In your case, a 3.06ghz GPU overclocked to 4.2ghz is right about a 30% clockspeed increase. Assuming your RAM is of sufficient bandwidth (it is, based on your sig), then a 1 hour render time would fall by APPROX. 30% That is to say, it would take 42 minutes instead of 60.
Just to be technically correct:
Increasing speed by 30% does not mean reducing encoding time by 30%.
^Using this, above logic - if you increase speed by 90%, your CPU is 1.9x faster, but gets the work done in 10% of the original time - it would have to be 10x faster to do that
you need to be careful with how you handle multipliers, as 1.3x faster is not equal to 0.7x time taken
1000 frames at 10fps takes 100 seconds
1000 frames at 13fps takes ~77 seconds, not 70
1000 frames at 15fps takes ~66 seconds, not 50
1000 frames at 18fps takes ~55 seconds, not 20
OC will cut down times, but won't revolutionize them. I'd honestly look towards encoding settings/methods as they can make much bigger dents.
As for GPU encoding, it's fast but not particularly efficient. If x264 at ~medium-veryfast preset is not good enough for you and you don't want to sacrifice resolution and you're already down to 30fps video, there's not too much to do i thinkEdited by Cyro999 - 3/19/14 at 5:35pm