以像源法與BTM模型探討長方形房間中隔板介入的聲學影響

Abstract

生活中的交談環境中的隔板影響聲音訊息的傳遞，不同聆聽者、對談者位置以及環境空間組成導致接收聲音的變化，若要設計或改善交談環境的聲音傳遞品質，分析並評估隔板介入造成的聲音變化為不可或缺的一環。本研究主要有三個目的：發展一套數值模型計算隔板介入前後的室內空間脈衝響應（RIR）、以實驗驗證數值模型在不同複雜空間的適用性、探討特定情境中隔板介入的聲學影響。本研究模型採用幾何聲學的觀點，將波動問題轉換成幾何問題來計算所有可能的聲音路徑，再透過像源法與Biot-Tolstoy-Medwin（BTM）繞射模型來計算特定路徑的響應，最後使用混合訊號電路以對數正弦啁啾法（logarithmic chirp technique）量測真實的無響、半無響、長方形空間的RIR比對模擬的結果，藉此驗證模型的適用性。本研究發現聆聽者位置的清晰度指數C50在隔板介入後下降1.7~ 6.7 dB，發話者位置的$C_{50}$則沒有明顯的變化（提升0.00~ 0.15 dB）。早期響應能量損失與清晰度指數下降，主要是因為隔板的介入使得直接音或部分早期反射改為繞射抵達，本研究更發現面對面坐相較於相鄰而坐的情況損失多 4 dB，且若桌子靠近牆邊，損失則會額外再多 2 dB。我們預期本研究的模型能預先在設計規劃空間階段評估最佳的聆聽位置、分析交談環境的響應，最終改善交談環境使人們的對談更加清晰。

In conversational environments, partitions would impede sound transmission. The resultant sound variation depends both on the room shape and the set of locations listeners and speakers. To improve the quality of the sound transmission in such environments, it is essential to analyze and evaluate the sound changes caused by partition insertions. This study has three main objectives: developing a numerical model to calculate the room impulse responses (RIR) before and after partition interventions, validating the applicability of the numerical model through experiments in rooms of different complexity, and examining the acoustic impact of partition insertions in specific scenarios. The proposed model adopts a geometrical acoustics approach, converting the wave propagation problem into a geometry problem to evaluate all possible sound paths. The responses of specific paths are then calculated from the image source method and the Biot-Tolstoy-Medwin (BTM) diffraction model. The measured RIR of anechoic, semi-anechoic, and rectangular spaces using the logarithmic chirp technique are compared with the simulated results to validate the model's applicability. The study finds that the clarity index C50 for the listener position decreases by 1.7 to 6.7 dB after partition insertions, while there is no significant change in the clarity index for the speaker position (increasing by 0.00 to 0.15 dB). The loss of early response energy and the decrease in clarity index are primarily attributed to the detour direct sound or some early reflections. Furthermore, our study reveals that sitting face to face results in a 4 dB higher loss compared to sitting adjacent to each other, and if the table is placed near the wall, an additional 2 dB loss is observed. It is expected that the model developed in this study can assess optimal listening positions and analyze the room response during the spatial design phase, ultimately improving the conversational environment and enhancing clarity in communication.