|
Hi All, As I have previous sent, as part of my PhD I am conducting large scale subjective testing of audio that has been processed using different Time-Scale Modification algorithms, and I would appreciate your assistance. To facilitate this I have used the Web Audio Evaluation Tool to enable subjective testing through a browser, and am conducting the research under Griffith University Ethics Reference 2018/671.
A massive thank you to everyone who has been involved already. As of this morning almost 32000 opinion scores have been given heading towards the goal of 37000 opinion scores (7 ratings per file). This means that there are only 92 sets out of 532 sets to go.
The subjective testing can be found at
http://www.timrobertssound.com.au/TSM/index.html with each set taking approximately 10 minutes to complete. You are able to complete as many or as few sets as you prefer.
Once testing is complete a MOS labelled data set will be released (with a paper of course) containing the 88 source files and 5280 processed files. 10 playback rates ranging from 30% to 200% have been used with 6 TSM algorithms, Phase Vocoder, Identity Phase Locking Phase Vocoder, Harmonic Percussive TSM, WSOLA, Mel-Scale TSM and FESOLA). The data set contains Speech (M&F), singing, solo instruments, percussion, complex music (recorded and synthesised) and sound effect style sounds.
If you have any questions or suggestions please feel free to contact me. Also, please feel free to pass the link through to anyone who you believe may be interested in this research.
Regards, Tim Roberts PhD Candidate |