New redistribution method using Final_Q in v1.26.5

[Insomniak4700]

I just tested the new "Use Final Q" feature for version 1.26.5 ..
I tried this on a episodic disc.

DVD-RB calculated the Q value from VTS1. If this is an episodic disc, shoudn't the Q value be calculated by taking into account all VTSes?? (I used the "Apply Redistribution using all VTSes" feature).
If only VTS1 is used, I know the Q value is good, but only for VTS1.

This particular disc uses one VTS per episode...
(Lost, season 2 NTSC R1)

[jdobbs]

No.

The Q value is used to put everything on an even playing field in terms of quality so the bitrate can be redistributed (based upon a constant qualilty) across the entire area to be "equalized". This is especially important on series discs. If you ran Q analysis against all the VTSs you may very likely get different Qs for each VTS -- which would be the exact opposite of the intent.

It may be possible to create a single AVS file that includes a sample across all the VTSs... but you'll find that on a series disc the deviation between episodes would be small... while the non-episode (extra) areas might move it off in one direction or the other. The best balance is to just use one VTS (that you know contains episodes) and use that as the constant quality baseline -- which is the reasoning behind why it is performed as it is.

[kumi]

The best balance is to just use one VTS (that you know contains episodes) and use that as the constant quality baseline -- which is the reasoning behind why it is performed as it is.
I think that's very sensible.

[Sharc]

A separate redistribution pass for the Extras for balancing the quality of the Extras can be very beneficial as the bitrate allocation of the Extras is often very careless. More critical even in the context of Steal Space from Extras.
Perhaps an additional redistribution mode in a future DVD-RB release..... ?

[Insomniak4700]

Thank you jdobbs... That's exactly what I wanted to know (my question probably wasen't very clear, but your answer was spot on) When I mentionned calculating Q for the entire disc, I somewhat had in mind an AVS sampling the whole thing, just like you mentionned, but you are probably right, the difference between episodes is probably next to nothing.