Currently to add sound like dialogue or music to a level, you need to:
What a hassle. Instead, we'd like that built into the functionality of the level editor itself. Just a simple interface that records audio, uploads, does the conversions and filtering and cleaning, and calls it a day. This project would be a good starting point for how to get the audio with HTML5. The filtering and conversion might happen in the browser or on the server, but probably have the server at least do the conversion. So maybe clean up on the browser, ask the user to confirm, then send the clean wave file to the server to be processed from there.
For the client side, the interface should be handled by Treema. See the existing Treema nodes for examples.
For the server, this requires some research! I imagine it would be Node calling command line conversion programs to get the job done.
The clean but uncompressed wave files should go in the db as well as the converted files so we can do more conversions later if we want (higher and lower quality audio for example).
Great idea, however such a feature would require to manage copyright violation, don't you think ?
We've got a CLA that people agree to when they contribute which stipulates they must own the work and grant us permission to use it.