B&C Engineer also mention quality of material and other variable, like expected life for your speaker. So there are a lot of variable, but one of interest of this thread is the music program material. In general, recorded music are more compressed and processed them live music. So for a while I think that is a myth that any music file would have Crest Factor of 3, it would be sine wave and not music, all instruments has fundamental frequencies and harmonics, so there will be summation and and crest factor will go higher.
lets see what people brings to the table.
Since you're the original poster, and have all the material readily at hand, you could perform the required "deep analysis" so that we can learn something? That would be infinitely helpful under the circumstances, as a favor to the community.
The crest factor of real music will go orders of magnitude higher than a sine wave’s, due to transients.
It isn’t clear what exactly you’re looking for here.
I've gone and captured the waveforms from a selection of about 10 of his tracks. Sure, the occasional waveform looks a bit squared off, but it's a fleeting effect at best. Synthesizers have the ability to generate sine waves, square waves, triangular waves, and no doubt a few others as well.
Below is one example of an Armin van Buren waveform. It's possibly looking a bit triangular early on, but it gets more sinusoidal further along.
Here's a section where at least one half of the waveform is looking a little more flat topped.
And here is something reminiscent of a saw tooth waveform:
Those pesky square waves seem to be quite difficult to catch in the wild. I think I need Sir David Attenborough's help with this. 🙂
Let me know the music so I can try to get a file to measure.
This one can be close to sine wave, but look to the amplitude (0,1) and compare to the sample from the first picture (0,8).
I think too, maybe DnB style. I remember in the 2000' in Brazil some CD flooding the car market just for testing purpose once the music were weird to listen, but they were hard for the system to reproduce. Unfortunately I have no sample.