Feedback would be awesome. The voice data is obtained via Flash for now.