We don't manufacture dipole speakers anymore because we don't see a need for them. With a discrete, full range surround format you don't want your rear channels to diffuse the sound. All of the speakers in the system should have the same radiation characteristics. Any ambience in the sound field will be mixed in by the producer and engineer, not by the speaker design.
With Dolby ProLogic, it was very helpful to have rear channels that diffused the sound. The rear channel was mono and there was no way to mix ambience into it other than the delay that the processor added to it. In such a crude system, a dipole does a good job of confusing your brain and keeping you from localizing the sound sources.
The NHT DSP speaker has a total of four analogue inputs. Two balanced and two unbalanced. There are no digital inputs. From the user's standpoint, the processor behaves as a multichannel power amplifier. They still need a preamp/surround processor with level adjustments, volume control, source selection, etc. The processors do not need to be linked together in any way.
We really wanted to put a digital input on the processor, but there are a number of technical and practical problems with doing this. There is no accepted consumer multichannel digital transmission format yet. If we use an SPDIF digital input, then the processor needs to have a volume control, which of course requires that units be ganged together. If it has a volume control, then we have to supply a remote control for it. Now the consumer has 21 remotes:) These are just a small example of why we kept all of the preamp functions out of the processor.
Our goal with this product is to give the user a product that sounds fantastic out of the box, and doesn't require lots of complicated setup. Most products with DSP give you too many options. This product is designed to make the DSP work for the user, not the other way around.
The processor is designed to drive any two speakers in the system plus have two extra line level analogue outputs to go to the subwoofer. A typical 5.1 system would have one processor drive the L and R speakers and send a feed to two subwoofers. The second processor would drive the LS and RS speakers and the third processor would drive the C speaker. You are of course left with one extra channel, which could be used to drive a rear center speaker.
I've been using DSP based loudspeaker controllers, to do crossover and eq functions, for a number of years, however the DEQX process is quite a bit different. It has very high slope, linear phase crossovers (0-300dB/octave) with low latency (delay time). I'm not aware of any other processor that can do this. On top of the crossover functions, the processor equalizes the phase response of the system to have a constant group delay. In the near future, the processor will support room correction processing. For more information on the DEQX process and existing products, go to www.deqx.com