Thanks for the reply.
All methods of channel zapping are about the same on my system (keyboard, OSB, remote). I can definitely rule out the input devices or overlay issues.
I traced this problem a little while back, and about 4 seconds were spent in the channel change itself, and the rest was spent rebuilding the graph. I will post the log output tonight when I get home.
Otherwise, I have compiled and traced this issue before from the svn source code (I am software engineer), so if you have any pointers I am happy to try and contribute.
Regards. |