DIGITAL AUDIO EDITING: A basic "how to".
Digital audio editing is simply the best fun you can have alone. I should know, I've been at it for over 5 years now...
The principles, basically, are these:
You have sounds recorded onto your computer's hard-drive; in the Wintel-PC users case these'll be .wav files, if you're rich enough to own a Mac they'll be aiff format. (Recording said sounds is a whole different matter and can be covered in it's own node). These sounds should be at a sample-rate of 44.1 kHz, that's 44100 electrical samples per second, (x2 for stereo sounds) and the standard for CD recordings, or 96 kHz, which is the top quality DVD audio can be recorded at. It should also be at 16 or 24 bit "depth", which should ensure the sound has a fuller range of tones. These formats are used as the default import settings on most digital audio software today and it's important to make sure that all of the files you're going to use in any one project are set to the same bit depth / sample rate as each other and your software's settings.
You can also use 48 kHz / 16 bit settings for DAT quality, but I found this pointless since I was going to have to degrade to 44.1 kHz to reach any kind of audience anyway, and this process takes a very long time, even on today's super fast machines.
(I tend to use the CD quality myself, as not so many listeners buy audio DVDs yet, and it still sounds perfectly crisp to me!)
Next you need to edit your sounds in a sound editor, (something along the lines of Cool Edit Pro or Wavelab) so that you can make sure your audio contains no glitches that'll cause you problems later on. Never forget, just because you can't hear it doesn't mean it's not there; And there are a couple of little jobs you can do to all of your files to make sure you don't hear any surprise clicks and pops in your final mix.
Always, no matter what, normalize. This process boosts or cuts the amplitude of the wave to a level set by the user, most commonly 97%, to keep it from clipping, that is, distorting due to being louder than the output can handle, literally clipping the tips from the loudest waveforms. It does need to be as loud as possible also, since the quieter the file, the greater the background hiss when you play it back. The other thing normalization does is adjust the dc of the file to 0%. This basically centers the waveform and is another way of preventing distortion
Then you must turn your attention to the zero crossings...
If your sample doesn't start and end at a point where the wave intersects with the center line, (the point at which the wave is at it's quietest) you will often hear a pop or click. Even if you don't, you may well later on, once the wave has been compressed or had any kind of effect placed over it which effects it's amplitude.
All you need to do is: Zoom in on the first and last few micro-seconds of your file, select the area of wave from the first zero-point to the cut, and then apply a "fade" envelope (from 0% to 100%, or vice versa) to smooth the ends. Also, try to make sure that, if the wave begins going up, that it also ends going down, or the other way round, otherwise it might just glitch...
Cool Edit has a "find zero-crossings" feature, and most audio editing applications have similar and these are fine for the-editor-in-a-rush, but the only way to ensure accuracy is to do the job by hand.