There's a good description of the newly released 'Selective' ratio mode in DVD2OneX 1.4 here:
http://www.dvd2one.com/forum/viewtopic.php?t=512

For the lazy ones, I'll quote directly:

Originally Posted by Rene Bakx
New version of DVD2one available!

Today we introduced our latest engine for DVD2one, The selective ratio engine.

This new engine is a different way of aproaching the video data!
Selectable in three pre-defined levels, this new engine determines the balance in distincting the important screen parts from the less important ones.


Things that are not in focus in the current frame, like backgrounds, are low in detail or as we define it, of low importance to the screen
and can be aproached differenty than the important ones.

With this new compression method, setting the slider to low will lead into a scheme where the important parts of the screen
are a bit more compressed than the less important ones. Compressing oflow important parts can result into macroblock because the original encoder
in a good encoder setup already achieved the best possible compression without loss of detail.

In medium the important parts are even more compressed than the non important parts, and in high the compression is almost done only on the important parts.

However in all three mode's the less important parts are still being compressed, because that is still mandatory to get the desired endsize.

By using this new method, the macroblocking (over compressed parts) will be less visible than with the conventional aproach.
If they occur they will be more eye friendly shaped, meaning they are not 100% square but a bit more dithered shape.
The human eye will notice them less in this non square shape and thus resulting in a more pleasantly viewable version of the picture.
He goes on to say that he usually gets the best results with 'low' and 'medium' selective ratio modes.

It would be nice if we could use this thread to post results of your experience with the different selective ratio modes...