This section initially supplied by Andy Church <> in 1994.

Andy writes: Although the specific methods used for sampling vary from platform to platform and program to program, the general procedure is the same. The first thing you need to do is connect your sound source, such as an electronic keyboard, to the input of your sampler. Then you should start up your sampling program (if you haven't done so already). Many programs have a "monitoring mode" where you can listen to what you will be recording before you actually record. If possible, take advantage of this so that you can adjust the sampler's input level and your sound source's volume before sampling to avoid problems with volume (see below).

When you have adjusted the volume, or if you do not have a monitoring mode, you can actually sample the sound. Usually this just involves clicking on a "Start" button, playing the sound on the sound source, and clicking on "Stop". Then you will probably be taken to an editor screen where you can see what you have recorded. You may discover one of two problems with your sample at this point (especially if you couldn't monitor the signal ahead of time):

  • If the overall volume is too low, you lose resolution and clarity in the sample. Generally, if the maximum amplitude of your sample (on a typical waveform display, represented by vertical distance from a central axis) is less than three-quarters of the total window height, your sample is too soft. For sounds that you want to be soft, you can adjust the volume later inside the module.
  • If the volume is too high, you get "clipping" when the input sound exceeds the limits of the sampler. If you see flat lines at the top and/or bottomof your sample display, or if you hear a lot of distortion, your sample is being clipped; you should reduce the volume or the sampler's input level and sample again.
Once you have successfully sampled a sound, what you do next depends on what kind of sound you have sampled. If it is a "one-shot" sound - a sound that only plays once and does not repeat, such as a drum - you are done, and you can save the sample to disk, ready to be loaded into your module. However, if the sample is a continuous or repeating sound, like a flute, you will need to determine where to place the looping section of the sample.

Determining the loop start and end points is usually a matter of careful listening and a lot of trial and error. Sometimes you will be able to look at the sample or a part of it and determine what part of the sample is repeating; often, however, you will just have to make a guess and play with the begin and end points until the sample sounds good. One thing to note when adjusting the loop points: since the sample right before the end point is immediately followed by the sample at the start point, those two samples should be very close if not equal in amplitude. Usually, this can be done by looking for "zero points" - places where the sample's waveform crosses the zero-axis - but any value will do. If the two samples are very different in amplitude, the result will be an audible "click" when the sample reaches the end of the loop. This is one easy way to tell whether your loop points are well-placed or not: if you hear a click while the sample is playing back which was not in the original sound, you need to move your loop points.


HTML Version of ABSM FAQ maintained by Mister X, © 1998/99.