[MPlayer-dev-eng] [PATCH] edl

[Michael Halcrow]

On Mon, 2003-01-20 at 05:53, Kövesdi György wrote:
> Yes, you are right, it needs optimizations. I have some ideas, but if 
> you have a newer version, i will wait.
> Do you included it in mencoder too? I would prefer if the code would 
> not be doubled in mplayer and mencoder.

I have not included it in mencoder yet. I've been working a lot recently
in my research lab (I'll lose my job if I spend too much time not
putting in hours because I'm hacking mplayer :-)

I'm including my old patch here (IMO, it is not quite ready for
application, but it works on my end), which has problems applying to the
current version of the mplayer source (this patch is a few weeks old),
but it's something to work with. It fixes the human-EDL muting
relationship (if the human mutes, the EDL will not unmute unless the
human unmutes too, etc.). I still need to refine the seek-recovery code
(it does a binary search, and then executes the record immediately
preceeding the current pts, which would be undesirable if the previous
record were a seek). I also include a vf_rectanglefill.c video filter,
but it does not correctly copy over the planes in memory (refer to the
recent vf_rectangle.c fix for an example of how it's supposed to be
done). These are all things I still need to get around to; someone else
can do it if he'd like in the meantime.

If we think that it does not add too much overhead to the playback, I
would not be totally opposed to re-searching the EDL list for the next
record to execute on every cycle, because I would rather not have to
think about *every* case where the EDL pointer would need to be re-set.
Of course, if we're going to do that, then the seek should definitely be
a binary search, which requires the EDL records to remain in an array.
If we just look at cases where the playback pts changes and code for
them, then future changes that affect the pts could easily break EDL.

Mike

-- 
---------------------------------------- | ------------------------
Michael Halcrow                          | mhalcrow at byu.edu    
Research Assistant, Network Security Lab | Dept. of Comp. Science  
                                         | Brigham Young University
"Campus sidewalks never exist as the     |
straightest line between two points."    |
  - M. M. Johnston                       |
---------------------------------------- | ------------------------
GnuPG Keyprint:  05B5 08A8 713A 64C1 D35D  2371 2D3C FDDA 3EB6 601D
-------------- next part --------------
A non-text attachment was scrubbed...
Name: edl0_6.diff
Type: text/x-patch
Size: 24340 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/attachments/20030120/24cf02fe/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 232 bytes
Desc: This is a digitally signed message part
URL: <http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/attachments/20030120/24cf02fe/attachment.pgp>