If we use Google Translate Audio output and process it so that its flat ambient speech, for instance in this case Market social-rumble and background crowd noise, as we produce the text and create the output audio file itself and because the Google Translate System is free, who controls copyright if there is any even though we provide the text for speech, the cript, the output is ours by content creation, the output is Google's by format file output creation and so on.

I'm going to use it anyway so if Google sue me for millions I'd like to see where they'll find them and I'll use the remaining ones. :P :D

But I'll be using about three vocal files created because I want some Chinese Background Hubub and I thought the best way to do it was to find a translation system that has a very nice text to speech audio-out and Google Translate if you've ever used it is wonderful.  All you need to do is feed it your text or Pinyin as I have below and off it goes.

You will need a morpher to change female to male afterwards if needs be. But the copyrights I couldn't find.  

Click the image to hear the wave file. Its in wav format.