[mythtv-users] does adjust fill work with vdpau?

[aaron]

On Sun, May 16, 2010 at 17:25, Brian J. Murrell <brian at interlinx.bc.ca> wrote:
>
>> Or, could the commercial flagger do this and store the info for future
>> use?  Shouldn't be much info:  frames x to y are this size, frames y+1
>> to z are this size, etc.
>
> Ahhhh.  Now, that sounds like a great idea.  The flagger is decoding
> frames anyway, so why not store the data that the software decoder would
> use to decide on auto-filling or not.  Awesome!

I understood (could very well be wrong) that the commflagger already
stored information about letterboxing (and etc) in the database. I had
wanted to see about getting the "auto" fill mode detection to use that
data rather than looking at each frame during runtime, but as I said
at the time, a team of monkeys could probably bang out a working patch
before I get enough free time to look into it. I still haven't had any
time to look into it.

aaron