A codec defines how audio (and video) is transported over a network and the quality of the voice. If the speech quality is too low it might make sense to try a different codec. A common reason for bad audio quality when using Voice over IP is insufficient bandwidth of the internet connection.
Generally, we differentiate between codecs which use compressed and uncompressed data. We also differentiate between lossless and lossy compression. A codec is always a compromise between used bandwidth, required CPU power for the compression of voice data and the overall speech quality. Wideband codecs make voice transmission in hi-fi Quality possible. Narrowband codecs allow voice transmission even with poor bandwidth. Consequently, speech quality drops. We recommend a bandwidth of 100 kBit/s in both directions for good quality.
Mean Opinion Score (MOS) is a measurement method for comparing codecs. A representative group evaluates how close speech quality of the codecs is to the human original. The scale is 1 (bad) to 5 (best). Everything better than 4 meets the speech quality of ISDN.
In the following table you can see how long it takes to translate one codec into another one and for which codecs the translation works in general. The unit of measure is microseconds.
H.261 is one codec for digital compression and decompression of video signals. Originally developed for video telephony via ISDN lines. H.261 was the first digital video standard codec. H.263 (from 1995) and H.264 (from 2001) are based on H.261. All standards are optimised for low data transfer rates. Constant data transfer rates are not defined. H.264, as the newest standard, can also handle high resolution images.
Askozia IP phone systems support a variety of free and licensed codecs.