基于建筑声学仿真模拟的视听一体化研究

Research on Audio-Visual Integration Based on Architectural Acoustic Simulation

文中选取莫扎特第四十交响乐多通道录音信号作为声源干信号,以天津文化中心音乐厅为例,通过建筑声学模拟软件ODEON进行声场模拟,将乐队按照乐器划分为49个点声源并分别对应各个乐器的指向性,在观众席设置5个典型接收点位置,分别计算每个声源点至接收点的双耳脉冲响应,再与每个声源点对应的乐器干信号进行卷积,将卷积得到的49个听音信号同步播放合成乐队多声源可听化信号。同时按照传统点声源的可听化方法,在交响乐队中心位置设置一个点声源,计算点声源至接收点的双耳脉冲响应,再与乐队49个干信号进行卷积同步播放得到交响乐单声源可听化信号。视觉方面首先在Sketch up中建立音乐厅的三维真实模型并给各个界面赋予真实材质,将Sketch up的视觉模型输出至VR虚拟现实模拟软件SimLab Composer,调整灯光、材质、环境等参数,对调整好信息参数的虚拟模型进行渲染以获得更好的沉浸感,通过HTC VIVE Cosmos眼镜输出VR虚拟场景,模拟厅堂的三维视觉。在建筑声学视听实验时分别对比多声源与单声源卷积的可听化信号分别在有无VR视觉信号的条件下混响感与ASW(感知声源宽度)这两个建筑声学空间感音质指标的主观评价差别。实验结果表明:采用多声源卷积的可听化信号可以显著提升听音的混响感和ASW;不同接收点位置的改善程度不同;由于单声源与多声源的可听化信号音质差异较为明显,实验过程中听音实验人员更多关注空间感音质的提升,加入VR视觉信号对可听化信号听音主观评价结果无显著提升。

In the article,the multi-channel recording signal of Mozart's Symphony No.40 was selected as the source dry signal,and the Tianjin Cultural Center concert hall was used as an example to simulate the sound field using the architectural acoustics simulation software ODEON.The symphony orchestra was divided into 49 point sound sources according to musical instrument,corresponding to the directivity of each musical instrument respectively.Five typical receiving point positions were set in the audience seat area,and the binaural impulse response from each sound source point to the receiving point was calculated respectively,then use binaural impulse response to convolve with the instrument signal corresponding to each sound source point,and finally the 49 resulting audio signals were synchronized and played together to synthesize the multi-source audible signal of the orchestra.At the same time,according to the traditional audibility method,set a point sound source in the center of the symphony orchestra,calculate the binaural impulse response from the point sound source to the receiving point,and then use binaural impulse response to convolve with the signal of the symphony composed of 49 dry signals and mix it into a single sound source audible signal.In terms of vision,firstly,the 3D model of the concert hall was established in sketch up and the real materials were given to each surface.The model of sketch up was output to the VR virtual reality simulation software SIMLAB composer,in which the parameters such as light,material and environment were adjusted.The virtual model was rendered and baked with adjusted information parameters to obtain a better sense of immersion.HTC VIVE Cosmos glasses were used to output VR virtual scenes and to simulate the threedimensional vision of the hall.During the architectural acoustics audio-visual experiment,the subjective evaluation differences of Reverberance and ASW(Apparent Source Width)of audible signals convoluted by multiple sound sources and single sound source were compared under the condition of with or without VR visual signal.The experimental results show that the audible signal convoluted by multiple sound sources can significantly improve the Reverberance and ASW,and the improvement degree of different receiving points is different.Due to the significant difference in sound quality between the single-source and multi-source audible signals,the listening test participants in the architectural acoustics and audiovisual experiment focused more on the improvement of spatial sound quality.The addition of VR visual signals did not significantly improve the subjective evaluation results of the audible signals.