[x264-devel] Re: Test results of trellis,brdo,and mixed_refs

[Tuukka Toivonen]

On Thu, 15 Dec 2005, Guillaume POIRIER wrote:

>> x264best2: baseline
>>"$OPTS:nomixed_refs:nobrdo:trellis=0:frameref=15:nob_adapt:b_bias=0:bframes=3:b_pyramid:ip_factor=1.0:pb_factor=1.5:direct_pred=1:noweight_b:i4x4:i8x8:b8x8mv:8x8mv:no4x4mv:8x8dct:me=3:me_range=24:subq=6:nochroma_me:chroma_qp_offset=8"

>Is there a reason why you added these specific options?

They gave around the best result in my tests.

> Is that to
>compare x264 VS reference H.264 code

No.

>, or did you notice that this
>setting looked better for your needs?

Yes.

> I've read several times that
>adaptive bframes,

Well, not for me. (but I ran my tests several months ago). Adaptive 
B-frames were quite bad, even, in average. (or maybe I had bad values
for example in maximum no. of B-frames, but i did try several).

> weighted bframes

Didn't have any significant difference for me so for simplicity I don't 
use.

> and pyramid were pretty much always
>improving quality.

Yes, pyramid indeed helps, it was enabled in these tests.

>As far as me_range is concerned, I've tested 2 encodes with uneven
>multi-hexagon search and me_range=16 (default) and me_range=64 (the
>maxme_range possible). From the top of my head, it slowed down
>encoding by a factor of 2 while adding just 0.01dB PSNR. Now, that was

Yeah, quite same for me: there was only small difference between
search range of 16 pixels and that of 24 pixels. But at least there was 
a clear difference, unlike between 24 and 32 pixels.

-- 
This is the x264-devel mailing-list
To unsubscribe, go to: http://developers.videolan.org/lists.html