Music Acoustics Research

One project I worked on involved using a laser vibrometer to measure the vibrations of a violin top in an attempt to visualize the mode shapes and their resonant frequencies. An impact hammer was used to excite the bridge, and the vibrometer was pointed at different places on the surface of the violin to measure its oscillations. A frequency response function was calculated, dividing the vibrometer frequency spectrum by the impact hammer's frequency spectrum, to determine where the violin top was most resonant at different frequencies. The visualization below takes measurements averaged over 5 trials at each position, mapping colors to frequency response energy at different points on the top.

An earlier project attempted to extract frequency information from high speed video footage of a violin top after the bridge is struck by an impact hammer. I used image processing to get pixel values along one vector of every frame of a high speed video clip. The user then selects an edge that they would like to track along that vector over the length of the video clip, and the software calculates which number pixel along the chosen vector crosses the selected threshold, interpolating position to the 100th of a pixel. This pixel position over time is then analyzed with a Fast Fourier Transform to determine what frequencies it was oscillating at. Image processing with this methodology did not prove effective in determining the frequency of violin top vibrations as the small amplitude of oscillation at frequencies in in the range of interest proved difficult to repeatably capture.