side loudspeakers, there is an additional
benefit to being wide and as horizontal as possible, providing both vertical and
horizontal dispersion without risking significant room reflection or diffusion of the main
front signals.
Soundlab has also produced a speaker called a Surrstat which is
essentially an Ambiostat turned on its side. Again being slightly concave toward the
listener it delivers ambient sound most efficiently over a wide angle without aggravating
room reflection modes. It rear wave should be absorbed, although the convex side,
radiating over a wider angle, reduces, on average, the intensity of the resulting
reflections impinging on the listening position in most rooms. Eventually, distributed
mode loudspeakers, that are essentially flat panels that radiate equally from their entire
surface, will be quite useful in this application.
Side and Rear Reverberation Loudspeakers
Side and rear speaker pairs are fed with largely uncorrelated
reverberation tails. Since in a concert hall, various reverb tails reach the listener from
virtually all directions, the ideal speaker would be a set of thin squares, which could be
hung on all the walls. I find, however, as above, that large electrostatics or ribbon
speakers do an excellent job, particularly if they can be mounted horizontally. One could
also use multiple small, inexpensive box speakers arranged on pedestals around the rear
half of the room. Again, in theory, each reverberation sound source should have its own
independent reverberation computer but the Japanese have shown that such speaker walls can
easily fool the ear-brain system even when some of the reverberation tail speakers are
correlated. Incidentally, there is no reason why Ambiophonic surround speakers need to be
matched if they can still be reasonably set to the sound level required.
Since the rear reverberant field often has a strong vertical component
coming from the auditorium balconies and ceiling, we have found it advantageous, but by no
means critical, to use one pair of rear speakers elevated as much as possible. These
sometimes provide a richer simulation and a better match to concert-hall design theory but
"better real" is not more real than "real" and this suggestion is,
perhaps, gilding the lily unless the measured hall response specifically includes
elevation data.
<
Chapter 5 | ^ Back to Top ^
| Chapter 7 >
|