K. P. H. Thai, P. Gournay, R. Lefebvre, and S. Charlebois, “Reconstruction of Mechanically Recorded Audio Signals Using White-Light Interferometry,” in Proc. AES Convention 138, May 2015, Paper 9300. [Online]. Available: https://aes.org/publications/elibrary-page/?id=17724
Thai KPH, Gournay P, Lefebvre R, Charlebois S. Reconstruction of Mechanically Recorded Audio Signals Using White-Light Interferometry. In: AES Convention 138. Audio Engineering Society; 2015. Paper 9300. Available from: https://aes.org/publications/elibrary-page/?id=17724
@inproceedings{Thai2015_17724,
author = {Thai, Khac Phuc Hung and Gournay, Philippe and Lefebvre, Roch and Charlebois, Serge},
title = {{Reconstruction of Mechanically Recorded Audio Signals Using White-Light Interferometry}},
booktitle = {AES Convention 138},
note = {Paper 9300},
year = {2015},
month = may,
publisher = {Audio Engineering Society},
url = {https://aes.org/publications/elibrary-page/?id=17724}
}
TY - CPAPER
TI - Reconstruction of Mechanically Recorded Audio Signals Using White-Light Interferometry
AU - Thai, Khac Phuc Hung
AU - Gournay, Philippe
AU - Lefebvre, Roch
AU - Charlebois, Serge
T2 - AES Convention 138
M1 - Paper 9300
PY - 2015
DA - 2015/05/06
UR - https://aes.org/publications/elibrary-page/?id=17724
PB - Audio Engineering Society
LA - en
AB - This paper presents a method to reconstruct a digital audio signal from a physical and analog sound-recording medium such as the Edison cylinder. A non-contact 3D optical profilometer based on white-light interferometry provides topographic information about small overlapping sections of the recording medium. For each of these sections, the natural curvature of the medium is compensated, grooves on the surface are detected, and short depth trajectories representing the audio signal are extracted. These trajectories are then concatenated to produce a digital audio signal that is post-processed to simulate the distinctive frequency response of an actual mechanical player. The effectiveness of this method is demonstrated on both tonal and vocal recordings using time, frequency, and time-frequency features as well as informal listening.
ER -