Difference between revisions of "Discussion Model System Overhaul"
(Imported from GH wiki) |
m (Category:Discussions_2012 + Archived) |
||
| Line 1: | Line 1: | ||
| + | [[Category:Discussions_2012]]{{Archived}} | ||
| + | ---- | ||
| + | |||
While LMR (the current Lua modelling system) is quite flexible, it is difficult to use correctly, particularly for people who do not have programming experience. LMR also has performance problems, which can only be partially mitigated through pre-processing and caching. There is a general consensus that we should move away from LMR for rendering normal models (that is, models created by artists, such as those for ships and buildings). | While LMR (the current Lua modelling system) is quite flexible, it is difficult to use correctly, particularly for people who do not have programming experience. LMR also has performance problems, which can only be partially mitigated through pre-processing and caching. There is a general consensus that we should move away from LMR for rendering normal models (that is, models created by artists, such as those for ships and buildings). | ||
The intention is to develop a more typical asset processing pipeline for game models, with models being created in a normal 3D modelling application such as Blender, and being exported (along with equipment attachment point locations, animations and so on) to some intermediate format (possibly [https://collada.org/mediawiki/index.php/COLLADA_-_Digital_Asset_and_FX_Exchange_Schema COLLADA]) and then processed by a command line tool to generate an efficient bianry model format. Rendering will become simpler and gain support for a standardised set of materials. LMR may be retained as a mechanism for procedural modelling (e.g., generating randomised buildings) | The intention is to develop a more typical asset processing pipeline for game models, with models being created in a normal 3D modelling application such as Blender, and being exported (along with equipment attachment point locations, animations and so on) to some intermediate format (possibly [https://collada.org/mediawiki/index.php/COLLADA_-_Digital_Asset_and_FX_Exchange_Schema COLLADA]) and then processed by a command line tool to generate an efficient bianry model format. Rendering will become simpler and gain support for a standardised set of materials. LMR may be retained as a mechanism for procedural modelling (e.g., generating randomised buildings) | ||
Latest revision as of 20:42, 3 January 2014
While LMR (the current Lua modelling system) is quite flexible, it is difficult to use correctly, particularly for people who do not have programming experience. LMR also has performance problems, which can only be partially mitigated through pre-processing and caching. There is a general consensus that we should move away from LMR for rendering normal models (that is, models created by artists, such as those for ships and buildings).
The intention is to develop a more typical asset processing pipeline for game models, with models being created in a normal 3D modelling application such as Blender, and being exported (along with equipment attachment point locations, animations and so on) to some intermediate format (possibly COLLADA) and then processed by a command line tool to generate an efficient bianry model format. Rendering will become simpler and gain support for a standardised set of materials. LMR may be retained as a mechanism for procedural modelling (e.g., generating randomised buildings)
