[MEncoder-users] new doom9 codec comparission (submission)

[Corey Hickey]

Doom9 Feedback Hotline wrote:
>>1. When will you let us know if we made it to round 2?
> 
> 
> Probably tonight when I'm done with the review, or tomorrow at the latest.
> 
>>2. If either codec makes it to round 2, are the cropping, scaling, and
>>bitrates for The Matrix Revolutions and Saving Private Ryan going to be
>>the same as in the 2004 comparison?
> 
> 
> Yes
> 
>>3. If either codec makes it to the final round, what will be the
>>cropping, scaling, and bitrate of Steamboy? I don't know if my copy is
>>anything like the E.U. one (and I'll have to adjust for PAL framerates)
>>but I'd like to give that a try.
> 
> 
> I put the size and resolution in the guidelines, 700 MB, 640x352. The MP3 
> audio is 113'729KB (I already had that from when I made the source 
> decision). I doubt he cropping parameters would do you any good considering 
> I have a PAL release (obviously I need to crop more). It will be just about 
> getting rid of the black borders, not some fancy (how much to cut into the 
> picture to have the lest possible aspect ratio error).
> 
> The thing is though, Friday is the deadline for the main round submissions, 
> so if you want to revise settings, it will have to be done by then. Since 
> this is the first time I'm doing the qualification, I didn't impose such 
> strict guidelines upon that phase, especially about the timeline, but 
> basically if you're in the qualification you have to assume you'll be 
> proceeding to the main round (and final round for that matter, SPR and 
> Steamboy only (anything else you probably should not submit in the first 
> place if you don't think your codec can handle itself against the top codecs 
> in the business)

I encoded The Matrix Revolutions last night. I was a little quick before
when I said there's not much difference between this and The Matrix --
the rain and lightning in the final fight scene are surely murderous to
encoding efficiency. Nevertheless, I couldn't find anything wrong and I
don't think there's any way I could get better results with different
parameters. If you don't hear from me before the deadline, go ahead and

As for Saving Private Ryan, I want to try mpeg_quant.

Steamboy will be more difficult for me. I've never done any parameter
comparison with encoding animated material, so I don't have much to
start with. At this point, if anybody feels like taking over the
Steamboy parameters from me, please speak up (I'd be happy if you did).
Otherwise, I'm going to start with a combination of some
non-controversial options and some options from Rémi Guyomarch's old tests:

mbd=2:mv0:trell:cbp:preme=2:predia=3:dia=3:precmp=1:cmp=1:subcmp=1:v4mv:qpel

-Corey