hcornea
Resource contributor
- Messages
- 2,388
- Country

I think I suggested this a long time ago, but perhaps with the evolution of ModelconverterX it may be more feasible.
To correctly batch, materials must be identical ... so across a library bgl, it would be a good idea to identify discrepancies between material settings (and correct them!)
A report function that identified the unique mateials and grouped them would make it possible to identify "rogue" materials that were meant to be the same.
A flagged material would have the same diffuse texture .... but at least one different material setting.
So (for example) ... similar to the "object" report that Modelconverter produces, this would be a report sorted by unique material
An output for part of a library might look something like this
Material: VHHH_Fence_trans.dds
Terminal_withFence.mdl 128 tvert
Peripheral_Hangar.mdl 128 tvert
Fence_4m.mdl 16 tvert
Fence_15m.mdl 16 tvert
Fence_64m.mdl 16 tvert
Fence_128m.mdl 16 tvert
Fence_256m.mdl 16 tvert
Material: VHHH_Fence_trans.dds ** duplicate use of texture
Fence_32m.mdl 16 tvert
In this example, the rogue material is in the Fence_32m model and is producing an additional (unnecessary) drawcall.
A simple ModelconverterX examine + edit (or maybe a "get material from xxx.mdl" function! -- another suggestion, that would be cool!) would fix the issue.
Alternatively, back to Max/gmax/TweakFSDS and fix the source file.
Thus, a library.bgl could be checked for efficiency
(please note the above example is a mock-up, and I am not working on Hong Kong international airport)
To correctly batch, materials must be identical ... so across a library bgl, it would be a good idea to identify discrepancies between material settings (and correct them!)
A report function that identified the unique mateials and grouped them would make it possible to identify "rogue" materials that were meant to be the same.
A flagged material would have the same diffuse texture .... but at least one different material setting.
So (for example) ... similar to the "object" report that Modelconverter produces, this would be a report sorted by unique material
An output for part of a library might look something like this
Material: VHHH_Fence_trans.dds
Terminal_withFence.mdl 128 tvert
Peripheral_Hangar.mdl 128 tvert
Fence_4m.mdl 16 tvert
Fence_15m.mdl 16 tvert
Fence_64m.mdl 16 tvert
Fence_128m.mdl 16 tvert
Fence_256m.mdl 16 tvert
Material: VHHH_Fence_trans.dds ** duplicate use of texture
Fence_32m.mdl 16 tvert
In this example, the rogue material is in the Fence_32m model and is producing an additional (unnecessary) drawcall.
A simple ModelconverterX examine + edit (or maybe a "get material from xxx.mdl" function! -- another suggestion, that would be cool!) would fix the issue.
Alternatively, back to Max/gmax/TweakFSDS and fix the source file.
Thus, a library.bgl could be checked for efficiency
(please note the above example is a mock-up, and I am not working on Hong Kong international airport)
