[CinCV TNG] MPEG-2 native lossless editing without re-encoding

Terje J. Hanssen terje at nordland-teknikk.no
Wed Dec 28 19:50:52 CET 2016

Den 28. des. 2016 16:45, skrev Einar R√ľnkaru:
> On 12/28/2016 04:13 PM, Terje J. Hanssen wrote:
>> Maybe a stupid question, but can someone clarify what happends when
>> typical a source mepg2 video file is loaded into Cinelerra, on the time
>> line and available as a Media asset in Ressources?
> Cinelerra decodes frames needed to fill timeline and compositor. Then 
> it waits for user input.
>> That is, does Cinelerra have (use) its own "internal format" for video
>> and audio, so that already after loading, the source file has been
>> decoded from the source mpeg2 ?
> Yes. The intened usage of Cinelerra is to modidify audio samples and 
> video pixels. One can control the internal format from 
> Preferences/Format.
>  And next to render this to the "same
>> mpeg2", is re-encode and/or passthrough used?
> When there are no edits, it is possible that cinelerra remuxes packets 
> from original file to the result. It happens automatically without 
> user control.
>> Doing this with Cinelerra-gg 5.1 and "try FFmpeg first" mode set, the
>> Media info in Ressources say the file format is "FFMPEG" and list more
>> in the details.
> Never tried gg. My explanations are for CV. I dont know wether HV or 
> GG have this feature. It is removed from CVE.


Thank you for throwing some light on this area.
If someone has more enlightment to add to this subject, it is welcome.

Terje J. H

More information about the Cinelerra mailing list