Sound Processing
Sound is a wavelength which displays the vibration of air molecules. Technically speaking, the sound-source that produces sounds vibrates the air molecules creating vibration wavelengths which, in turn, are recognized by our ears. Consequently, sounds can not be heard in a vacuum. The sound we hear is usually described in analog wavelength vibration and the amplitude of vibration is periodically divided into a frequency. The frequency of vibration wavelength is decided on within a second of its occurrence.
When we identify the existence of the reciprocal relationship(interrelated relationship) between the elements of sound signals and our auditory senses, it provides us with an opportunity to discover some ways to deal with sound medium more efficiently.
1. Audible Frequency
The audible frequency band range of a sound signal for people is between 20 Hz and 20 KHz. This audible frequency band range differs among animals and a bat, for instance, can hear ultrasonic waves that exceeds our audible band range. We can successfully simplify the structure of a hardware by segregating the human band range from others when a hardware for creating sound signals or receiving external sound signals is being designed.
The original signal of sound is in analog, but it has to be converted into a digital signal as computers can only identify digital terms. This conversion is done by sound cards. In the inside of a sound card, there exists an ADC(Analog-to-Digital Converter) to record the sound in the computer and also, a DAC(Digital-to-Analog Converter) is installed in order to reverse the process.
2. Sampling Technique
As has been suggested earlier, the sound signal is an analog and it needs to be converted into a digital signal for it to be executed in the computer effectively. Therefore, the sound sampling that is executed in a sound card is done in digital data. A microphone receives sound signals, then converts them into analog electrical signals. This signal is finally turned into digital signals and this process is called a `Sampling Technique'. Sampling here means a process by which every single band range is expressed in digits by measuring the signal consecutively in certain short periods of time. The crucial point in the sampling technique is the frequency of the sampling and this is closely related to the precision that is required.
It has been maintained that, according to the general sampling theory, sampling requires at least twice the amount of band range which the analog signal possesses. For a higher quality product, four times the amount is usually demanded. In accordance with this perspective, when a sound that we can hear is being sampled, over 40 KHz is needed for converting analog signals to digital ones to reconstruct the original sounds.
The above-mentioned process of converting analog signals into digital information is often called `digitalization' or `encoding'.
It is appropriate here to examine the PCM(Pulse Code Modulation).
3. PCM
PCM, based on the theory of `Sampling', is one of the most widely used technique in `Digitalization' techniques. PCM technique involves the method that divides analog vibration wavelengths consecutively into short intervals and indicates each of them in a rectangular shape(bar-shape) to measure their heights for digitalization.
For digitalization, using the PCM method, the sampling rate and sampling size are the most important factors to be considered. Sampling rate indicates the intervals and is measured in Hz (eg. 11.025 Hz indicates 11,025 intervals in a second). This sampling generally should have at least twice the actual quantity of wavelength to express actual sounds.
The indication of the sampling size involves the quantity of the beat that has been indicated by bar-shape value of the sound. 8-Beat can classify the sound in 256 sizes individually and 16-Beat can do it in 65,536 sizes. The sound can be expressed more accurately with 16-Beat, but its demerit is that it requires a lot of storage space.
4 Number of Channel
When digitalizing the sound, the number of channels also need be considered. There are Mono(using only one channel) and Stereo(using two designated channels) channels. High quality digital CD sound is generally known as a recording done with a sampling rate of 44.1 KHz and 16-Beat stereo.
Digital sound files require a relatively larger storage capacity where the file size is directly proportional to sampling rate and number of beats used. The sound size of a standard digital CD requires approximately 5.3 MB. When a sampling rate of 11.025 KHz is used, the size is reduced to 1/4 whereas it gets reduced to 1/2 when 8-Beat is used. For this reason, most of sound files use 11.025 KHz (16-Beat, Stereo) or 22.02 KHz(8-Beat, Mono).
One's own hard-disk can have some capacity limitations as the capacity required to store 1 hour of stereo in CD standard is around 500 MB, so DAT(Digital Audio Tape) is usually needed.
The database that is widely used to store large quantity digital data is CD-DA.
CD-DA can be used for general music reproduction CDs as well as for multimedia database. However, marketed CD-DAs are generally used for a multimedia production as CD-DA production itself requires some technical facilities.
Top
Copyright (C) 2000~2003, iMusicSoft All Rights Reserved. webmaster@imusicsoft.com