MP3 (MPEG-1 Audio Layer 3) is a digital audio encoding format. It greatly reduces the amount of data required to represent the audio recording and it uses a lossy compression algorithm. Even if the algorithm used is lossy in nature, the sound quality of the MP3 file is very good. A normal listener cannot differentiate between the MP3 file and the original uncompressed audio file.
The lossy compression algorithm used is very effective as it can be used to produce files that are 1/10th the size of the original uncompressed file. The compression achieved depends on the bitrate setting used for the compression. A mid-range bitrate of 128k bits will compress a file to 1/10th of its size. The quality of the compressed file also depends on the bitrate setting used for carrying out the compression process.
Perceptual Audio Coding is the underlying principle used for creating MP3 files. Under this, all the irrelevant and redundant parts of the sound signal are removed in order to reduce the size of the file. The parts of the signal that are inaudible to the human ear are removed. This reduces the size of the file by a considerable amount. The human ear can hear sounds in the range of 20 Hz to 20 KHz. This fact is used to decide the components of the sound signal that will be removed from the original uncompressed signal. Since these parts are already inaudible, removing them from the file doesn't make a difference to a normal listener. Therefore, the listener cannot differentiate between the uncompressed and the compressed file. The end result is a small file with good sound quality.
MP3 format is very popular as it's much easier to transfer an MP3 file than it is to transfer a file in WAV format. However, MP3 format suffers from the following limitations:
- The frequency resolution is limited and this decreases coding efficiency.
- There's no scale factor band for the frequencies above 15.5/15.8 KHz.
- Aliasing compensation stage is required for handling aliasing issues and this increases the amount of energy to be coded in the frequency domain. This results in a decrease in coding efficiency.
- The overall delay concerned with Encoder/Decoder is not defined.
