SSCSSS Version 2.0 - What is Aliasing?

There is a mathematical theorem to the effect that the highest frequency that can be sampled is one half of the sampling frequency. I am not going to get into all the math (mostly because I do not know all the math), but the basic idea is simple.

To represent a frequency, at least two sample points are needed: one high and one low. This is why the frequency must be no more that half the sampling frequency. Two sample points are needed at a minimum for each cycle. Aliasing is what happens when a digital recorder tries to record or a digital synthesizer tries to generate frequencies higher than half the sampling rate.

What does happen, if the "right" thing does not happen? The frequencies over one half the sampling frequency are "reflected" around one half the sampling frequency. For example, the EX's sampling frequency is 44,100 Hertz, so the highest frequency that can be successfully generated is 22,050. If it "tries" to generate a 25,000 Herz frequency, what will actually be generated is 22,050 - (25,000 - 22,050) or 19,100.

In general, the reflected frequencies will not be harmonics (integral multiples) of the intended pitch, so the "timbre" will change. How much it changes, is due to specifics like the relatives loudness of the reflected (aliased) frequecies relative to unreflected frequencies.