[mythtv] Trick play patch

Isaac Richards ijr at po.cwru.edu
Tue Aug 19 14:31:15 EDT 2003

On Wednesday 13 August 2003 07:03 pm, David Engel wrote:
> That's because the player class has direct control over the decoder.
> The TV class doesn't have this so it's not as straight forward.  The
> simple change for this is to have the player always make the decoder
> do exact seeks for single frame increments.  I just tested this and it
> works fine.  Should exact seeks also be done for 1 second increments?

Exact seeking is rather slow, but if it's only done (while the 'exact seeks' 
option is turned off) while paused, that's fine.

> One other thing I just noticed.  The first frame ahead/back now starts
> out slightly after where the video was initially paused.  For example,
> if I pause the video right after the pitcher releases the ball and
> then frame ahead, the next video shown shows the catcher already
> having caught the ball.  Frame ahead/back after that work as expected.
> I don't recall it doing this before.
> I did a cvs up right before testing this latest change.  Could one of
> the commits done in the last couple days have affected this?

I don't think anything would have that affect..  It's not happening here, 

> On Wed, Aug 13, 2003 at 02:51:53PM -0400, Isaac Richards wrote:
> > I'm just wondering if there's a better way to handle that, is all.  I
> > need to think about it s'more.
> OK.

Well, I couldn't think of a nicer way to do it, so I went and committed your 
patch =)  I'll probably add another variable sometime so it's not doing the 
dividing by play_speed all the time later.


More information about the mythtv-dev mailing list