Part of the difficulty, for a long time for me, and clearly for others, in assessing distortion is the coplete (to me) counterintuitive (not to say utter incompatibility with comon sense) of the fact that complex waves are made up of a whole bunch of sine waves in specific relationships to each other. SO, despite the fact it makes no immediate sense to me, truth is that a square wave consists of a fundamental plus odd harmonics. If the voltage of the fundamental is V, then the third harmonic (H#) is 1/3 of V, the fifth harmonic is 1/5 of V, the seventh harmonic is 1/7 of V, etc. (Note that because of ohm's law, one third voltage connotes less than one third power).
Note that since, AFAIK
, no amp in the world produces only odd harmonics when distorting, the even harmonics will make it impossible for the amp to ever actually produce a square wave, because a square wave contains no even harmonics.
The notion that complex waves are made up of sine waves is not new stuff. People ascribe it to J Fourier, a French ella who lived rom 1768 – 1830. So it ain't new stuff. A pretty good tutorial, with oscilloscope demonstrations etc of this with respect to a square wave is at
As a practical matter, those power-versus-THD plots we've seen above and in hundreds of amp tests come from application of a technology called FFT, which stands for Fast Fourier Transform. IOW, the test results depend, in an absolute way, on the accuracy of the Fourier concept that complex waves are made up of sine waves. If the underlying theory that the wave is made up of many sine waves is wrong, then whatever an FFT spits has to be wrong, too. So it would be pointless to continue discussing those meaningless results.
Moreover, AIUI, digital sound relies on FFT to recreate audio from the digital representation. So if the complex waves on CDs, or in the sound of things we get off the internet, are not composed of a series of sine waves, then the entire world has been laboring under a massive delusion that we actually get sound from CDs or off the internet. Not possible, if the underlying theory is wrong. Yet, I seem to hear sound when I play a CD, or when I watch things on youtube. Tends to make it likely that the theory is correct.
For me, seeing with my own eyes a sine wave transform toward a square wave on a scope as the higher harmonics are added made the concept, if not intuitive, at least concrete. I'm no longer uncomfortable with it . . .