by Sami Sumea

If you understand thoroughly what sampling actually is, you'll also understand what kind of limitations and possibilities it gives to you for sound sculpting.

The basic idea behind sampling is to transform a signal (usually a signal you can hear) into a stream of digital numbers by "sampling" the amplitude of the original signal at equal steps.

Every sampler (or sample playback system), no matter what kind of special features it has, is based on converting these samples back to "useful" data (into analogue or digital form), again at equal steps.

The samples can be transformed back to audible signal (or added into a pipe of digital signals in digital systems) by a set of rules specified by the user and the sampler. User specified things such as desired pitch of the sample usually affect the rate the samples are picked up from the memory of the sampler.

The user may also decide what kinds of effects or filtering should be applied to the signal.

It's up to the sampler how the sample information will be transformed into reasonable data. For example, when playing back a sample at higher pitch than the original, some older samplers just skip a certain amount of samples in the memory. In worst case this will produce horrible audible aliasing or a "chipmunk"-effect.

Some samplers (like NI's Kontakt) actually let you "stretch" the original signal by a combination of organising the samples into tiny groups and filtering them in a special way.

You may also be able to apply weird modulation or other effects to the resulting signal with your sampler, but the basic idea of picking up digital numbers from the memory and turning them into useful signal still stays the same.