I thought the graphs would be the same. Slight variations can be seen between 1-4khz. There is more of a tilting into the valley trend with the AVR and a bump right before 2khz. May be nothing.
I dont test using sweeps most of the time, the reason is as you stated, variability. I run pink noise and set the FFT to average the spl per frequency over time. I give it 30 seconds or so, but usually the graph has fully stabilized by the 7-10 second mark.
I use 1/12 octave resolution (twice what we are thought to be able to hear) because 1/3 or even 1/6 resolution can hide narrow high Q problems. Not that we are very good at hearing those either.
As a curiosity, sighted, I preferred the sound at first of the AVR over the power amp..... Once I started to turn it up however, the Amp remained sounding composed while the AVR started to sound brassy. Its hard to describe, but guitars start to sound more twangy with the AVR at higher volumes. Audible distortion perhaps? I would have never noticed it without a comparison like this.
I could be imagining things. I should re-run the measurements at a higher db and see.
I used the track "Timshel" from Mumford and Sons album Sigh No More for testing.
