平面電視音訊up/downmix系統之DSP實現

Abstract

現在三維立體音效在多媒體界中有顯著的發展，我們可以在杜比、SRS等特效上發現三維立體音效的處理，我們也將提出我們的方法來達到三維立體音效的效果。首先在upmix把雙聲道的音源做處理成為5.1聲道，對這5.1聲道做一些處理讓聲音具有立體感，且可用5.1聲道的揚聲器來播放，我們使用:1.殘響器法2.適應性法3.杜比法4.適應性權重法5.主軸分析法 來變成5.1聲道，之後在downmix過程中再把這5.1聲道轉換成雙聲道由普通的兩件式揚聲器撥放，在這過程中使用頭部轉移函數來對環繞聲道做定位處理，另外，揚聲器到人耳之間會有著交越干涉的現象，我們利用電子補償的方法來消除此現象。我們使用DSP板來實現這些音效，且對upmix和downmix的方法做主觀性聽覺測試，選出最好的方法來達到三維立體音效的處理。

The implement is realized with a 3D sound audio effect. We can see 3D effect designed from Dolby, SRS. Now we summarize the concepts of these designs and propose our methods. First we upmix stereo channel inputs into 5.1 channels to produce some surround effects and bass effect. Then, downmix the 5.1 channels into stereo channel outputs with traditional speakers. In the stereo channel outputs, we can sound the surround and bass effects. In the upmix process, we use several methods such as Adaptive Structure, Dolby Surround Structure, Reverb Structure, LMS Weighting Structure and PCA Structure to produce different center and surround sounds; in the downmix process, we use HRTF (Head Related Transfer Function) to localize the center and surround sound and create the surround effects. The algorithms will be realized with the ADI DSP board which is particularly applied in Audio and Video processes. Finally we have subjective listening test of the methods of upmix and downmix and compare every sound and spatial qualities of each algorithms.