maybe a good project for next years gsoc?
I think a year is a little long to wait.
Preferably getting it done via the donation system would, I think, be for the best.
That way it’s in Blender and with that a lot of the issues with importing/ exporting will be resolved, meaning a lot of bug fixing/ re-coding need not be done on exporters/ importers.
It would also make it feasible for studios and larger productions to work Blender into the pipeline effectively and with that there could be the potential for a few benefits for Blender, such as new add-ons and code from studios making use of it.
+1000 for alembic support - Ton is always talking about how he wants studios and professionals to hire their own blender coders, well this is a major barrier for entry for more studios to pick up blender in the pipeline.
However, to really make it work properly, it would have to be a bit more complicated than ‘just’ an importer/exporter. You don’t just want to import data like you do with an obj, but rather you’d need things like a cache modifier to dynamically fill the contents of your geometry from a disk cache. Other tools to be able to control when geometry data is loaded/unloaded from memory inside the app would be good too. Blender doesn’t have this at the moment (though there were a few patches with similar functionality).
Hi Broken,
I agree It’s not enough, but having a chance to have it as cache format , would permit facilities to test blender as pipeline tools . Particle cache is another thing that is in need to be nailed down in order to make blender suitable for vfx pipelines(not sure if Alebic is suitable for this role)
Alembic could work as just a Mesh/Particle cache. What is really nice about Alembic is that you can rebuild a mesh with the data that is stored in the cache format. So if you don’t have access to the mesh, you can always let Alembic build it for you.
But as for starters, it could be used as a Mesh/Particle Import/Export cache only. And later on extend it.
Alembic is one of the back bones for Katana, http://www.thefoundry.co.uk/articles/2010/06/29/31/katana/
regards
stefan andersson
+1 for alembic (I would donate as well)
I’m working in London atm and even though the bigger studios all have their own cache-format, everyone seems really interested. Compression and performance seem to be the selling point for them. The new Maya Service Pack supports read and write I think, Houdini has a read op. This could open many doors for the Blender community for painless exchange in pipelines.
If we could use this to replace the standard blender cache, I would imagine that it could speed up the write process for physics simulations. I wonder if it will be useful for blender cycles, to save GPU space for storing render data, maybe it doesn’t matter since the renderer might only load one frame at a time. Maybe someone more knowledgable would know whether this is misguided speculation. Regardless, I could really use something robust to get from my other applications to blender, and back out.
Having it as default cache format all over blender, it would permit (for example) to cache a fluid simulation and render it straight away, from renderman just using a procedural pointing to the path.
just what a vfx pipeline is in need of 
also, the further development improving alembic would give automatic bonus to blender.
If you could do that with Renderman and the fluid cache, then perhaps it would also be possible to rig up systems like exporting a stream of heavy animation data to an application like Mari where you could paint per frame, and it would stream the data back into Blender.
You could have a live connection between the two applications and Alembic would be the bridge.
Well that would be possible if Alembic actually supports textures/ UV maps etcetera, which I’d imagine that it does.
yep , is going to be available on all the major sw.
(that’s why would be good for blender to get it while still hot)
2c
I’ll add my vote for Alembic support. This is a huge development in VFX community. It would further allow blender to play well with others and even add some serious improvements to an all-blender pipeline. Didn’t Ton mention Alembic as a dev target during Siggraph, or was I just hearing what I wanted to hear?
I think we hear what we want to hear but I’m ready for donating money if a dev starts funding for alembic.
+1 for donating for the project
Ton updates on status of development fund. Proposals/ideas for new funded projects are open!
do anyone know where the proposals can be entered? I wrote in comment on code.blender.org about looking into focusing BFund towards finding devs to do alembic i/o …
if alembic gets chosen I will probably start monthly donation.
is anyone in touch with ton in here to try to propose it (o maybe is enough to drop an email to the ML…)?
I can talk to him on IRC on saturday/sunday.
thanks
L.