@bp,

reading the document from Berkeley (listed above somewhere I think) the P frame is based on previous I and P frames and the B frame is bi-directional so is derived from pre and post frames. The I frame has the bitrate value and the P and B frame values are quality based.

so this would make it more challenging to figure out how to set the P and B frames. Hopefully if DVD2AVI can be extended to dump out the I, P and B frame information (i'm not looking in Kwag's direction at all:), there may be enough data in there to figure out the correct values.

The information is in there somewhere so we just got to figure out how to extract it.

Cheers,

I'm working on it 8)

-kwag

@Kwag,

Now that SansGrip is back, maybe he can shed some light on how
to pick-up bitrate with each frame via DVD2AVI for at least the I
frame. Have you given up on using Tmpgenc to improve motion
search, picture quality and file size. :?: It sounded promising :)

-bp

I never give up! I'm just taking a break 8)

-kwag

i finally got my hands on visual C++. So i had a look at dvd2avi using the debugger and sure the debugger hangs as some variables are not initialised correctly, but as usually when compiled it runs fine - don't you just love those debugging issues!

Cheers,

Specially Microsoft's :roll:

-kwag