This whole discussion hints at a feature I have wanted from tube preamp manufacturers out there: implement HT pass-through WITHOUT the need for the tube unit to be powered on. Why is this so difficult? If all we want is to route a pair of inputs directly to the outputs, can't this be done through the use of one toggle switch? And in the other position, which perhaps is an open circuit, the preamp works in a normal way when powered on.
As for all the back-and-forth posts above, no matter how the Ref3 may differ from the LS25 sonically in "normal" mode, I gotta believe that they would NOT differ in HT pass-through if none of their active stages are in line for this mode. I suspect ARC's implementation (switches, wires and connectors) with HT pass-through is consistent from model to model. With that said however, it escapes me how insertion of any component (with its input and output connections and the additional IC) can result in NO kind of signal degradtion vs. the pre/pro's front channel outputs directly to the amp. In a high-resolution system, just that one extra pair of ICs can do far more damage than any inserted "pass-through" component. This may be a miniscule difference, and for the sake of HT usage, who cares, but I suspect there would certainly be a noticable change in dynamic contrasts even with the most linear ICs out there.
As for all the back-and-forth posts above, no matter how the Ref3 may differ from the LS25 sonically in "normal" mode, I gotta believe that they would NOT differ in HT pass-through if none of their active stages are in line for this mode. I suspect ARC's implementation (switches, wires and connectors) with HT pass-through is consistent from model to model. With that said however, it escapes me how insertion of any component (with its input and output connections and the additional IC) can result in NO kind of signal degradtion vs. the pre/pro's front channel outputs directly to the amp. In a high-resolution system, just that one extra pair of ICs can do far more damage than any inserted "pass-through" component. This may be a miniscule difference, and for the sake of HT usage, who cares, but I suspect there would certainly be a noticable change in dynamic contrasts even with the most linear ICs out there.