FiftyTifty

I'm looking to get started with modelling for WoW 3.3.5a, and I haven't been able to find much coherent information about the limitations and usage of .M2 and .WMO. From what I can tell, .M2 files are individual objects that are small in size. .WMO Files are a combination of .M2 files. But when should a .WMO be used, over a .M2? Is there an object bounds limit for .M2 files that .WMO doesn't have? For example, if I wanted to replace the entirety of Stormwind, is there a reason why I shouldn't just make a single large .M2 in 3DS Max, and instead split the model into sections as .M2 files, plonk them into a .WMO, and then place the .WMO into an .ADT? If I can use a single .M2 for a city, that would be ideal, as I would then be able to save on a huge number of draw calls by just making the architecture one big model.

Aha! There we go. So a .WMO is essentially a collection of .M2 meshes, which you use to create the bulk of the scene. Is it possible to make the .WMO itself just one big mesh in 3DS Max? Or does it only take .M2 meshes as it's data?

Unfortunately, WoW suffers from having very mundane post processing. There's no bloom, the ambient occlusion in even the BFA client is extremely mild, and the lighting does the game's toon textures no justice. So what I want to do, is port a couple post processing shaders to the WoTLK 3.3.5a client. But rather than using ReShade, which uses the LDR 8 bit frame buffer output, I would like to try making shaders that use the 32 bit frame that the anemic in-game post processing uses. I did some googling, but I couldn't find any information at all about this. Are there any resources out there that I could use as a reference and/or guide, for editing WoW's shaders?