Types of audio codecs:
-DST (Direct Stream Transfer)
-FLAC (Free Lossless Audio Codec)
-LA (Lossless Audio)
-LPAC (Lossless Predictive Audio Codec)
-LTAC (Lossless Transform Audio Codec)
-MLP (Meridial Lossless Packing)
-Monkey’s Audio (APE)
There is a huge amount of audio formats. The most common are formats such as MP3 (MPEG-2 Audio Layer III) and WAV. Usually, the type of format corresponds to the file extension (the letters of the file name after the period, for example .mp3, .wav, .ogg, .wma).
A codec is an algorithm for encoding and compressing data in an audio format. Some file types are assigned a specific codec. For example, the MP3 format always uses the MPEG Layer-3 codec, while the MP4 format can use a range of different codecs.
Many times, the notions of codec and format are used as interchangeable. Especially when a format always uses a single codec. However, it is necessary to understand the difference between a format and a codec. In simple terms, a format can be compared to a container in which a sound or a video signal that uses a particular codec can be stored.
Some formats, such as MP4 or FLV, can store both audio and video sequences.
In the general scope of codecs (for any type of data), we can classify them as follows, depending on whether the original signal can be recovered or not after coding:
With losses (lossy). In this type of codecs, after coding, it is impossible to recover the original signal. Most codecs manage to reduce the size of the bit stream to be transmitted or stored, due to the loss of information in said bit stream. Normally this loss does not produce a large decrease in the quality of the audio perceived by the end user, and if the decrease in quality is appreciable, it is that a lot of information has had to be lost to achieve a small bit stream size, that in many occasions it is necessary, especially in the transmission of audio at a distance (telephony, digital video, television …), although this is a compromise solution between the different codecs, an issue that we will discuss in the comparative section between codecs .
Lossless (loseless). In this type of codecs, after coding, the original signal can be recovered. These types of codecs are the least common. They are usually common especially in high quality audio applications, where the size of the bit stream or stream is not decisive. If the files are to be treated later, it is not advisable to perform loss coding, since one encoding with losses after another would significantly damage the audio quality.
There is another classification of codecs, depending on the type of algorithm used in the coding:
Waveform codecs:
used for all types of digital signals. The waveform of the encoded signal must be as similar to that of the original signal
Vocoders or source codecs: used only for coding voice signals. The original signal is analyzed and synthesized to give rise to the encoded signal
Hybrids: combine characteristics of the two previous types
The waveform codecs seek to produce a reconstructed signal of the signal to be encoded, whose waveform is as similar to that of the signal to be encoded. These codecs work without knowing how the signal to be encoded was generated, which implies that in theory its operation does not depend on the signal and can work well with all types of signals, even if they are not audio.
Hybrid Codecs
These types of codecs are a mix between waveform and source. Within the hybrid codecs, the most used are the codecs in the time domain of Analysis-by-SÃntesi
