Re[2]: [XviD-devel] Hadamard Transform

[Radoslaw 'sysKin' Czyz]

> Personally I dont see it the hadamard transform making much sense in software
> with the amount of mode decisions in MPEG-4. AFAICS the number of DCT transforms
> needed if you wanted to make the intra and 4v mode decisions in the DCT domain
> would only take a couple of percent of running time if those DCTs could all work
> from cache

DCT for mode decision? That sounds reasonable. But to make it faster,
I just have another idea - not-very-precise DCT (or hadamard transform)
which would work on 8-bit data and output the DCT block to different
buffer then source. Is it possible? I have no idea ;-)
But I think that the biggest slowdown would come from transfering data
from 8-bit picture to 16-bit buffer... I might be wrong, of course.
Just an idea.

Radek