Hi ,
now that alembic is been released ,
is there any chance to see it integrated into blender (even as addon if the licence doesn’t permit the actual lib linking)?

I for one hope so, can’t wait to get bullet physics into blender also (a bit of shame that maya, c4d… beaten us)

but to be realistic, there’s shit loads of stuf lined up to get merged… I’ve never seen so much stuf going to merge at once. there’s gotta be problem but we can hope it goes somewhat smooth.

i honestly can’t see alembic top prio.

unless someone just make a i/o script, maybe then later someone can add it to the source.

hi aermartin,
I know there’s quite a long queue of interesting projects waiting for integration , but if you think that pretty much all the nex sw releases (maya 2012/13 , houdi, and pretty much all the upcoming foundry releases) are going to integrate it as file format , you can easly spot why i think is quite important to have it :slight_smile:
(makes easier for a mesh to move from a sw to another which permits to blender to have a chance of a pipeline integration)

just 2c.

yeah the queue is C code though or C++ projects, what could be done is a ‘addon’ with pythonscript to import/export just to get it up and working quickly. what’s needed then later is a rewrite in C for speed when dealing with huge mesh data is suppose.

let’s just wait and see ;D alembic is pretty new stuf. it’s version 1.0 now and they might even change stuf. but I as you hope we’ll see this in blender asap.

We have been using Alembic at work with Maya for some time now, and so far… it rocks! I don’t think there is a smarter and faster point cache format out there. I really hope someone with some programming experience for blender picks it up.

stefan andersson

custom pipelines are all about reliable cache formats (ok not all , but a big part of them) and keyframe formats , everything glued with a smart db api.
I’ll keep the finger crossed.

“The Alembic initiative shows again the leadership and the vision of ILM and Imageworks. As visual effects and moviemaking processes are changing rapidly, Alembic enables increased levels collaboration throughout production. This will help the entire industry implement modern pipelines for the creation of digital entertainment. Autodesk is thrilled to support this initiative.”
Marc Petit – Senior Vice President, Autodesk Media & Entertainment

“We’re very excited by the possibilities of Alembic. Interchange formats need to be open and non-proprietary to achieve adoption, and innovations that fit these criteria but also come out of genuine production experience are what the industry needs. These days no software vendor or facility can be an island, and Alembic is a wonderful bridging idea that will remove another unnecessary hurdle in the creative process.”
Bill Collis – CEO, The Foundry

The Blender Foundation

I would donate $ for a fundraiser to get an Alembic i/o in native code, not plugin.

  • they have documentation ^^

that’s always good for a developer. in the video there’s some nice performance comparing to OBJ.

filesize: (testscene)

OBJ: 87Gb
alembic : 173Mb

write-time: 217 frames of testscene
OBJ: 95.0 min
alembic: 3.5min

read-time : 1 frame
OBJ : 65.0s
alembic: 0.2s

now OBJ aint the best format, the best format Lucasart has inhouse is PBC

PBC : 1262Mb
alembic : 353Mb

I have been using OBJ for years, all other file formats are hit and miss, between scene scale issues, empty UV maps, or extra crap coming in with the Geo (lights, extra nodes, god knows what else)…
Alembic sounds like a really great idea, which sidesteps the pitfalls of entire scene file formats like FBX and Collada.

I would definitely donate to an implementation of it in Blender. Also, Ton apparently met with the Alembic devs at Siggraph, at least i’m pretty sure I saw him mention that on Twitter… so perhaps the Foundation will get behind it before too long :slight_smile:

I’d also quite happily donate for this as well.

After watching that video I think it’ll be one of those ‘essential’ features that Blender will need in order to work in a larger pipeline that makes use of more than just Blender and/ or the import/ exporting of simple, and small models/ animations.

  • they have documentation ^^

yep , I’m quite surprised too :wink:

OBJ has worked only because its so old. Collada seems to work best when you set in world settings it to user metric instead of blender units. I always wondered why my collada stuf in blender was soooo tiny when imported with collada into apps.

small stuf like that I hope gets fixed from the get go. and somekind of indexation of animations, not just timeline frames. that was helpful with collada to be able via scripts to play animations simply. good for game making and on sites etc.

I’m coming from a web-perspective into this.

So I’m hoping this can be used not only to exchange from SW to SW, but also as import format for games/apps/sites.

mhh not sure I’m getting it…
Alembic is born with vfx/feature animation pipeline in mind… (ilm&sony) There are other better format (as collada) for other uses.

probably but we’ll see if alembic is so much better at reducing reading times (read-only almost only used in web-games for 3d assets) and can store animations etc there’ll probably be parsers for it.

I wonder if we could start a fund-raiser site w/o a developer? or is it money wasted…
you know what, someone should mention it to Ton, they’re getting around $700 a month in monthly donations now, how about spending some $100 a month for a dev to put time into getting alembic i/o

It seems they have quite clear ideas in which context it has to be used :wink:
(sorry for reposting the whole content… )

A found raise seems a good idea to me,
but I’d like to understand if the blender foundation sees it as a priority or not.



  • …Is a data representation scheme for storing computer graphics scenes

  • …Distills the results of artist disciplines for handoff to other artists in other disciplines

  • …Is focused on the greatest common divisor between applications, the ‘periodic table of cg primitives’

  • …Is extensible to support new workflows and new tools

    Alembic Is Not…

  • …A dependency graph, nor a procedural data transformation tool

  • …A replacement for native application scene file formats

  • …An asset management application

  • …A general rigging storage solution

    Alembic Would Be Used…

  • …To bake the results of an animated scene for hand-off to lighting & rendering

  • …To hand off an animated creature for cloth or flesh simulation

  • …To store the results of a cloth or flesh simulation for use in lighting & rendering

  • …To hand off animated geometry to a physical simulation engine

  • …To store the results of a physical simulation engine for use in lighting & rendering

    Alembic Would Not Be Used…

  • …To transport complex procedural animation rigs between different applications

  • …To make lossless round trips out of and into the same computation context

  • …To construct complex networks of procedural tools

oh there you go :slight_smile: so it’s not a collada slayer.

I would also donate for this. You can’t really compare Collada and Alembic as they are two completely different things.


To everyone who likes to organize a pay-per-feature or a more generic “open source kick-starter” initiative; please try to get it organized? I’ve asked the ‘foundations mailing list’ (where 100+ boards of OS foundations are member), and this idea unanimously gets rejected as not-feasible or unwanted everywhere…

  • Ton


I guess we could try and find a dev just to look at alembic specs etc, estimate how long it takes to code it. from that we have a sum to fundraise to.

Hmm… Seems cool. Do you think this could provide one standard format solution to selling 3d models online? Exporting models and verifying they open up as expected in other 3d apps is always a pain. :frowning:

Well, If we don’t have any luck finding a dev who would be up for implementing this as a solo project, just be sure to donate and/or subscribe to the Foundation and definitely let them know this is a popular request. I’m sure if there is enough interest, the Foundation would get a few heads on it.

If there is one thing Blender does great, it’s IO formats, and really great implementations at that.