This is where you are wrong I believe. If the DSP is built into a full-range amp, the entire frequency rage signal has to suffer the processor. Only if you split the signal first and send bass only to an amp with DSP while the rest goes to another amp without DSP can you avoid the problem described. I have to agree totally with Amir that the top end is degraded as a result of being subjected to this complex processor and this can be easily demonstrated with high quality speakers. In fact the DSP in my amp can only adjust sub-500 Hz yet the top end is clearly slightly spoilt.
Well, I realize that you may not even share the same beliefs as me so we may as well be talking past each other. But in the interest of making my argument easier to follow, I will list all my underlying reasons why DSP is effectively transparent to any part of the frequency range that you leave unmolested.
- First, digital signals ONLY consist of two elements: 1's and 0's, and timing (jitter).
- Transmission of 1's and 0's is mostly perfect. If it isn't, there would be obvious signal dropout which is audible as interruption to the music.
- Jitter can be reclocked. If bits are not changed, then there is no difference whether it goes directly from CD to DAC, or from CD, transmitted to Timbuktu, transmitted back, then reclocked and into the DAC.
- Running a signal through a processor, and having the processor NOT do any manipulation with the signal, will result in exactly the same signal coming out the processor.
If we can agree with the above premises, then we can move on to the next points:
- It is possible to send a full range signal through DSP and only apply DSP to the frequencies you are interested in.
- Somebody mentioned delay. It is also possible to delay part of the signal and leave the bits completely unmolested so that when it is reproduced by the DAC, it is still bit perfect and will result in exactly the same signal.
Since you mention "high quality speakers", I will make a bold claim that there is no tweeter on this planet that is more resolving than mine, except for other manufacturers which use the same technology: plasma tweeters. No inertia, no momentum, no distortion, and no "character" from materials used to construct the tweeter. Sound literally emerges from thin air (or rather, a ball of plasma). I also use horns, like you. It is probably the only active DSP controlled speaker with a plasma tweeter in the world, and if it isn't, I would be interested in meeting other members of what must be a very small club. I am not using some low end DSP solution either, I built my own PC which upsamples everything and processes it at 64 bit with 65536 taps. The signal is then sent to a Merging NADAC which reclocks everything.
So yes, I can hear every little change that goes on in my system and I can tell you that if I do not apply any manipulation to the top end, there is absolutely no difference.