On Sun, Jun 30, 2013 at 3:20 PM, Tomas Neme <lacrymol...@gmail.com> wrote:
> I don't exactly know why, but the segfault disappeared. Now I'm back
> to the speed problem. If I run this, the "current position" counter in
> melt stays put in 0
>
> $ melt -producer avformat:bm.mp4 -consumer posixshm
>
> but if I run it with the -debug flag, I get this:
>
>
> [producer avformat] bm.mp4 0
>     V pkt.pts 636636 pkt.dts 632632 req_pos 152 cur_pos 151 pkt_pos 157
> ....
> [producer avformat] bm.mp4
>     V pkt.pts 644644 pkt.dts 636636 req_pos 153 cur_pos 152 pkt_pos 159
> ...
> [producer avformat] bm.mp4
>     V pkt.pts 652652 pkt.dts 640640 req_pos 154 cur_pos 153 pkt_pos 161
>
> so the consumer thread is running, but apparently the position isn't
> changing. Should I be doing something manually?

You should add this in your consumer thread after your call to output().
mlt_events_fire( properties, "consumer-frame-show", frame, NULL );

> I see the sdl consumer does some queuing of frames, and waits until
> the proper time to show it. Should I be doing something similar? To

The consumer controls the timing. That is the audio clock in something
like the sdl consumer, but decklink and sdi consumers use their
SDK/driver for timing. In your use case, you probably do need some
kind of a timer.

> me, it'd make more sense to request the frames as needed, because
> there's some filters that will depend on real-time data, plus the
> input right now is a file, but it could very well be a live stream, or
> whatever, so I'm not sure what I'm doing wrong, but the sdl consumer
> is not an exact model for what I want to do.

The other process's producer that reads the shm could use some IPC or
signaling to this consumer, but I do not have a specific
recommendation here.

--
+-DRD-+

------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Mlt-devel mailing list
Mlt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mlt-devel

Reply via email to