XMPEG vs. Gordian Knot Method...

[Quicksix]

I am looking for some specific details on the differences between the resizing/cropping methods used by XMPEG vs. those used by Gordian Knot (Nandub). I am interested in the methods used by the two. Obviously the Gordian Knot method generally produces better results, as all of the rips I have done with GK that I have compared to XMPEG have been superior. The XMPEG resize method (by using the output pad) will obviously allow you to create a video with the same frame size on screen as when you play an actual DVD (as you can select 16:9 for input res and then 1:1 for the pixel aspect). However in GK, using 16:9 input and then 1:1 pixel aspect will create a distorted picture. After adjusting the video size through cropping the black borders off, the final output resolution is lower than that which has been created through cropping. I am guessing this is because DVD pixels are not square, and GK is more accurate in maintaining aspect ratios, and will downsize to maintain them. Does XMPEG use some kind of scaling method which is not explicitly stated that would cause the difference?

Sorry for the long rant, but it wasn't easy explaining what I am after....

[khp]

I think the main difference is that Xmpeg will automatically correct the aspect ratio before letting u chop the picture.
With GKnot you chop the picture, before correcting the aspect ratio.

[Quicksix]

Thanks.....that seems to make sense. It could also be that XMPEG (being based on Flask) just uses lower quality processing/doesn't offer the same options as GK.....

[khp]

I don't think there is much of a quality issue, decoders are probably next to identical and they can both do biqubic resize.

But GKnot definetly has more options.