[Bf-committers] Re: render passes & workflow

trip somewhere trip0o at gmail.com
Thu Dec 7 17:08:49 CET 2006


First real test image
http://img467.imageshack.us/img467/7466/tonpassesle6.png

Forgive me on any missing to dos I know your still working on it. I'm
just very excited.

Two Ui add ins
Auto name the Viewer nodes per whatever the pass is That way it is
easy to see what each out put is

Z still out puts nothing unless map to is used, there must be a
cleaner way to just get a closer Z image at start.

Color wires in a path that are much further than half way get very
confusing to find them, i know they have a gradient fall off but if to
far away it shows nothing unless moved around to see it tail.

Shadow is weird, shadow should not have color. Might just be me...

On the nodes auto connections in between node connections. Aka add a
rgb curves between a pass and a viewer, aka then a mix and whatever,
instead of having to reattach them allover again

Weeee fun

On 12/7/06, trip somewhere <trip0o at gmail.com> wrote:
> Oooo more to read ! :D just updated cvs testing now.
>
> On 12/7/06, Ton Roosendaal <ton at blender.org> wrote:
> > Hi,
> >
> > I'd suggest to read my logs:
> > http://www.blender3d.org/cms/Render_Passes.829.0.html
> >
> > It even has images! But also read the text, like in bottom "todo".
> >
> > -Ton-
> >
> > On 7 Dec, 2006, at 15:48, trip somewhere wrote:
> >
> > > Hmm so what I can get from all of this is that every change that is
> > > made to a pass has to be re-rendered there by having to re-render
> > > everything over again.. =?
> > >
> > > If that is so then at least it is a great start.
> > >
> > > On 12/7/06, Johnny Matthews <johnny.matthews at gmail.com> wrote:
> > >> Now one thing that would be exceptionally cool is if the image node
> > >> could then read these layers and provide outputs for each layer of the
> > >> inputted file. So if you saved an anim as exr with passes, you could
> > >> get it back in super easy, with all the passes available. Or do the
> > >> same with other layered file types like PDF.
> > >>
> > >> On 12/7/06, Ton Roosendaal <ton at blender.org> wrote:
> > >> > Hi,
> > >> >
> > >> > OpenEXR was designed to be multi-layer multi-pass. And since it's an
> > >> > open format, something we should support first, and discuss with the
> > >> > ILM team if we can somehow standardize naming for layers/passes.
> > >> They
> > >> > then can take care of promoting this among the other application
> > >> > developers, like for the plugin in Photoshop.
> > >> >
> > >> > For the time being, I can at least ensure Blender reads/writes such
> > >> > multi-exr files well, also for image window and image-input nodes.
> > >> >
> > >> > Some interested developer then can also look at adding the
> > >> > Combuston/3DS RPF (Rich Pixel Format) format. Or, also seems to be
> > >> > supported: Irix RLA Files (currently already in Blender, in the
> > >> past we
> > >> > added a Z to this (IriZ), but that appeared to be not readable in
> > >> any
> > >> > other program than Blender...).
> > >> >
> > >> > -Ton-
> > >> >
> > >> >
> > >> > On 7 Dec, 2006, at 13:58, Roland Hess wrote:
> > >> >
> > >> > > Almost any article I've read on using pass rendering seems to
> > >> focus on
> > >> > > doing your final renders in passes for greater flexibility down
> > >> the
> > >> > > pipeline. Now, if the rest of your pipeline is Blender's
> > >> compositor,
> > >> > > then it's not an issue. However, if you're using something else as
> > >> > > your target -- putting this in with a bunch of other apps, or
> > >> handing
> > >> > > the files off to another "department" for compositing/sweetening
> > >> --
> > >> > > you would want to be able to have all passes saved into a file
> > >> format
> > >> > > straight out of the renderer in which each pass is a layer (per
> > >> the
> > >> > > image editor of your choice) set with the appropriate blend mode
> > >> flags
> > >> > > so that rendered files can be opened in either a still editor or
> > >> > > compositing app with layers (passes) intact, blended properly and
> > >> > > ready to be tweaked.
> > >> > >
> > >> > > Unfortunately, OpenEXR isn't up to this sort of challenge right
> > >> now --
> > >> > > the file format may be able to handle the channels, but there
> > >> doesn't
> > >> > > seem to be support for having those appear in image editors in a
> > >> > > useful way. Of course, it wouldn't hurt if Blender could read such
> > >> > > files right back in at a later date, allowing you access in the
> > >> > > compositor to each channel set appropriately.
> > >> > >
> > >> > > On Dec 6, 2006, at 4:33 PM,
> > >> bf-committers-request at projects.blender.org
> > >> > > wrote:
> > >> > >
> > >> > >> Now that the basic code work was done, it's time to check on
> > >> usablity
> > >> > >> of pass rendering.
> > >> > >> (No Trip, not usuability of compositor, that's another project!).
> > >> > >>
> > >> > >> Two ideas I got in mind now:
> > >> > >>
> > >> > >> 1 Make the "Combined" pass (optional) exclusive passes.
> > >> > >>
> > >> > >> My assumption is that you only use passes for specific cases,
> > >> like
> > >> > >> only
> > >> > >> the shadow pass or only the reflection pass. In that case, a
> > >> composite
> > >> > >> now is difficult to use when "Combined" has this pass info added
> > >> (for
> > >> > >> example blurring shadow).
> > >> > >>
> > >> > >> 2 Give RenderLayers two new options:
> > >> > >> - Light (Group) override
> > >> > >> - Material (incl Node tree) override
> > >> > >>
> > >> > >> This option will render everything in the RenderLayer with a
> > >> specific
> > >> > >> group of Lamps, and/or with a specific Material. That's a much
> > >> more
> > >> > >> powerful way than define a "custom pass", since then you have
> > >> the the
> > >> > >> RenderLayer result again with all existing passes available.
> > >> > >> It also gives a nice use for the to-be-added Python Node
> > >> shaders. :)
> > >> > >
> > >> > > Roland Hess
> > >> > > IT Manager/Digital Prepress
> > >> > > Reed & Witting Company
> > >> > > Pittsburgh, PA
> > >> > > 412-682-1000
> > >> > > rolandh at reed-witting.com
> > >> > >
> > >> > >
> > >> > >
> > >> > >
> > >> > > _______________________________________________
> > >> > > Bf-committers mailing list
> > >> > > Bf-committers at projects.blender.org
> > >> > > http://projects.blender.org/mailman/listinfo/bf-committers
> > >> > >
> > >> >
> > >> ----------------------------------------------------------------------
> > >> --
> > >> > --
> > >> > Ton Roosendaal  Blender Foundation ton at blender.org
> > >> > http://www.blender.org
> > >> >
> > >> > _______________________________________________
> > >> > Bf-committers mailing list
> > >> > Bf-committers at projects.blender.org
> > >> > http://projects.blender.org/mailman/listinfo/bf-committers
> > >> >
> > >>
> > >>
> > >> --
> > >> Johnny Matthews
> > >> johnny.matthews at gmail.com
> > >> Check out my blog at - http://johnnygizmo.blogspot.com
> > >>
> > >> "Any sufficiently advanced technology is indistinguishable from
> > >> magic."
> > >> - Arthur C. Clarke
> > >> _______________________________________________
> > >> Bf-committers mailing list
> > >> Bf-committers at projects.blender.org
> > >> http://projects.blender.org/mailman/listinfo/bf-committers
> > >>
> > > _______________________________________________
> > > Bf-committers mailing list
> > > Bf-committers at projects.blender.org
> > > http://projects.blender.org/mailman/listinfo/bf-committers
> > >
> > >
> > >
> > ------------------------------------------------------------------------
> > --
> > Ton Roosendaal  Blender Foundation ton at blender.org
> > http://www.blender.org
> >
> > _______________________________________________
> > Bf-committers mailing list
> > Bf-committers at projects.blender.org
> > http://projects.blender.org/mailman/listinfo/bf-committers
> >
>


More information about the Bf-committers mailing list