[FFmpeg-devel] [PATCH, v2 2/4] avc/avcodec: add AV_CODEC_CAP_VARIABLE_DIMENSIONS flag

[Anton Khirnov]

Quoting Michael Niedermayer (2020-02-19 19:07:28)
> On Tue, Feb 18, 2020 at 05:42:19PM +0100, Anton Khirnov wrote:
> > Quoting Fu, Linjie (2020-02-18 16:06:10)
> > > > Is there even a sufficiently strong use case for this? Why not just
> > > 
> > > Like transcoding reinit-large_420_8-to-small_420_8.h264 (from 352x288 to 240x196)
> > > and prefer to keep the resolution changing in the output stream. 
> > > 
> > > IMHO not modifying the resolution unless it's required by user is kind of natural.
> > > 
> > > > create a new encoder instance?
> > > 
> > > Would you please help to elaborate more about "create a new encoder instance?"
> > > 
> > > What I intended to do is to flush and close the encoder when resolution changing,
> > > and reinit/recreate a new one.
> > 
> > Yes, that's what I mean. Flush and destroy an AVCodecContext and open a
> > new one. But then why is there a need for this flag?
> 
> some sort of capability check would be needed for flush destroy open too
> because there are a range of cases where this would not produce a decodeable
> file.
> for example some codecs like msmpeg4 do not store the resolution in their 
> header and instead its stored in the container only once per stream.

Right, but that is a property of a given codec ID and should be a flag
in AVCodecDescriptor, not in a specific encoder implementation.
Actually I had some plans to add such a flag and then write a bitstream
filter for concatenating streams where some extra processing is
required.

-- 
Anton Khirnov