[FFmpeg-devel] [PATCH] lavfi: add fade filter
Tue Mar 8 13:01:20 CET 2011
On Tue, Mar 08, 2011 at 11:39:01AM +0100, Stefano Sabatini wrote:
> On date Tuesday 2011-02-22 04:53:13 +0100, Michael Niedermayer encoded:
> > On Mon, Feb 21, 2011 at 10:22:00PM +0100, Stefano Sabatini wrote:
> > > On date Monday 2011-02-21 18:06:50 +0100, Michael Niedermayer encoded:
> > > > fade=in:10:20,fade=out:50:60
> > > > could be supported as:
> > > > fade=in:10:20:50:60
> > > >
> > > > and
> > > > fade=out:10:20:50:60
> > > > to black a part out (currently not supported with individual filters easily)
> > > >
> > > > and more than 2, that is N such fade in/outs in a single filter would also be
> > > > usefull
> > > > for example consider that someone concatenates videos from several scenes to
> > > > make a movie. She might want to create fade out+fade in at the transitions
> > >
> > > Uhm, so it may support a list:
> > > fade=out=S:D;in=S:D;out=S:D;...
> > fade=out=S:D:S:D:S:D:...
> > doesnt need escaping and is shorter
> > you cant have 2 out or 2 in after each other so this is not ambigous syntax
> What about having two distinct fade and multifade filters?
iam against that
> My problem with the multifade is that allowing #F and time format
> creates problems since it isn't possible to reliably predict the exact
> timing of a duration expressed using a number of frames, so you could
> have fade-in/fade-out intervals overlap or nesting.
If the user specifies inconsitant parameters thats his fault.
we just should detect it and print an error.
with mixing time in seconds and frames that can only be detected at runtime.
Seems no big issue to me, if the user mixes these 2 he is asking for it.
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Observe your enemies, for they first find out your faults. -- Antisthenes
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 198 bytes
Desc: Digital signature
More information about the ffmpeg-devel