The process of dropping samples from a stream of oversampled digital audio data. Typically decimation is employed to lower the density of the data to a point where it can be more conveniently stored. For example, in a 44.1 kHz sample for CD, the original sample/recording may have been done with 64 times oversampling, which results in an effective sample rate of just over 2.8 MHz, but decimation can reduce this data back to what is effectively a 44.1 kHz rate. This is important because there are many benefits for taking the original samples at higher rates that are not lost when the data is thinned out later. Of course, some would argue that we’re throwing away perfectly good data, hence the increased popularity of higher sample storage rates (96 kHz, 192 kHz, etc.).