Page 2 of 8

Re: OpenGEX as .Nif alternative for OpenMW

Posted: 26 Aug 2016, 07:38
by Methelina
Chris wrote:AFAIK, all we'd need is an OSG loader for OpenGEX files. Anything OSG can load, OpenMW should be able to use.

What will be more difficult is things like animation, since the engine relies on custom data being supplied in certain nodes of the model. With NIF, these are done with specific Node types and extra data containers stored in the nodes. I don't know enough about how OSG (or OpenGEX) is designed to know how easy it is work with that. I know there was some talk before about needing to have custom node types to support particles and animation in .osg files, which would obviously cause a problem with modeling software that doesn't know about those node types (which thusly can't create particles or animations, kinda defeating the purpose of supporting them in the format).
The import code for the Ogex looks easy to implement OpenGex-Import.zip into OSG interpreter.
Anyway, i think the basic feature of the .nif - node system with unique flags and tags are disadvantage, if we are look into the universal and wide supportable future of the engine. Where all important features are stick to the long abandoned plugins in a third-party software with need to edit in another third-party software before the Game Engine. If we will look into the Unity3d for example, it have a importing menu for the raw data. This Import menu used to set up the unique flags and tags, names for the animation sequences and materials, same as we did in .nif. In this case we have a great advantage - we can use any kind of RAW data - mesh data or animation data and than we can set the needed settings and the tags inside the Engine, where Any software can be used to create the assets and then exported it in the unified raw data, such a OpenGEX and than natively load via engine.
In same way will be easy to implement other formats if it will be needed. For example .OSG with animation (in future) can be loaded in same way with the Import Settings in the middle and will have same features as .nif can have.

Re: OpenGEX as .Nif alternative for OpenMW

Posted: 26 Aug 2016, 21:46
by Pherim
Sounds like a great idea!

Re: OpenGEX as .Nif alternative for OpenMW

Posted: 28 Aug 2016, 19:04
by Berandas
The more simplier way of getting the asset into the OpenMW the better. Let's finally get rid of the slow and rigid Nif production pipeline.

Re: OpenGEX as .Nif alternative for OpenMW

Posted: 31 Aug 2016, 14:41
by Methelina
Yeah, the time is come....
the Great Developers, hear our cry! XD

Re: OpenGEX as .Nif alternative for OpenMW

Posted: 31 Aug 2016, 15:37
by MiroslavXO
We need to cast some spell influence.

Re: OpenGEX as .Nif alternative for OpenMW

Posted: 31 Aug 2016, 17:52
by Methelina
Or may be we need a make a human sacrifice?

Re: OpenGEX as .Nif alternative for OpenMW

Posted: 31 Aug 2016, 18:11
by DestinedToDie
A slave tasked with writing the code should be sufficient. Otherwise you canĀ“t really rush the developers if there is no one willing to take the task at this moment.

Re: OpenGEX as .Nif alternative for OpenMW

Posted: 06 Jun 2017, 09:16
by psi29a
https://www.phoronix.com/scan.php?page= ... 0-Released
The Khronos Group has released the ratified glTF 2.0 specification this morning as their transmission format for portable 3D assets. glTF 2.0 can be integrated now not only within Khronos graphics API using applications/games but also within Direct3D and Metal.

The glTF 2.0 update is big for the real-time delivery of 3D assets. The big change with glTF 2.0 is support for Physically Based Rendering (PBR) rather than being tied to GLSL shaders. glTF 2.0 still has an extension for supporting OpenGL shaders, but by using the new PBR approach, glTF models can now be imported into Direct3D or Metal applications, or basically any graphics rendering API with now not relying upon GLSL shaders. PBR also allows for higher-quality materials.

Other improvements beyond PBR definitions and making it graphics API neutral is support for morph targets and improvements to binary glTF and other performance work.

Among the items Khronos will be exploring for future glTF revisions are mesh compression, progressive geometry streaming, point clouds, a lighting extension, and extensions for other APIs and language specifics. For mesh compression they are looking at Google's Draco work, for geometry streaming is Fraunhover SRC, and for the compression they are looking at the Basis format from Binomial.

Along with the updated specification, Khronos will be publishing their source-based validation tool for the 2.0 specification and other open-source tooling updates. The glTF format for distributing 3D assets is quite exciting and continuing to be more broadly supported, even with Microsoft planning to make use of glTF in future 3D creation programs. Oculus, NVIDIA, Adobe, Autodesk, Google, and Blender are among other organizations employing glTF.
Looks like this is the way forward and there is Blender and Autodesk(Maya/3DSMax) support, it should only be a matter of time before an OSG plugin is there to support this.

Here is the spec:
https://github.com/KhronosGroup/glTF/tr ... cation/2.0

Oh... looky here:

https://github.com/KhronosGroup/glTF/tr ... animations
Animations

glTF supports articulated and skinned animation via key frame animations of nodes' transforms. Key frame data is stored in buffers and referenced in animations using accessors. glTF 2.0 also supports animation of instantiated Morph Targets in a similar fashion.

Re: OpenGEX as .Nif alternative for OpenMW

Posted: 06 Jun 2017, 09:23
by psi29a
There seems to already be a OSG plugin... but for something called OSG.js
https://github.com/cedricpinson/osgjs/b ... terGLTF.js

Re: OpenGEX as .Nif alternative for OpenMW

Posted: 14 Aug 2017, 10:04
by Br0ken