by, if you are concerned about Faldage's use of the term "sampling error" be assured that he is correct. In sampling anything, there is a requirement that to prevent aliasing you must sample at a frequency beyond the "Nyquist frequency", which is twice the frequency of the maximum frequency you are interested in accurately reproducing. If there are frequency components higher than that in the process you are sampling, then they are "aliased" and it is indeed referred to as "sampling error". Anyway, with the wheel turning much faster (more than half as fast) than the movie frames are going by, you experience a real-life version of sampling error. I was trying to draw a nice graph on Matlab yesterday to help explain this but was sidetracked by "real" work.
A non-graphical example would be the sampling rate of CDs. The human ear is sensitive to sounds (on average) between 20 Hz and 20 000 Hz (20 kHz). So if I want to accurately reproduce the highest frequencies I can hear - the 20 000 Hz ones - I must sample at least twice as fast as that. So CDs are typically sampled at about 44 000 Hz (I think it might be 44 100), or there's another standard that calls for about 48 000 Hz. Both are twice as large as 20 000 Hz, so aliasing of audible frequencies shouldn't happen.
DISCLAIMER: Please don't start arguing with me about the sound of CDs and analog and why one is better than the other. I was just using the CD as a sampling-theory example.
