diff mbox

[FFmpeg-devel,1/3] avcodec/vp56: Require 1 undamaged frame for 5 damaged frames for concealment to be used

Message ID 20170316100037.GI5776@nb4
State Not Applicable
Headers show

Commit Message

Michael Niedermayer March 16, 2017, 10 a.m. UTC 
On Wed, Mar 15, 2017 at 08:44:25PM -0400, Ronald S. Bultje wrote:
> Hi,
> 
> On Wed, Mar 15, 2017 at 6:12 PM, Michael Niedermayer <michael@niedermayer.cc
> > wrote:
> 
> > On Wed, Mar 15, 2017 at 07:18:30PM +0000, Kieran Kunhya wrote:
> > > >
> > > > > > I have tons of testcases for h264 that are 1KB and can make error
> > > > > > concealment run for ages.
> > > >
> > > > and how is this related to a fix for th vp* decoder ?
> > > >
> > >
> > > My point is you can spend a lifetime fixing obscure conditions that cause
> > > error concealment to take a long time.
> >
> > the vp56 EC code
> > just sets every block to the reference frames content
> 
> 
> I don't see how this helps the user experience.

Like i said previously if there is damage, its better with error
concealment than without. Thats true in general not specific to vp56.

also without EC we will need to change fate this way:


as the last frame in the sample is damaged, probably truncated
we cannot output a partly uninitialzed frame nor can we use it as
reference frame afterwards.
If there where frames afterwards they would use a wrong/different
reference frame or be discarded as well until the next keyframe.


> 
> I once again think you should consider removing the vp56 EC code until we
> have a better idea of requirements of and need for such a feature.

As the one having implemented it i see both requirement and need.
So theres no sense or way i could consider removing it on the basis
that these are missing or until they are found.

I can remove it becouse people want it removed.

also for this specific patch here there is possibly another solution
which ill test and if it works submit a patch

[...]
diff mbox

Patch

--- ./tests/ref/fate/vp5        2017-03-13 15:59:01.833910756 +0100
+++ tests/data/fate/vp5 2017-03-16 09:32:21.934865428 +0100
@@ -249,4 +249,3 @@ 
 0,        243,        243,        1,   233472, 0x6f530ac6
 0,        244,        244,        1,   233472, 0x94f7466c
 0,        245,        245,        1,   233472, 0xa8c1d365
-0,        246,        246,        1,   233472, 0xbf73f1b7