[QUOTE=terry5;100425]Hi diddleydoo,
Firstly, 2-pass does not support CUDA encoding;
Secondly, 2-pass will calculate twice, which costs more time(as you said twice the amount of time as the one-pass) and provides high quality. What's more, you can see that, accurate account of the Encoding Method in Conversion Settings page, "Fast encoding(1-pass)" and "High quality encoding (2-pass)".
Wish it helps.[/QUOTE]
Well I'm aware that 2-pass will inherently take longer than 1-pass.  Quality and analysis doesn't come for free.
The point I was trying to get at was "2-pass does not support CUDA encoding".  This was my suspicion but I had a hard time finding this out officially.
If this is so, may I ask why?  Surely the encoding itself is similar (the second pass) to a 1-pass operation?  What are the differences that CUDA can't be used for 2-pass but can for 1-pass.  I have wasted my money on a nVidia GPUI in this case.