As far as the crest factor is concerned (towards the end of the article) - there is no such thing as 'peak RMS level'. Crest factor is a ratio between peak and RMS level of the signal. Period. When you say: "A steady-state sine wave has a crest factor of 0 dB, because its peak level and its average level are the same", it does not make any sense and only attempts to prove that there is no difference between peak and RMS levels, despite what you have said a few lines earlier (factor of 0.7, which corresponds to roughly 3 dB).

You're right, of course!

That passage was written quite some time ago (I was younger, cuter and sloppier), and I was trying to finesse the idea of the loudest RMS value (hence it's "peak" value - cognitive chaos!) in comparision to an "average RMS level.

Sorry about that!

I would now say something like:

"Crest factor expresses the relationship between the long-term average level (Leq) and the typical peak magnitude of the signal trace. For a random signal such as pink noise, that crest factor is usually a little less than 12 dB. For a constant level sine wave signal the crest factor is almost zero, because there is no variation in the peak magnitudes of the signal trace"

I would avoid a discussion of the difference between peak and RMS values here, because they don't represent differences in level, just differences in measurement metric, whie crest factor expresses level variation.

Thanks for pointing out the error of my younger, cuter and sloppier ways. I've been advised by my doctor that it won't be happening again!

Best regards,