Whoa there is a lot here but I will try to address what I can and remember from your post.
I use -20dBFS b/c a 0dBFS can overload the bass management circuits if all speakers are set to small. It's an industry standard to use -20dBFS for multi channel testing. I do check 2CH audio with 0dBFS to ensure the product can handle fullscale digital. I don't show every test I do else the reports will be too long and boring for most readers.
Maybe you provide a link to additional results for those interested. Of course I know this is wishful thinking, I understand the time it takes to do this and most of your readership might not be interested in it.
For the FFT, I calculate how low the 2nd or most dominant harmonic is down from the fundamental. I also do total THD + N for the power measurements too. IF you want to calculate total distortion from the FFT, one can sum all the harmonics, but I see little point to this. The point of those plots is to show the harmonic nasties and look for any misbehavior.
I understand how to calculate it, I simply wanted to point out the usage of the term THD + N, which to implies all harmonic Noise
As for effective bits, Johnson noise will never allow you to achieve a full 24 bit resolution in the analog domain. 18-19 bits is about the best you can ever achieve in real life.
I somewhat agree with you. It is possibly to get @22 bits out the DAC & low level circuitry but by the time one goes through the power amps you are lower
HDMI crosstalk levels are ALWAYS higher than analog crosstalk measurements from what I've measured on all products. The levels are still low enough to not be of any concern IMO.
I test the preamps to output clipping and always show what they are capable of driving under a 600 ohm load. There are no appreciable differences if the preamp is driving output levels below clipping vs say 500mV under the same power drive to the amp section. Preamp circuits aren't sharing the same power supply as the power amp for virtually any receiver I ever test. As long as a preamp can cleanly drive at least 1.2Vrms, it should be enough gain to work with virtually any power amp. I prefer the preamp to be able to drive at least 2Vrms and I discuss that in our measurement standard.
I do FFT plots at various power levels and typically show them for 1 watt and fullrated power and yes fullrated power is always higher. Some amps are better behaved than others in this regard.
I was unable to find a 1 Watt FFT plot for the power amp for either 3000 or 1010, hence I spoke. A mid level plot might be useful from the standpoint that, that is where is most will likely operate this equipment.
I test SNR at 1 watt and full rated power and calculate the 1 watt SNR figure at full rated power to ensure linearity at all power levels. I usually find no surprises here.
I agree with the above, just that it is not indicative of max SNR and is not where most users operate
I scaled back many of my analog tests b/c most people these days are using the HDMI connections between the receiver and blu-ray player. I test TrueHD and DTS HD signals to make sure there is a perfect bit to bit transfer which would indicate there should be no audible differences between various players when using the HDMI interface. This assumes upconversion/downcoversion circuits are disabled b/c if they are not, they will fail the BER (Bit Error Rate) test.
Ok I think I got most of your stuff. Do you have an Audio Precision? If so, I'd be happy to get some help doing testing to cut down the time. My testing is already quite time consuming and I can rarely dedicate any more time to test individual products than what I already do. I've actually taken away tests such as output impedance b/c they are way too time consuming to pull AP data into Excel and calculate the impedances.
