[x264-devel] Re: no gain in 2PASS for very low bitrates

[shauli rozen]

>b_stat_read=1 only on the 2nd pass, right?
yes
>What resolution? At 40kbps, it better be <= QCIF.
it is a qcif movie indeed
>How many frames? 2pass doesn't do much on short videos, other than giving
>more accurate bitrate. It only matters when there's many scenes of varying
>complexity.
the movie is 300 frames long , maybe that is not big enough for the 2PASS
to take effect?




On 12/21/05, Loren Merritt <lorenm at u.washington.edu> wrote:
>
> On Wed, 21 Dec 2005, shauli rozen wrote:
>
> > Hi All.
> >
> > I am trying to encode the a test vector with some movement to low bit
> > rate (25 fps at 40kbps ).
> > I tried using both abr and 2pass , and even though 2pass gives me a
> > more accurate bit rate results
> > I don't see much gain in quality over the abr.
> >
> > I am using the two pass by setting both b_stat_write and b_stat_read to
> 1,
> > and then I run the encoder twice.
> >
> > Is it true that there is no much gain in using 2pass for really low bit
> rate
> > ? I expected much better results in the 2pass.
> >
> > Maybe I am doing something wrong?
>
> b_stat_read=1 only on the 2nd pass, right?
> What resolution? At 40kbps, it better be <= QCIF.
> How many frames? 2pass doesn't do much on short videos, other than giving
> more accurate bitrate. It only matters when there's many scenes of varying
> complexity.
>
> --Loren Merritt
>
> --
> This is the x264-devel mailing-list
> To unsubscribe, go to: http://developers.videolan.org/lists.html
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.videolan.org/pipermail/x264-devel/attachments/20051222/051e0cfb/attachment.htm