[MEncoder-users] Backup policies
Peter Cordes
peter at cordes.ca
Wed Nov 28 07:36:17 CET 2007
On Mon, Nov 26, 2007 at 05:28:55AM -0700, Loren Merritt wrote:
> On Sat, 24 Nov 2007, Peter Cordes wrote:
>
> > OTOH, some guides I've seen say that some options don't really work with
> > crf, e.g. trellis=1. (Also, some say that trellis and dct_decimate don't
> > work well together, so I usually use trellis=1:no_dct_decimate).
>
> They're just wrong. If they're wikis, please provide a link so that I can
> fix them. If they're not wikis, please provide a link so that I can flame
> the authors.
Here's one: http://www.avidemux.org/admWiki/index.php?title=X264
> I'm not sure how the rumor started, unless it has something to do with the
> previous rumor that xvid's trellis is harmful with CQ. Which is both wrong
> in that it was due to incorrect measurements the benefits of trellis, and
> wrong in that it assumes x264 is the same.
One of them actually has a justification that sounded reasonable to me (not
knowing exactly how x264 works, but having some idea of video compression in
general):
dct_decimate throws away some detail and it's better to let trellis make
all the decisions about what to throw away.
I can't find that doc with google, even though I think I did stumble across
it again recently.
So you're saying that the dct blocks dct_decimate throws away would not
always get thrown away by trellis when it would be Better to do so. (for
some definition of better, such as PSNR-optimal.)
Err, I've also seen it said in several places that trellis is not a good
idea with crf rate control. e.g. in this doc from june 2006:
http://www.digital-digest.com/articles/x264_options_page6.html
Is this true now, if it ever was?
I've seen akupenguin (x264 lead author) say that crf is close the the same
quality as 2pass, and that would be more likely if there aren't options that
are only good with 2pass.
Thanks for setting me straight.
--
#define X(x,y) x##y
Peter Cordes ; e-mail: X(peter at cor , des.ca)
"The gods confound the man who first found out how to distinguish the hours!
Confound him, too, who in this place set up a sundial, to cut and hack
my day so wretchedly into small pieces!" -- Plautus, 200 BC
More information about the MEncoder-users
mailing list