The use of sound in multimedia can be extremely powerful.  When done correctly, it adds a spectacular dimension to any presentation.  However, if done incorrectly, sound can ruin a presentation.  Thus, much study has taken place in the area of multimedia sound recording. 

Perhaps the most common representation of sound in multimedia is digital audio.  Any sound can be digitized to transform it from its analog nature to discrete samples.  These sound samples are stored as digital information in bits and bytes.  Samples of a sound are taken in equal time intervals, called the sampling rate.  The sample size indicates the amount of information captured concerning each sample.  A collection of samples played out over time make up a sound segment that can be used in a multimedia project.  Once all samples have been recorded, you can use software to edit the sound segment until it possesses the characteristics you desire.

As technology advances, users are forced to make decisions concerning sound quality versus file size.  Generally, the better the quality, the larger the file size.  This phenomenon is a result of the sound being recorded either at a higher sampling rate or with larger sample sizes.

Sounds recorded at higher sampling rates produce clearer, brighter, and seemingly louder sounds than those with lower sampling rates.  Recordings at higher rates capture more details of the original analog signal, such as the attack of the word and the speaker's breath.  Thus,  the sounds seem more live and less recorded.

Another sound quality element is the sample size.  A sample size usually consists of 8, 16, or 32 bits.  Recordings with larger sample sizes are less grainy than those with smaller sample sizes.  Thus, larger sample sizes result in better sound quality.

The ideal sampling rate and sample size often depend on what is being recorded.  For example, if the rate at which a musical piece is recorded is too low, many of the instruments may be muffled or perhaps not heard at all.  However, a single voice recorded at that same rate may be clearly audible.  Generally, the more dimensions there are to a sound, the higher the sampling rate and sample size required.  The recording must be such that all of the critical aspects of the sound are captured.  Otherwise, the effect produced will not accurately represent the original sound.