说话人位置不可知性：基于距离的幅度平移算法的改进
Speaker Placement Agnosticism: Improving the Distance-based Amplitude Panning Algorithm
摘要：
洛修斯等al介绍了基于距离的幅度平移算法（DBAP），以实现艺术和科学背景下扬声器放置的灵活性。该算法允许2D平面中的任意扬声器位置，使得虚拟声源可以导航2D空间。每个扬声器的增益计算为源到每个扬声器的距离的函数，从而创建声场。这给听众的印象是一个源移动通过扬声器的领域。本文介绍了一个开发的DBAP强大的变化，纠正错误的假设，在实施Lossius。具体而言，本文开发了一种方法，用于扬声器领域外的声源，其中Lossius版本产生失真的听觉印象和各种扬声器的增益中的空间不连续性引起的剧烈波动的幅度。在平滑虚拟源的空间印象中，我们还能够完全消除凸包的计算，这是原始实现的必要组成部分。这大大简化和减少了二维或三维空间所需的计算。
英文摘要：
Lossius et. al introduced the distance-based amplitude panning algorithm, or DBAP, to enable flexibility of loudspeaker placement in artistic and scientific contexts. The algorithm allows for arbitrary loudspeaker locations in a 2D plane so that a virtual sound source may navigate the 2D space. The gains for each speaker are calculated as a function of the source's distance to each loudspeaker, thus creating a sound field. This gives the listener the impression of a source moving through the field of loudspeakers. This paper introduces a heuristically developed robust variation of DBAP that corrects for faulty assumptions in the implementation of Lossius. Specifically, this paper develops a method for working with sound sources outside the field of loudspeakers in which the Lossius version produces distorted aural impressions and wildly undulating amplitudes caused by spatial discontinuities in the gains of the various loudspeakers. In smoothing the spatial impression of the virtual source, we are also able to eliminate the calculation of the convex hull entirely, a necessary component of the original implementation. This significantly simplifies and reduces the calculations required for any space in either two or three dimensions.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
