Please explain clipping

Clipping occurs when an amplifier stage cannot supply either enough output voltage swing or current to the load. It is not really related to frequency, except that speakers often have a lower impedance in a certain frequency range (usually lower freq) and that is where the amp would tend to run out of gas when attempting to respond to a high amplitude (loud) passage in that range. This has been discussed in other threads at some length, you should search. The result of the rapid change in voltage / current when the limit is reached, is generally a high frequncy transient (spike) that typically would just be too much for the high frequency drivers, thus fried tweeters. Such damage is therefore usually the result of inadequate electronics rather than inadequate speakers. It has also been pointed out in another thread that clipping can occur in any preamp / amp stage and have sonic implications, but is usually thought about with respect to the final amp because of the physical consequeces. :)

Sf- Clipping is the inability of an amplifier to increase the amplitude of the input signal while maintaining the same waveform shape as the input signal. The peaks of a sinewave will no longer be round, but "clipped" flat. (This would be an example of severe clipping). Some amplifiers produce excessive amounts of distortion when this occurs and don't recover quickly or gracefully from clipping. Regrettably, without nearly limitless amounts of power, clipping is bound to occur sometimes. I seem to remember an MIT test where the reproduced sound of a pair of scissors still showed signs of slight clipping even with 2,000 watts available. Single-ended triode designs,and tubes in general, tend to handle this situation in a manner more pleasing to the ear, though they obviously have greater measureable distortion products than other topologies.
Hope this helps.

Sfrounds, Jcbtubes has given the best explanation. As an aside, there is evidence that between well designed amplifiers, the amp's recovery characteristis from clipping may be a key determinate in listener sound preference. Some amps go into short-term instability after clipping, while others seem unfazed.

Jcbtubes, I thought I knew what clipping was, but now I wonder. From your post, it appears that clipping is not amplitude related, as I thought it was -- did the 2kw amp produce clipping even at very low volumes? Was it more a function of slew rate or other factors? Are there two kinds of clipping -- one related to design and one related to overstepping the design? Sorry for all the questions, but I would appreciate any insight. Thanks for the post.

Clipping is simply the amplifier trying to output more power than it's power supply is capable of. It is not a function of slew rate but simply the amplifier is being asked to provide more power than it's power supply is able to handle. The term clipping came into use because of the oscillographic picture of a sine wave output of an amplifier in clipping looks as if it had the top cut off. Clipping causes distortion based in the odd harmonics. Slew induced distortion or Transient Intermodulation Distortion is a function of amplifier bandwidth or speed. If an amplifier has power in reserve it cannot be considered in clipping until that reserve is used up. That is why the spec of headroom is important. If an amplifier has a headroom spec of 3db it can pass 2 times the power it is rated for. Unfortunated speaker loads are not linear over frequency. While the nominal impedance may be 8 ohms the impedance may at some frequencies be as high as 20 ohms or as low as 2 ohms. It is possible that at the majority of frequencies our amplifier can supply enough power to drive the speakers but if the impedance requires more power than the amp can muster, the result is, you guessed it! Clipping.