b-vops on single pass are optimally placed?

[PiXuS]

Section C6. of Crusty's FAQ make it sound like single pass is really only useful for fixed-quant or real-time encodes. I agree to that.

Though, I searched his FAQ and this forum and couldn't figure out if the b-vops or b-frames are optimally choosen on a single-pass.

Right now, I am doing an encode with a target quant of 4. B-frames are enabled (2/1.50/1.00). I see them being distributed here and there (via the "XviD Status" window). When they are choosen, is it because the encoder figured that that frame was compressible enough to be encoded by a b-frame?

Asked differently and more succinctly, is it safe to use b-frames on single-pass encodes?

Thanks.

[Manao]

As safe as with two passes, since b-frames are chosen during the first pass.

[PiXuS]

Okay, so on the first pass, position of I, P and B frames are chosen. Then, on the second pass, the position remain the same, but the quant of each frame might change.

That's it?

Thanks!!

[sysKin]

In theory, P/B decision alghorithm is "tuned" to work in quant 2 (ie with high-quality frames). So in theory, if you're doing single pass at quant 12, there might be less b-frames than optimum.

I wouldn't worry about that at all, especially since P/B decision is mostly a random combination of random factors that somehow appear to work ;)

Radek

[Koepi]

You mean like in.... the frame type decision code may contain nuts? :p

[PiXuS]

Originally posted by sysKin
I wouldn't worry about that at all, especially since P/B decision is mostly a random combination of random factors that somehow appear to work ;)

Me crosses fingers in hope you are just being humble. :)

[lordadmira]

I'm allergic to nuts. No, seriously, I'm allergic to them. They can ruin my day.


_