In addition to the experiment Jim suggested, did you try this experiment I suggested earlier:
... if you are still using the Plinius Odeon amplifier you mentioned in a post a few months ago you might try changing the setting of its ground-lift switch to the opposite of what it is currently set to. It would be prudent, though, to turn the amp off and wait about 30 to 60 seconds before changing that setting.
While Bryston provides schematics at their website for many of their models, unfortunately they don’t for the SP3. And neither does hifiengine.com. But I took a look at the schematic for the SP1.7: http://bryston.com/PDF/Schematics/SP1pt7_SCHEMATICS.pdf
As is the case in many designs it appears that the muting function is accomplished with relays that short the various outputs to ground when the mute function is activated. It also appears that the outputs are AC coupled via 100 uF capacitors, which are "ahead" of the relays in the signal path. I would expect that the SP3 is similar in those respects. If perchance you have the ground-lift switch on the amp set to the lift position, while I can’t formulate a complete explanation I’m thinking that the issue might be related to the fact that with the switch in that position and no signal existing within the SP3 the AC coupling would allow the input of the amp to "float" to an uncontrolled and arbitrary level relative to its chassis and to AC safety ground and to the chassis and the circuit ground of the SP3, **if the muting relays are not activated.**
That may or may not have anything to do with the problem, of course, but intuitively it seems to me that it's a possibility. And in any event if you are using the Odeon the experiment I suggested is easy enough to do. Also, as Jim suggested contacting Bryston and describing your findings to them may be worthwhile.