[vlc-devel] [PATCH 00/17] Fixing normal_time handling

[Rémi Denis-Courmont]

Hi,

No it's not the hard part. You have demuxers where the NPT can be computed properly, and it will be a step-wise offset of the timestamps, most often 0 or VLC_TS_0, with the notable exceptions of TS and discs.

And then you have some pathological cases like seekable TS without EIT. But computing an NPT offset is not harder or more jittery than whatever is being done right now to estimate the NPT value.

Le 12 mars 2021 12:05:33 GMT+02:00, Francois Cartegnie <fcvlcdev at free.fr> a écrit :
>Le 11/03/2021 à 16:28, Rémi Denis-Courmont a écrit :
>
>> And my point is that we should just report that offset (second bullet
>point)
>> from NPT to PCR whenever it changes.
>
>That's the hard task. I tried to bind both references on my side and 
>there's more than one corner case.
>
>As they usually don't have the same update frequency, depending on when
>
>you create a reference point, or update it, that creates unwanted
>jitter 
>or unaligned start offset.
>
>-- 
>Francois Cartegnie
>VideoLAN - VLC Developer
>_______________________________________________
>vlc-devel mailing list
>To unsubscribe or modify your subscription options:
>https://mailman.videolan.org/listinfo/vlc-devel

-- 
Envoyé de mon appareil Android avec Courriel K-9 Mail. Veuillez excuser ma brièveté.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/vlc-devel/attachments/20210312/b788db5d/attachment.html>