Soft codec HE-AAC

The audio signal compression standard AAC was developed and published in the context of the ISO/IEC 13818-3 (MPEG-2) and ISO/IEC 14496-3 (MPEG-4) standards. The AAC standard was positioned as an effective replacement for the widely spread audio processing standard MP3 developed in the context of ISO/IEC 11172-3 (MPEG-1 Audio Layer III) standard. In comparison with MP3, the AAC standard has a great number of advantages. The most important of them are the following:

•          Higher frequency resolution when audio signal spectral processing – AAC has 1024 spectral features compared to 576 for MP3;

•          ААС has a feature to use backward prediction that secures higher efficiency of one tone audio signal coding (for example, sound of camertone);

•          Stereo signal coding is more flexible in comparison with the analogue algorithms of MP3, this allows more adaptive changing of output stream rate of an audio coder;

•          The level of Haffman scrambling is enhanced – when coding, four-bit bytes of spectral features are often used, moreover, choice of Haffman tables on the coder side is realized more flexibly;

•          Instead of cascading (hybrid) filter bank, as it is made in MP3, AAC uses standard MDCT (Modified Discrete Cosine Transformation) with 5.3 ms impulse response and 48 KHz sampling frequency, which essentially reduces the impact of unwanted artifacts of preecho;

•          Algorithms of comfort noise forming are added, which is especially helpful for transmission of speech on low streams.

Today the AAC standard is unconditional leader in performance of audio information transmission, including musical, starting with bit rate of 24 KBit/s for stereo signal.

Soft sound codec AAC implements ISO/IEC 14496-3 (HE-AAC) standard.

In the current version of the codec HE-AAC version is implemented, which includes the algorithm of spectral reflection of a signal band, so-called SBR, allowing attaining of audio CD quality already with 48 KBit/s bit rate – FM radio broadcast performance.

In addition to the above, the codec allows, at the user’s option, tagging output stream rate of audio coder, being constant or adaptive to the dynamic of the audio signal, which is more flexible for systems engineering that include sound channel. 

In comparison with more widespread at present time sound codec MP3, implemented soft codec HE-AAC gives up to 70% advantage in stream rate with equal quality of regenerated signal on receiving side.