Different Types of Audio Compression

Audio compression helps in reducing the size of audio files as generic data compression algorithms are not advisable for reduction of audio files. There are two types of audio compressions used today,, lossless and lossy audio compressions.

When you speak of audio compressions, it refers to a type of data compression that reduces the size of audio files. It is not advisable to use generic data compression algorithms for reducing audio file sizes as it performs poorly with audio data. With generic data compression the file size is reduced not more than 87% of the original file size and is basically not designed for real time use.

In the case of audio compressions, the algorithms are found in the computer software as audio codecs. There are basically two types of audio compression used today; lossy and lossless comparisons where lossy algorithms is the most commonly used algorithm that provides better compression ratios which are used in mainstream consumer audio devices.

Information redundancy is reduced in both forms of algorithms through coding, linear prediction and pattern recognition to reduce the amount of information used in describing data. Lossless audio compression formats like Monkey’s Audio, FLAC and Shorten have grown in popularity because of less expensive file storage and communications bandwidth which is easily available. With lossless audio compression, it is possible to maintain a permanent archive of audio files.

Audio engineers, audiophiles and people who want preservation of exact copies of their audio files use lossless audio compression. This is in contrast to the irreversible changes that arise from lossy compression techniques like MP3 and Vorbis. High definition DVD formats are also being introduced with lossless formats like Dolby True HD.

It is basically difficult to maintain data in audio streams while achieving substantial compression because most sound recordings are complex and recorded from the real world. Even computer generated sounds have complicated waveforms that form a challenge to compression algorithms. This is mainly due to the fact that audio waveforms are difficult to simplify without any conversion to frequency information as picked by the human ear.

Audio data conversion is also difficult to perform as the value of audio samples change quickly making it impossible to use generic data compression algorithms for audio data conversion. In addition to this, strings of consecutive bytes don’t tend to appear often. However filter convolution slightly whitens the spectrum wherein traditional lossless compression found at the encoder helps the decoder restore the original signal with integration.

FLAC, TTA and Shorten all estimate the spectrum of signal with linear prediction. The estimator’s inverse is used for whitening the signal at the encoder by removing spectral peaks. However the estimator is used for reconstruction of the original signal found at the decoder.

Today you find lossy audio compression being used in numerous applications. Not only direct applications like computers and MP3 players use lossy audio compression, it is also used in digital television, most video DVDs, satellite and cable radio, streaming media on the internet and in terrestrial radio broadcasts. You find greater compression rates with lossy compression than lossless compression by the discarding of less-critical data.

Lossy audio compression uses psychoacoustics to prove that not all data found in audio stream can be picked up by the human auditory system. This is why perceptual redundancy is reduced by first identifying sounds that are hard to hear like high frequencies and sounds that occur together. These sounds are either coded with decreased accuracy or no coding at all.

With the removal of these ‘unhearable’ sounds only a small percentage of bits are saved in lossy compression. If this is not sufficient in achieving compression for an application, more lossy compression is required. However sometimes this too may not produce perceptible differences as speech is usually compressed more than music.

The bit rate is usually used to express the target rate of data that has to be achieved through lossy compression schemes. Depending on the bandwidth and storage requirements, lossy compression may lead to a reduction to the audio quality which ranges from none to severe which is usually unacceptable to listeners. With the removal of data in lossy compression which cannot be recovered by decompression, many people don’t prefer lossy compression for archival storage.
Source: http://www.flacsite.com

Niche Profit Sites by