Brian Lenharth,
1997-10-31 19:04:44 UTC
At recent conferences in USA and England, a new kind of flat
loudspeakers has been presented. They are called Distributed Mode
Loudspeakers (DML), and work on a quite different principle than the
normal direct radiating units. The vibration of a thin, light plate is
made as random as possible, obtaining a sound radiation which is almost
perfectly omnidirectional (even at higher frequencies), mixed-phase
(this means that no interference occurs between different speakers or
even between the mirror images created by room reflections) and
incredibly efficient (nearly 50%, compared to 1-3 % of direct radiation
loudspeakers).
These flat units are inexpensive, and they can be easily incorporated in
the ceiling or in lateral walls. Their only limit is toward low
frequency (they start at about 100 Hz), but a single subwoofer can cover
very weel this range.
Due to the intrensically diffuse nature of the radiated sound, no "sweet
spot" occurs. Obviously the localisation of concentrated sound sources
is ligthly blurred, but at listening tests I founded these units very
pleasant and natural, particularly with classic music. I listened also
at discrete 5.1 soundtracks during the projection of a film sample
(actually the projection screen WAS the frontal loudspeaker), and the
surround effect was certainly better than with traditional loudspeakers!
I think that this kind of loudspeakers can be a good starting point for
introducing in our homes 8 discrete sound sources (or even more). It is
impossible to fit in my living room 8 normal sound boxes, I already have
problems with 2!
I think, anyway, that using DML loudspeakers the Ambisonics decoding do
not work anymore, because this is based too much on the phase of the
wavefronts: with DML there is no more the wavefront, nor the "phase" as
usually is defined... These loudspeakers really need some time to be
understood!
I have read some of the info on these new-fangled speakers. Maybe I'm justloudspeakers has been presented. They are called Distributed Mode
Loudspeakers (DML), and work on a quite different principle than the
normal direct radiating units. The vibration of a thin, light plate is
made as random as possible, obtaining a sound radiation which is almost
perfectly omnidirectional (even at higher frequencies), mixed-phase
(this means that no interference occurs between different speakers or
even between the mirror images created by room reflections) and
incredibly efficient (nearly 50%, compared to 1-3 % of direct radiation
loudspeakers).
These flat units are inexpensive, and they can be easily incorporated in
the ceiling or in lateral walls. Their only limit is toward low
frequency (they start at about 100 Hz), but a single subwoofer can cover
very weel this range.
Due to the intrensically diffuse nature of the radiated sound, no "sweet
spot" occurs. Obviously the localisation of concentrated sound sources
is ligthly blurred, but at listening tests I founded these units very
pleasant and natural, particularly with classic music. I listened also
at discrete 5.1 soundtracks during the projection of a film sample
(actually the projection screen WAS the frontal loudspeaker), and the
surround effect was certainly better than with traditional loudspeakers!
I think that this kind of loudspeakers can be a good starting point for
introducing in our homes 8 discrete sound sources (or even more). It is
impossible to fit in my living room 8 normal sound boxes, I already have
problems with 2!
I think, anyway, that using DML loudspeakers the Ambisonics decoding do
not work anymore, because this is based too much on the phase of the
wavefronts: with DML there is no more the wavefront, nor the "phase" as
usually is defined... These loudspeakers really need some time to be
understood!
'optimism challenged' but this all sounds too good to be true. It sure
would be nice to have a low-cost virtually perfect transducer that's
unobtrusive, highly efficient, etc. But I'm not holding my breath at
this point.
A new decoding theory is needed, based on the simple energy vector
panning of Ville Pulki (JAES June 97). I shall experiment this as soon
as I get 8 DML loudspeaker in my laboratory.
Let us know when you get them. Don't forget to put one on the ceiling, justpanning of Ville Pulki (JAES June 97). I shall experiment this as soon
as I get 8 DML loudspeaker in my laboratory.
like they do at Holiday Inns.
bel
--
Brian E. Lenharth Tele: 206-335-2389
Lake Stevens Instrument Division/ MS10 FAX: 206-335-2828
Hewlett-Packard Co. ***@lsid.hp.com
8600 Soper Hill Road; Everett, WA 98205-1298