Loudspeaker Power Response and the Reverberant Sound Field (RSF)
I have decided to address this in a little detail
here. This is a topic that has been a distillation in the back of my mind
for many years. I am more concerned at looking at it from a real world
perspective rather than an academic one. In my case, "real world" is from
the speaker designer's view. I have to make speakers work in rooms that vary
but I have to expect that rooms are reasonably well behaved. You
cannot please all rooms (they vary almost as much as people) and you can
only make a speaker that will work well in a reasonable room as no
loudspeaker will sound good in a bad room.
reverberant field response in a reverberant room is dependent on the
sources power response – not its anechoic response."
The designer of the speaker must
get this aspect of his design right. You start with the wrong speaker and
RSF for your room cannot be optimised properly. The power response is not a
single discrete measurement made in an anechoic or pseudo-anechoic (time
edited to omit reflections) measurements. It takes several of these
measurements from carefully selected microphone locations, then averaged out
with correct weighing, to give us an indication of the power
Unless you are a speaker designer
there is little than you can do other than choosing the right speaker for
your room. My job, as I see it, is to design a speaker with a correct power
response but also with a controlled degree of directivity.
The following is not designed to be the last word on
the subject, but to point to practical things you can do.
My late father was an acoustic engineer of some note, so the following is
kinda in my blood. When I was only twelve years old he did an acoustic trick
that had engineers from many parts of Europe looking closely. It was done at
the famous Parken Stadium in Copenhagen, which on the occasion was used for
a multi-language convention. The Parken is a tight soccer stadium that held
50.000 with sharp 90° corner angles. Three
languages [translators] were to be spoken at the same times at various portions of the
programme. If you were sitting at one of the near 90 degree corners of the
stadium, you should be able to hear your language well defined, despite the
fact that you were only sitting a few metres from the other quadrant side of
the stadium where they were listening in their language.
This feat was repeated a second time a few years later with an even greater
improvement and precision. It all had to do with an understanding how the
human ear works in relation to delay, phase and amplitude and something
called the Haas Effect.
Back to the subject:
I have decided to explain (attempt as best as I can)
the key issues separately so as to understand the whole better. They are:
The Ear (you are stuck with them, so safeguard them)
The Room & its
RSF (Reverberant Sound Field)
The Loudspeaker (and its Power Response)
Largely the last two needs to be compatible, but they
do not need to be exactly tailored to each other, as some have suggested.
Maybe in a one off situation, you can have such a luxury. But there are
things you can do to tune a room and make a nice improvement.
1. The Ear. I want
to keep it simple. It is really the Ear-Brain as it is impossible to
separate the two. When the ear has a chance, it will latch on to the leading
edge of a sound wave. The ear gives prominence to direct sound over
reflected sound, a proximity effect. But having said that, the ear does NOT
ignore the overall sound field. The reverberant sound field can colour the
direct sound, make it sound brighter or duller. This is easily demonstrated
and have done so many times. It is possible for the
reverberant sound field to build up to overpower the direct sound or
sometimes subtly confuse it. We cannot alter the direct sound much (that is
the speaker designer's job) other than experiment with placement and angles,
so it is the
RSF that needs adjustment. The loudspeaker designer needs to understand
these basics, at the very least. The ear? Keep it in good shape, protect it
and keep the wax out. :-)
2. The Room. There
is this thing known as the Haas Effect - and I am sure it gets a mention or
two on some of the DIY Audio forums. The Haas Effect allows the ear to
distinguish between direct and indirect sounds and allow us to locate sounds
based on proximity. If two sounds arrive at the listener from two different
sources, one direct and the other indirect, the closest one will have
audible precedence. This is noticed when conducting a conversation in a room
where many others are doing the same thing only a few more feet away.
To me, a good room has been
optimised to take advantage of the Haas Effect. Along with the Haas Effect,
it comes down to these three, delay, phase and amplitude (all these
secondary sounds are part of the RSF).
Delay: Look carefully
at primary reflections. In a conventional room there are often three. The
Floor (I like it to be carpeted), the Ceiling and the Side Wall. Coffee
tables are a definite no-no! These reflections are delayed. As delays gets
greater, the ear works less hard. It is short delays in the range of a few
milliSeconds that causes aural confusion. IF delays become larger than 30mS
they become less harmful and
we should start to perceive an echo (in normal rooms that should not be a problem
as they are not large enough). Very low delay time (differences) means
reflections become more difficult to discern from direct sound. That's
bad. So reflections sub 10mS should be carefully looked at. It takes just
under 3mS per metre, so
if you have a near Side Wall (just outside the speaker), then it is the
'difference' at the listening position that matters. This is normally in
the order of milliSeconds, or around 1.5 Metre. This delay is critical.
But we do have additional help...
Phase: When the
reflected sound arrives later, there will also be a shift in phase. It
is the in phase response from two stereo speakers that allow us to
perceive a strong centre image. Change phase and location is affected. So
the same applies to Side Wall (and other) reflections. The phase in
tangent with delay helps the ear to distinguish between direct and
indirect sounds. But we have another 'helper'...
reflected sound has a longer path to travel, so its amplitude should be
less. For each doubling of distance we lose minus 6dB. The material that
is the reflector also contributes. It can absorb (but not necessarily in a
linear frequency manner - high frequencies absorb better and is highly
tuneable) and also
diffuse, which is often a good option.
So it is a combination of above in conjunction with
the Haas Effect that helps us to
distinguish direct and indirect sound, the latter being
RSF, the reverberant sound field. The Haas Effect should mean that the
RSF has no negative effects and may in fact compliment the listening
experience provided it is not over done.
Why not shut out the
RSF completely? Kill the sound field? That will allow the ear to
completely latch on to the direct sound from the speaker, right? Maybe so,
RSF is a bit like salt, without some of it, the sound in that room will
sound flat, bland and dull. In every woken moment we are surrounded by some
reverberant sound field. Take it away and you have been transported into a
very sterile environment. We need some reverberant field but one that does
not obfuscate the direct sound.
Whatever the room we have, we
should seek to find ways of adjusting the RSF and like salt, adjust by ear
and by taste.
IF in doubt, diffuse: Perhaps the missing
link... I mean missing piece, is diffusion. The beauty of diffusion
is that it optimises the Haas Effect like nearly nothing else. It doesn't
kill the reverberant field (although a considered amount of absorption is
still required). Yet it still delays, does useful things to phase as well as
a more than useful reduction in amplitude.
The time it takes a reverberant sound field to
effectively die is called RT60. There
will always be arguments as to what RT60
to aim at (and at what frequency). We don't need to seek the answer here but
we will aim at a good audible result and the RT60
will look after itself.
It does not mean the
RSF entirely dies, but how long it takes for the decay to reach minus
60dB. Eventually all sound in our room has to decay to nothing. So we have
to have some absorption. We should also have diffusers, especially useful
for primary reflections that would otherwise reach the listener.
Then there is an often overlooked fact, the boundary horn effect. If two or
more surfaces meet together and approaches the perpendicular, there is a
horn effect. The horn effect is an efficient acoustic device. This can cause
audible ripples of sound. When avoided the room seems to take on a calmed
quality. To some it may even strike as being deader sounding, but standing
mid room and clap your hands and you'll realise this is not necessarily the
case. (You can judge a fair bit about the room using the clapping test, such
as zingy high frequencies.)
Do You Like Triangles?
Triangular shapes that stops sound getting into top
corners are very effective. Make the outside of the triangles reflective and
the inside, facing the corner, will benefit from having an absorbing
surface. This may strike as counter intuitive, but the aim is not to absorb
sound but reflect it. As for any sound that gets past the outside reflector
and behind and into the corner, this should be absorbed (or reduced) before
any chance of it escaping into the room.
Strips that are long and with similar
reflective/absorbing qualities can be applied to the boundaries between
corners, walls and ceilings. These are slightly less effective than the
triangles, so do this only after the triangles have been fitted.
There are already specialist companies making
commercial products, such as
www.eighthnerve.com. They use 15 inch Triangles and 48 inch by 8 inch
for corners and wall/ceiling joints.
Of course, the room can be treated using Gyprock
(plasterboard) to achieve the same reflective effects.
To end the discussion, analyse your room carefully,
add only as much absorption to the room as needed, no more. A carpet is a good idea as it will
counter the most direct reflection which is from the floor and has the
least/smallest time differential. Some additional absorption will still be
required. I like a series of smaller wall carpets, but don't overdo it.
Listen to the tonal balance when doing this, clap in the middle of the room
and listen to flutter echo effects. These must be eliminated but no more.
Look at potential side wall reflections, a shelf with
books do a nice job of partially absorbing and diffusing. Look at other ways
to diffuse - use your imagination. Finally add the triangles and boundary
At the very end, see if you can lessen the
amount of absorption. You will hear noticeable shifts in tonal balances. As before listen to this tonal balance and let that be
your final guide. Also experiment with the toe-in of the speakers as they
also affect the tonal balance and changes reflection ratios. Keep doing this until you are confident that
you have the best result.
Please note this: Although this is not part of the
direct discussion here, I have noted that in many rooms there is a
particular spot where the speaker sounds best. If this is the case, then
your first aim is to find it, then optimise it to suit your choice of
speaker. The problem is that you may get discouraged
before you find it, but do not get deterred. Especially the distance from
the rear wall can be critical and too close (the wife likes them there) and
as little as 15-20cm can make a huge difference.
I hope you will have found the above to be worthwhile.