The DSP allows the amp very precise control over the cone's movements. I'm not 100% sure of the inner workings. But by delaying the signal slightly (DSP subs need additional time delay applied to the other speakers in the receiver) it can make changes to the waveform to counter act the physical limitations of the driver. Sort of how an amplifier uses feedback to look at the output signal vs. the input, and then apply an inverted difference (the distortion) to the input to reverse-distort the input so the output is cleaner. If you know the inertial effect of the cone, surround, air, then you can pre-compensate for how you want the cone to move vs. how it is actually moving. I don't know if this is done from reading some sort of feedback, or from pre-computed tables for each system. But I've read that this absolute control over a speaker cone can only be applied up to a few hundred Hz. Which is why it can't be done with anything but subs (for now).
