[XviD-devel] Two Encoders

[Christoph Lampert]

Hi,

I had a few bad ideas about SMP. What's your oppinion about this one?

Instead of higher speed (which is difficult because of data dependencies),
we could go for higher quality: 

Encode every frame in parallel with two instances of the encoder using
different options (e.g. with or without inter4v, adaptive quant,
H263/MPEG, for the branch it also could be GMC, etc.), and then choose the
"better" result (smaller bitstream, higher PSNR). 

After each frame we would have to sync the two encoders, i.e. copy
the information from the winner to the loser so both have the same
reference image etc. 

Do you have an idea how much work it would be to do so? Not too much, 
I guess, although I didn't really check.

Of course it would be better to have perfect decision heuristics when
inter4v or adaptive quant is benificial and when it's not. But at the
moment, I don't see any.


gruel