基于高阶Ambisonics的2.5维近场声源合成
发布时间：2018-09-04 19:11
【摘要】：针对空间声信号处理中二维声场合成与实际声源维数失配问题及近场声源合成中的近场失真问题,本文提出了一种2.5维声场合成方法和一种改进的维纳滤波的近场补偿方法.2.5维声场合成基于高阶Ambisonics的球谐函数展开.改进的维纳滤波方法用于补偿高阶Ambisonics合成2.5维近场声源时引入的近场失真.基于连续扬声器阵列假设理论,推导出了扬声器激励信号的一般表达形式.声场合成实验结果表明:与余弦正则化方法产生的相对误差0.23相比,所提方法合成2.5维近场声源的相对误差减少到了0.031;与高通滤波法相比,所提方法可以提供更大范围的有效听音区.
[Abstract]:Aiming at the problem of two-dimensional sound field synthesis and actual sound source dimension mismatch in spatial acoustic signal processing, and the near-field distortion problem in near-field sound source synthesis, In this paper, we propose a 2.5-D sound field synthesis method and an improved near-field compensation method for Wiener filter .2.5-D sound field synthesis based on the spherical harmonic function expansion of high-order Ambisonics. The improved Wiener filtering method is used to compensate the near-field distortion introduced in the synthesis of 2.5-D near-field sound sources by high-order Ambisonics. Based on the hypothesis theory of continuous loudspeaker array, the general expression of loudspeaker excitation signal is derived. The experimental results of sound field synthesis show that compared with the relative error of cosine regularization method 0.23, the relative error of the proposed method for the synthesis of 2.5 dimensional near-field sound sources is reduced to 0.031, and compared with the high-pass filtering method, the proposed method reduces the relative error of the proposed method to 0.031. The proposed method can provide a wider range of effective listening areas.
【作者单位】： 北京工业大学电子信息与控制工程学院;
【基金】：国家自然科学基金(No.61231015,No.61201197) 教育部博士点新教师基金(No.20121103120017) 北京市教委科技计划(No.KM201310005008)
【分类号】：TN912.3;TN643
本文编号：2223090
[Abstract]:Aiming at the problem of two-dimensional sound field synthesis and actual sound source dimension mismatch in spatial acoustic signal processing, and the near-field distortion problem in near-field sound source synthesis, In this paper, we propose a 2.5-D sound field synthesis method and an improved near-field compensation method for Wiener filter .2.5-D sound field synthesis based on the spherical harmonic function expansion of high-order Ambisonics. The improved Wiener filtering method is used to compensate the near-field distortion introduced in the synthesis of 2.5-D near-field sound sources by high-order Ambisonics. Based on the hypothesis theory of continuous loudspeaker array, the general expression of loudspeaker excitation signal is derived. The experimental results of sound field synthesis show that compared with the relative error of cosine regularization method 0.23, the relative error of the proposed method for the synthesis of 2.5 dimensional near-field sound sources is reduced to 0.031, and compared with the high-pass filtering method, the proposed method reduces the relative error of the proposed method to 0.031. The proposed method can provide a wider range of effective listening areas.
【作者单位】： 北京工业大学电子信息与控制工程学院;
【基金】：国家自然科学基金(No.61231015,No.61201197) 教育部博士点新教师基金(No.20121103120017) 北京市教委科技计划(No.KM201310005008)
【分类号】：TN912.3;TN643
【相似文献】
相关硕士学位论文 前1条
1 查梦芳;基于Ambisonics的多区域声场重放方法研究[D];北京工业大学;2016年
，本文编号：2223090
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2223090.html
