Compressibility Test with Gordian Knot?

[cca]

Hello. One the most important features of Gordian Knot for me is the compressibility test. For DivX 3 & 5 GKnot can do it automaticaly. In case I use XviD, if I generate the proper .avs script with GKnot and the run a first pass with XviD manually, the resulting stats file loads in GKnot. My question is; is this method correct, or I'm getting result wich have nothing to do with reallity? I searched this forum prior to posting this, but it seems noone has tried it? Any replies are apprieciated.

[cjv]

Yup, that's right.

cjv

[cca]

So, I'm doing it right then. Anyway, the main reason that lead to this question was that the same compressibility test with DivX 5 is giving me worse result with XviD. And also changing the resizing method from e.g. sharp bicubic to bilinear has more effect to the DivX than to XviD. So, I thought maybe the stats from XviD is no good for GKnot's comp. test. Anyway, I'll do more testing with XviD.

[manono]

Hi-

I run the Compress test in GKnot and then test the first pass stats file afterwards as well. The percentages given match almost perfectly with the number given by Koepi's StatsReader when divided into 1. So, those percentages are accurate. And it also stands to reason that a DivX5 Compress Test on the same material doesn't matchup because of the greater comppressibility afforded by DivX5's B-Frames (although that will change soon).