Jim,
No, I didn’t measure the amp’s power output, but I have an SPL meter that shows peak volume levels. My typical listening, while sitting on my couch, has peaks around 70db. The speakers are about 2 metres away.
My red disc was made in photoshop using the circular marquee tool and the paint bucket tool set to a pleasing shade of red and opacity at 15%.
Really, at the levels I typically listen, I could power my speakers with my headphone amp. Their sensitivity is about 100db for a watt at a metre.
I would wager that someone with small inefficient speakers powered by a big clean solid state amp has more distortion in practice because the speaker’s midrange driver is pretending to be a woofer.
Joe
Your comment about the 'red disc' answers a different question to the one I had in mind.
However now you've said you use a peak reading meter, that helps with my intended question/point. How do you know the reading *is* a true peak for the level into the speaker, and is reliable?
That said, yes, it looks like you are avoiding limiting *if* the readings are reliable and *if* the plots of distortion versus power into three resistances of light-bulb for a sinewave can be used for music into your speakers. But that still leaves factors like the way output transformer performance deteriorates for lower frequencies and asymmetric waveforms. Sinewave-into-resistor-at-1kHz is a particularly iffy way to rate the behaviour of many valve power amp designs. And designers can 'target' this, optimising for the 'usual test method' at the expense of other situations - either deliberately or being unaware of the problems.
Have a look at the table on
http://www.audiomisc.co.uk/HFN/HeadphoneDAC/HeadDAC.html
showing how a small signal transformer's THD varies with frequency as an illustration. That was, of course, a tiny signal transformer used at low current levels. But a similar pattern tends to arise in larger transformers with the distortion rising rapidly as you get to low frequencies unless you have very large transformers. This gets worse for some designs of power amp when there is some LF offsetting the 'balance' of the transformer sections as you play HF.
Again, all that said, the output impedance also varies and affects the result, and this tend to be so even at low powers. Result is a form of 'tone control' that may give a sound someone prefers.
The basic problem here is that although valve amps look simple in design terms, the need for output transformers is a big challenge with very complex behaviours that most users, reviewers, and I suspect designers, don't really take into account. Hence no surprise if they 'sound different'. Simple standard measurements won't show why.