Far and away the most commonly used method today is ``sampling synthesis'' (also called ``wavetable synthesis''). ``Sampling'' an instrument means simply to record it under a wide variety of useful playing conditions. For example, every note of a piano can be recorded at a wide variety of key velocities. Sounds corresponding to velocities in between the recorded velocities can be approximated by interpolating the two sounds recorded at the nearest lower and higher key velocities. Similarly, not every key needs to be recorded, and it is quite typical to use one recorded sample to cover several adjacent keys by modifying its playback rate to scale the pitch. Additionally, memory can be saved by looping the steady-state portion of the sound. Sampling synthesizers have been around for at least a quarter century (the Fairlight sampling synthesizer was available in the late 1970s [174]). The great majority of current synthesizers use this method with various refinements.
The great advantage of sampling synthesis is static fidelity. That is, the quality of the sound produced is limited only by the quality of the original recorded sound. Since any sound an instrument makes can be recorded, there is no fundamental lack of generality in sampling synthesis. Anything is possible at some price.
The two major drawbacks of sampling synthesis are