On Thu, 3 Apr 2008, Florent wrote:

> As you may (skinkie) or not (hi koos) know, i am developing a video
> compositing engine on top of clutter and gstreamer.
>
> Attached is a preview (work in progress) of a simple animation wrapper
> in python, with an effect example. It wraps the depth, opacity and
> position behaviours. It features a "wait_for_completion" handler
> (allow or not an animation to be fired before completion) and offers
> some default parameters (overridable) that lets try them out fast.
> It's far from complete, and it lacks a plugin architecture (on our
> roadmap) but i'm very curious about what you guys may think of such an
> approach.
>
> I'd be glad to know:
> * what language you will/want to develop your glSMIL player(s) in ?

Preferably C (for the clutter related part).

> * what you think of this preliminary proposal (it's not that elegant
> nor scalable, yet i implemented a bunch of effects quite easily with
> such wrappers)

I'll take a look at your code :) I'm not against Python as prototyping
code, and it might be possible to interface with Python effects all
together (so it is possible to have dynamic effects).

> * the thing with SMIL is that it's not taking 3D effects into account
> (AFAIK). How to keep both worlds in the same rendering and compositing
> engine ?

I guess we could just propose extentions to the transition effects (for
SMIL3+ versions). Or do you think the amount of parameters are a problem?

> * for the SMIL effects, there are a lot of cropping-based effects, i
> don't know how well clutter would perform by re-clipping incrementally
> a texture using a timeline...
>
> On my side, i am very interested in SMIL support as well, yet not only
> for playback but rather for "recording"/animation archival purposes.
> Please let me know when/if you join forces !

You might know that dvswitch wants to use compositing too. I am interested
in SMIL from an output perspective. Tonight 22:30 (GMT+1) we are joining
forces again in #clutter-smil lets brainstorm and divide work :)


Stefan

-- 
To unsubscribe send a mail to [EMAIL PROTECTED]

Reply via email to