[Bf-committers] Geometry in Compositor or Quadrangulation???

[François T.]

you should talk to Jeroen Bakker who is working on the redesign of the
compositor. I know he has some plan to implement something close to Daniel's
proposal. Maybe you guys should talk so you can go in the same way, or maybe
help him with that, I don't know.


cheers

F.


2011/4/13 pete larabell <xgl.asyliax at gmail.com>

> If this is ok with blender peeps that I implement something like
> this... I can start work on it.
>
> Is there any major reasons I *shouldn't* do something similar to that?
>
> On Tue, Apr 12, 2011 at 10:18 PM, Daniel Salazar - 3Developer.com
> <zanqdo at gmail.com> wrote:
> > Ok here's the ultimate sucky mockup for a 2D masking environment. I
> > focuses on a getting the basics of masking well done. leaving better
> > integration with 3D scene for later.
> >
> > http://www.zanqdo.com/tmp/MockUpMask.png
> >
> > Daniel Salazar
> > 3Developer.com
> >
> >
> >
> > On Tue, Apr 12, 2011 at 3:54 AM, Daniel Salazar - 3Developer.com
> > <zanqdo at gmail.com> wrote:
> >> after a chat with Matt we agreed that there are two areas to take care
> >> now or eventually:
> >>
> >> 1- masking tools: simple, independant from 3D view, interactive,
> >> probably work inside *image editor*. rotobezier, power windows, etc go
> >> here
> >>
> >> 2- integration of 3D scene: this should be done by making a better
> >> *blender internal*. the renderer should handle non obstructive
> >> rendering of custom jobs, like it does with material preview renders.
> >> It should handle different cameras or resolution (to be able to render
> >> outside of main view's screenspace) and it should be able to render
> >> simple geometry like shadeless objects or normals as fast as possible.
> >> This is actually simmilar to an old design of mine that uses the 3D
> >> view instead of the image editor but still focuses on tweaking Blender
> >> Internal to be more flexible
> >>
> >>
> https://docs0.google.com/drawings/edit?id=1VLs3SkOjn9TVYtet0KvNTVLWvcgaH4gVYIUrMOMp_xo&authkey=CIK37ecL&hl=en
> >>
> >> what Pete is doing is probably a mix between the two and that's
> >> probably a bad idea :s
> >>
> >> cheers!
> >>
> >> Daniel Salazar
> >> 3Developer.com
> >>
> >>
> >>
> >> On Tue, Apr 12, 2011 at 12:04 AM, Daniel Salazar - 3Developer.com
> >> <zanqdo at gmail.com> wrote:
> >>> to elaborate the problems with everything that has to be with scene
> >>> data and compositing together are speed/interactivity and integration.
> >>> what do you suggest?
> >>>
> >>> Daniel Salazar
> >>> 3Developer.com
> >>>
> >>>
> >>>
> >>> On Mon, Apr 11, 2011 at 11:58 PM, Daniel Salazar - 3Developer.com
> >>> <zanqdo at gmail.com> wrote:
> >>>> What do you suggest? (not focusing on the pure topic of masking)
> >>>>
> >>>> Daniel Salazar
> >>>> 3Developer.com
> >>>>
> >>>>
> >>>>
> >>>> On Mon, Apr 11, 2011 at 11:51 PM, Matt Ebb <matt at mke3.net> wrote:
> >>>>> On Tue, Apr 12, 2011 at 3:35 PM, Daniel Salazar - 3Developer.com
> >>>>> <zanqdo at gmail.com> wrote:
> >>>>>> Indeed! :) *but!* there are other uses for having a geometry node in
> >>>>>> the compositor like bringing geometry normals, vectors, alphas and
> >>>>>> what not and all interactive (no need for regular render). It's what
> >>>>>> other compositors do to integrate the 3D view with the compositor.
> We
> >>>>>> can see this as a step of integration. What do you think Matt?
> >>>>>
> >>>>> I think it's a bad idea. Blender already has a renderer and that's
> >>>>> what it's for. Duplicating code to make an entirely separate renderer
> >>>>> that's only used in the comp would end up in a world of
> >>>>> overcomplicated pain. If there are problems with the workflow of
> >>>>> rendering elements to be used in comp, then that should be worked on
> >>>>> itself, I don't think the solution is to ignore it and build an
> >>>>> entirely separate thing.
> >>>>>
> >>>>> But that's all putting the cart way before the horse anyway, when so
> >>>>> much of blender's compositor is still at quite a basic level for 2d
> >>>>> manipulations.
> >>>>>
> >>>>> Matt
> >>>>> _______________________________________________
> >>>>> Bf-committers mailing list
> >>>>> Bf-committers at blender.org
> >>>>> http://lists.blender.org/mailman/listinfo/bf-committers
> >>>>>
> >>>>
> >>>
> >>
> > _______________________________________________
> > Bf-committers mailing list
> > Bf-committers at blender.org
> > http://lists.blender.org/mailman/listinfo/bf-committers
> >
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> http://lists.blender.org/mailman/listinfo/bf-committers
>



-- 
____________________
François Tarlier
www.francois-tarlier.com
www.linkedin.com/in/francoistarlier