Mailing List Archive

VideoSync class
Hi Mark

Since you removed avsync and renamed avsync2, the synchronization
classes in vsync.cpp are now not needed. However, the VideoSync Object
is still being created in mythplayer.cpp and a sync method is being
reported, but as far as I know it is not actually used for anything
relevant. Do you agree that vsync.cpp and vsync.h should be removed
together will all references to the classes defined in vsync.h?

Peter
_______________________________________________
mythtv-dev mailing list
mythtv-dev@mythtv.org
http://lists.mythtv.org/mailman/listinfo/mythtv-dev
http://wiki.mythtv.org/Mailing_List_etiquette
MythTV Forums: https://forum.mythtv.org
Re: VideoSync class [ In reply to ]
On Wed, 8 Jan 2020 at 20:01, Peter Bennett <pb.mythtv@gmail.com> wrote:
>
> Hi Mark
>
> Since you removed avsync and renamed avsync2, the synchronization
> classes in vsync.cpp are now not needed. However, the VideoSync Object
> is still being created in mythplayer.cpp and a sync method is being
> reported, but as far as I know it is not actually used for anything
> relevant. Do you agree that vsync.cpp and vsync.h should be removed
> together will all references to the classes defined in vsync.h?

Hi Peter

I hadn't realised it was effectively no longer used:)

So - yes, they should probably go if everyone is happy that the new
avsync code is working as expected.

Regards
Mark
_______________________________________________
mythtv-dev mailing list
mythtv-dev@mythtv.org
http://lists.mythtv.org/mailman/listinfo/mythtv-dev
http://wiki.mythtv.org/Mailing_List_etiquette
MythTV Forums: https://forum.mythtv.org
Re: VideoSync class [ In reply to ]
> Wiadomo?? napisana przez Mark Kendall <mark.kendall@gmail.com> w dniu 09.01.2020, o godz. 09:39:
>
> On Wed, 8 Jan 2020 at 20:01, Peter Bennett <pb.mythtv@gmail.com> wrote:
>>
>> Hi Mark
>>
>> Since you removed avsync and renamed avsync2, the synchronization
>> classes in vsync.cpp are now not needed. However, the VideoSync Object
>> is still being created in mythplayer.cpp and a sync method is being
>> reported, but as far as I know it is not actually used for anything
>> relevant. Do you agree that vsync.cpp and vsync.h should be removed
>> together will all references to the classes defined in vsync.h?
>
> Hi Peter
>
> I hadn't realised it was effectively no longer used:)
>
> So - yes, they should probably go if everyone is happy that the new
> avsync code is working as expected.

Mark,

Is it possible to add back old sync method for some time just for testing purposes?

Reasons:
1. new method has some side effects for me.
Namely:
-on nvidia playback is smooth but I can notice with very little (but noticeable) micro jumps (best seen on new scrolling bars).
They are random - every few to 10sec - but noticeable.
Old sync hasn't this effect - so it will be good to compare and make sure new method not causing this issue.
-on stateful decoders (amlogic) playback sometimes enters jumpy video (audio is still ok). To fix - i need to exit playback.
I’m not sure where issue is - but it will be good to exclude new sync root cause here.

2.Good test for new sync method is not regressing is to make sure all sync playback issues (if present) - are manifesting on both methods.
This will make us sure that new method not regressing when compared to old one in corner cases….

br


_______________________________________________
mythtv-dev mailing list
mythtv-dev@mythtv.org
http://lists.mythtv.org/mailman/listinfo/mythtv-dev
http://wiki.mythtv.org/Mailing_List_etiquette
MythTV Forums: https://forum.mythtv.org
Re: VideoSync class [ In reply to ]
On Thu, 9 Jan 2020 at 18:34, Piotr Oniszczuk <piotr.oniszczuk@gmail.com> wrote:
> Is it possible to add back old sync method for some time just for testing purposes?
>
> Reasons:
> 1. new method has some side effects for me.
> Namely:
> -on nvidia playback is smooth but I can notice with very little (but noticeable) micro jumps (best seen on new scrolling bars).
> They are random - every few to 10sec - but noticeable.
> Old sync hasn't this effect - so it will be good to compare and make sure new method not causing this issue.
> -on stateful decoders (amlogic) playback sometimes enters jumpy video (audio is still ok). To fix - i need to exit playback.
> I’m not sure where issue is - but it will be good to exclude new sync root cause here.
>
> 2.Good test for new sync method is not regressing is to make sure all sync playback issues (if present) - are manifesting on both methods.
> This will make us sure that new method not regressing when compared to old one in corner cases….

Piotr

Adding back old code is not a priority for me. Time is currently
limited for development work so I would rather focus on fixing issues
rather than trying to work out whether they worked before or not.

And you seem confident that the old sync worked anyway?

I'd rather see some playback logs and understand if the issues are
repeatable and fixable.

Thanks and regards
Mark
_______________________________________________
mythtv-dev mailing list
mythtv-dev@mythtv.org
http://lists.mythtv.org/mailman/listinfo/mythtv-dev
http://wiki.mythtv.org/Mailing_List_etiquette
MythTV Forums: https://forum.mythtv.org