"Why does digital clipping occur? I would think that if the input level reaches the maximum level of the converter, thus creating a digital representation of all 1's, anything over that would be ignored. How can a sample be distorted if the information is only 1's and 0's. Shouldn't the clipped sound just be registered as full scale (all 1's)?"Hate to answer a question with a question, but what good would that do? If the signal were truly all 1's there would be no way to differentiate a clipped flute from a clipped bass guitar. (Actually there's more to this - all 1's would never work at all. See WFTD Two's complement.) Basically it's an unrelated sound at that point either way. Although looking deeper at your question it is important to point out that all samples have some amount of distortion. We understandably continue to try to apply our analog understanding of the world to digital processes, which just doesn't work most of the time. In analog systems the distortion rises gradually once you begin to approach the limitations of the system. This produces a character of sound we have learned and adapted to. In digital the distortion actually goes down as the levels get higher, until you reach full code, and of course after that it is ruined. This is true not because the digital sample is distorted; the converters essentially do ignore anything over the maximum voltage they can mathematically capture. It is loosely analogous to the most extreme form of analog clipping: set up an amp with plus and minus 10 volt rails and try to get it to reproduce a 50 volt peak to peak signal. The resultant output will not resemble the input much at all. If you start with a sine wave you will end up with a square wave of the same frequency. If you start with a complex music signal you will end up with something resembling noise.In typical scenarios (read not extreme) where analog signals are clipped the voltage doesn't stay the same for very long. In order for it to be characterized as any kind of sound it has to be a periodic waveform (otherwise it is just a DC voltage). Consequently the signal spends a statistically significant period of its time in what we have to call, for lack of a better term, a partially distorted state. Remember, in analog electronics distortion rises gradually as you approach the limit of the system. This increased, but not complete, distortion is a state that doesn't really exist in digital, at least not in the same way it does in analog. It is no doubt theoretically possible to produce a digital quantization system that more closely approximates the distortion behavior of analog. It is also possible to turn tin into gold, but there are other more practical ways of obtaining gold so we don't bother. Why not just operate within the linear range of digital, letting it do what it does best, and get the analog 'sound' you want from analog equipment? However, it is interesting to ponder that designing digital converters with the idea of having a 'sound' of their own based on someone's idea of how some analog system behaves could potentially launch a whole new distinction in the audio industry. No longer would people be agonizing over jitter and clock rates. Instead there would be wide spread conversations pertaining to the way converter systems color the sound at various energy levels and frequencies. Non linear operation would be the rage, but not just any non-linearity; you've got to have the XYZ brand because their math was developed by... I think we see where this is going. Marketing people, this Tech Tip will self destruct in 10 seconds!