An adaptive multi rate wideband speech codec with adaptive gain re-quantization
C. Erdmann, Peter Vary, Katrin Fischer, Joachim Stegmann, Catherine Quinquis, D. Massaloux, B. Kövesi
Abstract
C. Erdmann, Peter Vary, Katrin Fischer, Joachim Stegmann, Catherine Quinquis, D. Massaloux, B. Kövesi
Abstract
This paper describes an adaptive multi-rate wideband (AMR-WB) speech codec proposed for the GSM system and also for the evolving third generation (3G) mobile speech services. The coder is a multi rate SB-CELP (subband-code excited linear prediction) with five modes operating at bit rates from 24 kbit/s down to 9.1 kbit/s. Our basic approach consists of an unequal band-splitting of the input signal into two subbands (SB). A variable rate, multi-mode ACELP coder is applied to the lower subband (0-6 kHz). The various bit rates are integrated in a common structure where the scalability is realized by exchanging the fixed excitation codebooks while leaving all other codec parameters invariant. For the GSM related modes (9.1-17.8 kbit/s), the upper subband (6-7 kHz) is coded using a very low bit rate representation based on bandwidth expansion techniques. In case of the 3G application (24 kbit/s) the upper band is coded using a 4 kbit/s ADPCM coding scheme. In addition the analysis by synthesis (AbS) coder of the lower band employs a novel closed loop gain re-quantization technique controlled by the character of the speech signal. Thereby the codec achieves an enhanced performance for background noise while maintaining its clean speech quality.
OpenAlex reports 3 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper describes an adaptive multi-rate wideband (AMR-WB) speech codec proposed for the GSM system and also for the evolving third generation (3G) mobile speech services. The coder is a multi rate SB-CELP (subband-code excited linear prediction) with five modes operating at bit rates from 24 kbit/s down to 9.1 kbit/s. Our basic approach consists of an unequal band-splitting of the input signal into two subbands (SB). A variable rate, multi-mode ACELP coder is applied to the lower subband (0-6 kHz). The various bit rates are integrated in a common structure where the scalability is realized by exchanging the fixed excitation codebooks while leaving all other codec parameters invariant. For the GSM related modes (9.1-17.8 kbit/s), the upper subband (6-7 kHz) is coded using a very low bit rate representation based on bandwidth expansion techniques. In case of the 3G application (24 kbit/s) the upper band is coded using a 4 kbit/s ADPCM coding scheme. In addition the analysis by synthesis (AbS) coder of the lower band employs a novel closed loop gain re-quantization technique controlled by the character of the speech signal. Thereby the codec achieves an enhanced performance for background noise while maintaining its clean speech quality.
Key concepts: Full Rate, Adaptive Multi-Rate audio codec, Computer science, Speech coding, Harmonic Vector Excitation Coding, Codec2, Code-excited linear prediction, Codec