[Foundation-l] Why don't we re-encode proprietary formats as Ogg?

Mon Jun 8 21:57:39 UTC 2009

We have done a good amount of work with archive.org to ensure that their 
archive is interpretable. I know from the present vantage point it does 
not seem helpful to have media on archive.org... but as features like 
the add_media_wizard get deployed it will make a lot more sense why it 
does not matter so much where the source media is hosted.

In terms of source files... I think the problem is people don't 
necessarily have the bandwidth to upload their raw DV footage...  If 
they do then we should also upload a copy to archive.org. Using firefogg 
its easy to add a js function call to also send a copy of the raw 
non-ogg encoded footage to archive.org all from within our commons 
upload interface.

Its of course always better to have the original. But I would argue (for 
the time being) we should store that original on archive.org rather than 
build out and maintain all the trannscoder & raw footage storage 
infrastructure internally. In the future if we do have time and or 
resources (volunteer or otherwise) to support transocding on wikimedia 
commons... then thats great and we can support that too.

In terms of encoder updates: Firefogg lets us control the encoder via 
Firefox extension auto updates. Firefogg is already running the 
thusnelda encoder branch. In the future we can push out other free 
encoders (dirac speex etc). This makes it much easier for someone to 
build a collaborative video wiki since they don't need to build out 
transcoding infrastructure.

--michael

Tim Starling wrote:
> Peter Gervai wrote:
>   
>> On Sun, Jun 7, 2009 at 17:26, David Gerard<dgerard at gmail.com> wrote:
>>     
>>> It would be a simple matter of programming to have something that
>>> allows upload of encumbered video and audio formats and re-encode them
>>> as Ogg Theora or Ogg Vorbis.
>>>       
>> As a technical sidenote, it should be mentioned that recoding a lossy
>> format to another lossy format results _always_ a worse quality output
>> than the source lossy format. The amount of quality loss depends on
>> countless factors and usually do not render the result useless, but
>> the quality difference may be still audible/visible.
>>     
>
> But if we can do resizing and quality conversion post-upload, then we
> can encourage users to upload their videos with the best possible
> quality, they won't be forced to upload at a quality suitable for web
> download. We can store the source video unconverted for archival
> purposes. When we reduce the quality of a video for a web user, the
> process will be under server control and can be incrementally
> improved, instead of using whatever outdated software the user has on
> their computer. So the net effect of the feature should be a
> significant improvement in video quality.
>
> -- Tim Starling
>
>
> _______________________________________________
> foundation-l mailing list
> foundation-l at lists.wikimedia.org
> Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/foundation-l
>   