[MEncoder-users] Re: forcing encode of lost frames as duplicates

[Josh Sutton]

Corey Hickey <bugfood-ml <at> fatooh.org> writes:

>> Yep.  Reading between the lines, I don't think this is what the harddup is
>> actually meant for, so it's not totally suprising that it doesn't do what 
>> I want it to do.
>> 
>> What I'm looking for is a tool that will blindly write new frames to disk 
>> even if they weren't received.  So that frame rate and hence playback time is
>> preserved.   Perhaps there is a better way to achieve this than using these
>> tools, but so far I have not found it.
> 
> Yeah, I was afraid that's what you meant.
> 
>>>> I suspect because the frames are missing, not duplicates.
>>> What exactly is the video data you're feeding mencoder? Is it an MPEG,
>>> or a series of JPEGs, or what?
>> 
>> The input stream is an mpeg stream from an rtsp session.  The mencoder 
>> command line and output is below.  If frames are not received, then the 
>> transmit time stamp on the encoded video jumps from say 11 seconds to 13 
>> seconds.  What I want is for the the encoded video to hold the frame
>> received at 11 seconds for a full 2 seconds and then move on to the frames
>> received at 13 seconds.
> 
> I can't think of any way to do that with mencoder, short of making a
> patch for either the demuxer or muxer to generate those blank frames. I
> don't have the time to figure out how to do that, though.
> 
> If anybody else has any good ideas then he or she should chime in.
> Perhaps there is some creative use of mencoder or another program I
> can't think of.

Thanks anyway.

Just one more thought.  Is it possible to have mencoder encode the video stream
from the video server and also an audio stream of silence from a separate
source?  That might force mencoder to keep writing blank or duplicate a/v frames
because it needs to keep encoding the audio?  The man page doesn't mention
anything specific, but it seems like an obvious feature of the muxer to be able
to take audio and video from separate streams.