Have you considered adding audio material for "immersion" translation? This would be, I imagine, beneficial to both duolingo and those wishing to improve their aural language skills.
I don't know the copyright status of current event broadcasts (i.e., audio recordings from Radio France International), but if available would be extremely relevant material and a chance to translate the spoken language.
I think the problem is less the copyright and more the fact that anyone can translate anything, and you have the serious problem of dealing with trolls. (There's enough people on the boards posting things like duolingo is awesome and then wirting nothing. Or things like this is hard what's going on.)
I'd like the implementation of it, but there's the problem of policing it.
Thanks for the reply. I'm not certain I understand your objection - I had in mind that it would work like the "immersion" section, only for transcription of audio clips. The translation would be regulated by the community, just as it is for the written.
I'm personally interested because my weakest skill is aural comprehension. The aural comprehension in the lessons is a good start, but could be extended to "real life" exercise.
I'll try to explain it better, then, using random names.
Basically, say Ana uploads a clip of herself reading the first paragraph of War and Peace in Russian.
Brian would then come along and translate that into English, uploading a clip of himself saying it in English.
Then Carrie comes along and edits Brian's clip to fix up a few word choices.
And lastly, Dan shows up, and being a new user, or a troll, or just in a bad mood, replaces the whole thing with a clip of inappropriate language, or an advertisement, etc.
Then suppose Emma, who's in middle school and studying Duolingo with the rest of the class, ends up on this aural translation including off-limits words. Her teacher hears about it, the Duolingo program is shut down.
I suppose it could be on a tier system—once you reach tier 3+ or so, you can access aural translations. That would avoid most of the problem, since by tier 3 or so people are already invested.
There's also the problem of how to give points for it—it could be by seconds, but that doesn't always guarantee anything. Some people talk 400 words a minute, others talk 100 words a minute.
There's also the problem dealing with microphones and the various buzzes you hear when recording.
A simple idea would be to take /dictation/. You have a clip, you take dictation from it and practice Russian (as in the example above.) Then it's moved to regular immersion (or a special immersion section) where you can translate it, and finally it's taken to a place where people can practice speaking and will receive upvotes.
This would not only transcribe and possible subtitle videos, but it would also ensure people can take dictation (if you were learning Russian, you'd take dictation and that would be crowd-checked), can translate (utilizing what is already there), and then can practice speaking.