dan2
(dan)
December 16, 2019, 7:58pm
103
ah thanks! that’s what it was only remembered the experimental section, why is that off by default?
SteffenD
(Steffen Dünner)
December 16, 2019, 7:59pm
104
Experimental options were “removed” from the preferences (or rather enabled by default if I understand this commit correctly):
https://developer.blender.org/rBafcdea3961c2cc9a63c2e8613f8bbfa0ba49f583
USD support isn’t in the buildbot builds yet because the libraries still need to be updated.
It’s still all there, that committ is about something else…
Yeah, it would be nice to have it on by default…
John4
(john cavanagh)
January 5, 2020, 1:10am
107
For Blender to have a good USD/Hydra integration it needs to change it’s core data structure to match it, otherwise it won’t benefit from it besides having an importer/exporter.
I wouldn’t put my hopes high on BF as they have taken a different direction from what the rest are doing and there is no easy way to trackback.
Blender wil probably be the last Software to get it in this new decade unless some Studios who are already using Blender in their production and need this tech decide to do it & push the development upstream, of course Blender Foundation has to agree to these fundamental changes before we can see anything, but i do hope that it would be well integrated like what Autodesk & SideFx have done so it can work with Eevee & Cycles and other external engines.
eobet
(eobet)
December 1, 2022, 8:34am
108
Anyone know if there’s any plans at all to be able to link to a USD file in Blender?
I’m talking about a workflow where the USD contains the geometry and you use this in Blender, Max, Omniverse, whatever but you never make duplicates of the data (possibly only add layers to it),
1 Like
Kologe
(Kologe)
December 1, 2022, 12:21pm
109
Short answer: It’s being discussed, but lacks a clear design yet, thus no tangible plans yet neither.
Long answer:
Pipeline Assets & I/O Meeting 22-09-08:
Brian, Sonny and Michael had a meeting last week to align further on a common design.
The ‘simple’ case of the Collection I/O project , adding import/export features to Collections in Blender, while still keeping all data manipulation in Blender area, is being owrked on by Sonny.
There was a long discussion over the ‘complex’ case of handling USD data (i.e. not simple import/export, but rather editing/authoring of USD data itself in Blender), with many open questions:
Do we need a special type of collections, or can we rather work with ‘compositing’ nodetree attached to a regular collection?
Would such a collection authoring tree only affect I/O, or should it be also actual part of Blender scene evaluation?
How can current Blender collection/collection instanciation system work with USD ‘collections’, e.g. when it comes to transformations? Do we need some neew type of data, ‘group of objects’?
Can we use a ‘simple’ layer-based UI/UX in the outliner for authoring of USD?
There is need more concrete plans and/or prototypes to better understand where to go, and what are exactly the issues that need to be fixed here.
[source]
Pipeline Assets & I/O Meeting 22-11-17:
Collections for Import/Export Design & USD
Blender is actively looking for a senior dev to work at Blender and coordinate and lead the whole USD project. We are also looking for some official suport from a big USD player (like a big studio, or the ASWF), and stakeholders (people or studios already using USD a lot, and willing to help define and test Blender’s integration of USD). Brian thinks that ASWF and their USD working group is probably the best way to find these.
Sonny is working on curve export (trying to bring Project Heist data into USD as test playground).
Brian and his team keep working on Hydra integration, they updated the design task , and expect to be ready for sharing they work publicly in a few weeks. Next step, once required library work is done (for Blender 3.5), would be to move Hydra integration into the main USD development branch .
Zhen asks is there is a roadmap for USD improvements in upcoming releases (Blender 3.4 and 3.5). Bastien answers that besides regular improvements of existing I/O code, there is no clear roadmap yet, as lack of resources at Blender are currently making progresses slow. It would be good though to get basics of Collection I/O finalized and ready for master though.
[source]
greetings, Kologe
3 Likes