Eiron wrote:
Serge Auckland wrote:
Cessna172 wrote:

speaker manufacturers, to make sure their product is technically
brilliant, yet get them into a room and they will sound anything from
horrible to superb.


I've yet to see a product that didn't work properly when the sums were
done correctly, rigorously, and applying the rules. That's not to say
that an intuitive designer can't get a good sound without doing the
sums, what I'm saying is if you then analyse such a design, you will
find that the sums come out right as well.

What about the perfect speaker - a point source with a flat frequency
response?
It will sound perfect if your room has no walls or floor.
If you design the perfect speaker for your room, it may not work well in
mine.

Two points from the above:-

Firstly, there is no practical way of producing a point source with a
flat frequency response and sufficient audio output for music
reproduction. Quad has come the closest, I suppose, with a sound-field
synthesis for a point source, but this is a long way from a real point
source. Consequently, there are no "sums" to be done with this example,
as it is not realisable.

Secondly, if you were to design the "perfect" 'speaker for my room, that
is exactly what it would be, the "perfect" 'speaker for *my* room, not
yours or anyone else's. You would have to design a different "perfect"
'speaker for every different room. This is clearly not a practical
commercial proposition, but theoretically of course, it could be done.
Incidentally, you would need to define carefully what "perfect" meant,
but let's accept, for the sake of this argument that we both understand
what we mean by "perfect".

In pursuit of the above "perfect" for every room, several manufacturers
are now including DSP control with measuring microphones so that their
'speakers can be better matched to the room. Whilst an improvement, you
cannot DSP away the floor and walls which cause echoes, nasty
resonances, frequency response anomalies caused by frequency-variable
absorptions in furniture, (although this last is probably the easiest
for DSP to make a difference) and a whole host of other stuff that we
all live with.

Many years ago, I did some 'speaker measurements in an anechoic chamber,
and as a working environment it wasn't very pleasant. I suspect this was
because the acoustic space ( effectively no walls, floor or ceiling)
didn't match one's visual space, which was that of a small room.
However, listening with eyes closed and the room darkened, did actually
sound very good. More recently, I visited a Scandinavian Broadcaster
who had a number of IEC standard listening rooms of different sizes in
their centre. They were using B&W 807s, and the sound was quite
extraordinarily good. If you ever get a chance to visit any UK
commercial radio station that still has studios made during the old IBA
Code of Practice days, you will again hear what a really good room can
sounds like.

S.