Abstract:
PURPOSE: A noise filling method, an audio decoding method, recording medium, and a multimedia device having the same are provided to generate a noise signal without additional noise information on an encoding unit by using a Norm value transmitted from an encoding unit for noise filling. CONSTITUTION: A decoding unit performs a spectrum decoding process of bit stream(1810). The decoding unit performs envelope shaping of a spectrum by using an encoded Norm value included from each sub band unit(1830). The decoding unit generates a noise signal about a sub band including a spectrum hole(1850). The decoding unit shapes the sub band which the generated noise signal fills(1870). [Reference numerals] (1810) A decoding unit performs a spectrum decoding process of bit stream; (1830) The decoding unit performs envelope shaping of a spectrum by using an encoded Norm value included from each sub band unit; (1850) The decoding unit generates a noise signal about a sub band including a spectrum hole; (1870) The decoding unit shapes the sub band which the generated noise signal fills; (AA) Start; (BB) Return
Abstract:
오디오신호의부호화/복호화방법및 장치가제공된다. 오디오신호의부호화는중요주파수성분에대응하는잔차신호를선택적으로부호화할수 있다. 오디오신호의부호화는 PNS기술을최종합성신호에적용함으로써추가된노이즈의특성을유지할수 있다. 오디오신호의부호화/복호화는 attack이강한신호에서 pre-echo를최소화시키며, 주파수도메인에서자기상관도가높은프레임에대해서부호화효율을향상시킬수 있으며, 합성필터를거친후에발생할수 있는 spectral hole 부분에적절한노이즈를추가함으로써자연스러운합성음을생성할수 있다.
Abstract:
무손실 부호화/복호화 장치 및 방법이 제공된다. 무손실 부호화 장치는 메모리로부터 콘텍스트 그룹(context group)들 각각에 대응하는 확률 모델 (probability model)을 읽을 수 있다. 이 때 메모리에 저장된 확률 모델은 콘텍스트가 그룹핑되어 생성된 것일 수 있다. 무손실 부호화 장치는 확률 모델을 이용하여 심볼(symbol)을 부호화하고 비트 스트림을 생성할 수 있다. 무손실 부호화 장치는 부호화 효율(coding efficiency)을 높이고 메모리의 필요 공간을 줄일 수 있다. lossless coding, 무손실 부호화, context, context group, probability model
Abstract:
PURPOSE: A bit allocation method, an audio encoding/decoding method, an apparatus thereof, and a multimedia device thereof are provided to calculate the number of maximum allocation bits in a sub band by using knowledge modeling and to restrict the number of maximum allocation bits. CONSTITUTION: A Norm estimating unit(210) obtains a Norm value corresponding to average spectrum energy. A Norm encoding unit(230) performs lossless encoding and quantizing of the Norm value. A bit estimating and allocating unit(250) performs estimation and allocation of the number of necessary bits by using the Norm value. [Reference numerals] (210) Norm estimating unit; (230) Norm encoding unit; (250) Bit estimating and allocating unit; (AA) From conversion unit; (BB) To encoding unit; (CC) To multiplexing unit
Abstract:
PURPOSE: A device for encoding/decoding spatial parameters and a method thereof are provided to eliminate ineffective parameter information by using correlations between spatial parameters. CONSTITUTION: An effective range determining unit(302) determines an effective range of spatial parameters by using correlations between the spatial parameters showing the inter channel feature relation of a multi channel audio signal. A parameter quantization unit(303) quantizes the spatial parameters based on the effective range. A parameter encoding unit(304) encodes quantized spatial parameters.
Abstract:
멀티 채널 신호의 부호화/복호화 장치가 개시된다. 멀티 채널 신호의 부호화 장치는 멀티 채널 신호의 특성을 고려하여 멀티 채널 신호의 위상을 변형시킨다. 위상이 변형된 멀티 채널 신호에서 추출된 다운믹스 신호와 잔여 신호를 이용하여 멀티 채널 신호에 대한 부호화된 비트스트림을 생성한다.
Abstract:
PURPOSE: A lossless encoding and decoding apparatus and a method thereof are provided to increase encoding efficiency. CONSTITUTION: An extracting unit(110) extracts a context of the first symbol among consecutive symbols. A decision unit(120) decides a group in which the extracted context is included among a plurality of context groups. A memory(130) stores a probability model corresponding to the plurality of context groups. An encoder(140) encodes the first symbol based on the probability model and the decided group. The encoder generates a bit stream.