When is music too loud for a headphone? - Page 2

When my room mate comes into my room telling me to lower the volume on the speakers.

Distortion due to the limited range of the movement the diaphragm is capable of. But I admit this is indeed redundant because it would fall into another already mentioned category anyway.
 

So playing a square wave (the most clipped waveform possible), regardless of frequency, and even at a relatively low level, is dangerous. Can you give a physical explanation of this, or a source ? I do not necessarily doubt it, but more information is better.

Of course, high SPL is dangerous by itself. But I am more interested in how a "distorted" signal well below the rated power is damaging a driver, and especially actual numbers. In the best case, real word data on how the power handling capacity of any particular transducer changes with the frequency and level of clipping applied to a sine wave (right until it essentially becomes a square) would be the most useful. Alternatively, if I take a musical signal, and start to apply clipping to it, with no change to the gain, such that the peak level is gradually reduced from the maximum to zero, at which point is the damage potential the highest typically ?