It's a nice idea, but it looks like you're just querying <a href="http://www.soundcat.ch/" rel="nofollow">http://www.soundcat.ch/</a> and then taking the file and playing it back, correct? I noticed in the Chrome developer console I can click a link right to download the MP3 of the song. For example I searched, "atlantic thrice" and it worked, but I could also get the song by going: <a href="http://www.soundcat.ch/download.php?id=163850" rel="nofollow">http://www.soundcat.ch/download.php?id=163850</a><p>I also noticed that playback stops and a it tries to load the file from a domain: mystuffsite.ch and I get a 404 error. Not sure what is going on there. I also think some improvements could be a play and stop button, the ability to create a simple playlist and perhaps save it using HTML5 localStorage.