Seems like I'll Justin Schoeman wrote: > Ori Pessach wrote: > > > > Justin Schoeman wrote: > > > > > Please don't knock it before trying it. > > > > Ok. I read the code, but I couldn't run it over the weekend. I'll try it in the next > > few days. > > > > > And by the way, all my high quality capture applications use > > > post-filtering+stretching to match the streams. This technique is only > > > supposed to be use for vcr-like applications, where it is more important > > > to get a once-off recording, than absolute quality (in which case you > > > wouldn't be using streamer anyway). > > > > The audio processing can be done in real time. I'll give it a try. > > It could be, but it would be very difficult - you would have to buffer a > large amount of data ahead of time, calculate the average audio drift > over this period, and then compensate the audio. If you make the > compensation period too small, you will have to make too many small > corrections, and it will not sound good. Agreed. > If you do get a capture engine written up with this core, let me know - > I would like to use it. Sure thing. Don't hold your breath, though... Seems like the right way to do this is the only way to go, after all. Oh, well... Ori
