1db. Wtf. Audio is a strange game.
And in no small way, because the "mind plays tricks..."
As Craig inferred, this is always fun to do for the sake of fun, but realistically, one requires a very controlled setup to determine the most valid results.
It has long been published in science literature that human audio memory is very short. Without an instantaneous switching method the observations anyone makes (and believes are real) are subject to bias. It doesn't matter if a person had a 20 year musician in the room, or their wife who never heard the setup the first time, etc. etc.
All that being said, and taking Craig's comments re: room atmosphere alone into account, the two graphs produced are EXTREMELY similar (overlay the two and you will see how little difference exists; anything at delta 1dB for a minor frequency range would not likely be audible). If repeated and averages over multiple trials, i would doubt there would be any statistically significant differences in those curves and for the most part, that was already demonstrated for most of that spectrum reproduction.
The suggestion that the AVR is being strained in its power output is certainly plausible depending on the playback SPL and speakers being driven. I have found our AVR to have a change in sound as SPL increases much earlier than i ever expected, though the point at which i "perceive" that to be happening is still pretty loud (>85 dB at 11 feet).
