Jeremiah Morris wrote:
On 26 Jan 2005, at 5:02 PM, Ed Wildgoose wrote:
Having synchronised in and out audio will be essential if we ever want to support accoustic echo cancellation.
That would be very cool. I'd also love to have dynamic volume control, based on ambient noise. With pause and commercial skip taken care of, now our biggest TV annoyance is having to crank the volume when we're making dinner and running the faucet, boiling water, and so on. The algorithm's a lot simpler too, so it could be a guinea pig for the audio input in case you run into snags on the echo cancellation. (I'm almost positive I'd be of no help there, though I'd be glad to pitch in on the other parts.)
This is fairly straightforward. Have a look at the mplayer code for an example. There is probably the algorithm already in ffmpeg?
Basic algorithm is just to look at RMS of some small history of audio samples. Change volume to bring this up to some preset level. The trick is to do something sensible when the volume jumps up again (loud explosion or something). Perhaps it's best to do soft clipping on the loud explosion and then slowly ramp down the volume...? Or perhaps it's better just to immediately cut the volume and build it up slowly again if needed...?
Ed W _______________________________________________ mythtv-dev mailing list [email protected] http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
