[x264-devel] x264-devel Digest, Vol 45, Issue 20

Ajit Deshpande ajitsdeshpande at gmail.com
Fri Feb 11 13:12:27 CET 2011


> Message: 3
> Date: Fri, 11 Feb 2011 01:50:53 -0800
> From: Jason Garrett-Glaser <jason at x264.com>
> To: Mailing list for x264 developers <x264-devel at videolan.org>
> Subject: Re: [x264-devel] x264-devel Digest, Vol 45, Issue 17
> Message-ID:
>        <AANLkTim2O9JiaK7T82fevEYkAumy1LCxSRny5nXixY+q at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> On Fri, Feb 11, 2011 at 1:42 AM, Ajit Deshpande
> <ajitsdeshpande at gmail.com> wrote:
> >
> >>
> >> On Feb 10, 2011, at 12:17 PM, Patrick White wrote:
> >>
> >> >
> >> > Hi Dennis,
> >> >
> >> > There is indeed a parameter to change, in the x264_param_t struct, the
> >> > "i_slice_max_size" variable.
> >> >
> >> > There are also related options for macro-blocks and slices.
> >> >
> >
> > Just curious about how "i_slice_max_size" would work exactly. Would
> setting
> > this parameter influence the - Motion estimation, mode decision algorithm
> &
> > rate control blocks in the
> > encoder such that this max slice size limit is adhered to?
> >
> > thank you.
> >
> > -AD
>
> Encode MBs until the slice max size is exceeded.  Then back up to the
> one just before and end the slice.
>
> Jason
>
> Hello Jason,
  But what happens if FMO is on? In that case  would it mean, that a user
defined MB allocation map to make a slice of those MBs is over-ridden by
i_slice_max_size?
Could it happen that when encoder backs up to the preceeding MB to obey the
slice max size limit, and ends the Slice/NAL unit there, and thus does not
encode as per the user entered configuration of MB allocation to a slice.

thank you.
-Ajit.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x264-devel/attachments/20110211/ae7fb22c/attachment.html>


More information about the x264-devel mailing list