Hey all , i need some help with some questions regarding SP maps , i hope someone can answer or atleast point me in the right direction.

Recently like i said in the last thread , i have not used CEs engine in a year , or more due to having to work in Unreal Engine for Uni , that exploit can be seen in this latest dev video , and yes it is related as ...

1: Its Sonic
2: I was (and still am) concidering porting it back to CE , as the original was supposed to be CE content.

http://www.youtube.com/watch?v=iqA0ABkm2o0

^Not all of the map is shown in that one video.

Anyways , i found myself getting some creativity back into my head and starting to draw ideas for the SP Sonic mod I was working on with my team , and i need some questions regarding the map size and stuff.

As most probably know , apon opening sapien and loading a SP level from normal Halo , the level is cut into sections , take A30 for example , i think if i remember rightly thats split into 2 parts , A30 A and A30 B or something if i remember rightly anyways heres my questions related to that.

Question 1 : When making a level , take for example A30 , does each part of the map have to be modeled seperately or can it be one mesh that the engine splits into 2 sections.Also , if the answer to this is no , then does it require the BSP mesh pieces to be compiled into the map similar to permutation models so that it can load them? , i dont think its the correct understanding , but when you look at the map tags , it obviously says A30 only , no subsections are mentioned in the maps file , but they are in sapien.

Question 2 : Can a Single Player level have a higher poly count than recommended for a multiplayer level , i know this will most likely be yes because of the details bungie had in the original campaign.

Hopefully , someone can answer these questions , which will help me remember everything i knew about modding CE.

question1: im not sure.... i think it should be diferent meshes but into 1 gbx model thing.....im nor really sure

question2: yeah they can have more, specially when its played on the local network and with our high-end pc´s far supperior to 8 year old ones

It has to be modeled separtly, but each bsp needs to overlap at the end to create a zone wher ethe bsp can switch. The smaller the BSP chunks, the less tagspace it takes up.

As far as polycount, peopel these days have much more modern computers than in 2004, so theres no reason not to push the original limitations.

yes different parts of the map are different meshes. you compile each separately. if i were you, i wouldn't go above 200k tris for any bsp if you can portal well
if you cant, stay below 100k