Hello all,

Just been doing the tutorial on optimisation on the GECK wiki.

I followed it well enough, reading it definitely makes it sound worse than it is hehe but now I know how it is done it actually seems fairly straightforward.

Being the curious person that I am I wanted to look at how Beth had set up room markers in their content and now I'm more curious because there seems to be no rhyme or reason to their approach on this.

There are sizable locations with absolutely no room markers at all and others with the entire cell covered only by 2 room markers.

Where only 2 room markers are used presumably this still helps by only rendering objects in the 2nd room marker when the portal is in line of sight? That's how I understood the description from the tutorial, is anyone able to confirm this?

I'm assuming some of the inconsistancies in gamesas's approach when making new cells are simply down to some of the devs do it, others don't. Regardless, if making a mod it seems it would be better to make use of room markers/portals just out of good practice, what are your thoughts on this.

Thanks

In my vault mod, I only use room markers in one of my interior cells. But that is because there are locker rooms with hundreds of lockers in each, which caused some slow-down when the player turned in their direction. The room markers fixed this by not drawing the locker room unless the player was near/n the locker rooms.

These are really only useful if you have large cells, or you want slow/weak computer players to have a better time playing your mod. Thats my 2 caps worth.

I cannot speak for why Bethesda picked what-should-have-roombounds vs. what-should-not. I know that in the case of a number of my own maps, I feel like I just look, and know. Like... http://www.finhosting.fi/~fallout/screenshots/megacell-roombounds.jpg .

My sense is - if someone's CPU might start to get stressed over like, anything, including lots of detailed textures in use or, lots of followers, or maybe lots of combat and explosions, its good to make the rendering side of it be easier. Or in the case of my screenshot map - - a deliberate, uncommonly large, single-cell sort of build.

Grouping rooms into a single roombound is, I think, exactly as you perceive it. The goal with putting in roombounds is to reduce the rendering workload, not to necessarily be a real room-divider sort of deal.

Yea, pretty much what WillieSea and Tarrant said. In my mod, I have several medium sized cells at larged so far, and I will only be room-markering the most cluttered/destroyed ones. Like one is an office, theres mostly static desks, and small clutter like pencils, and not much rubble. This I will not worry about, as it should play fine on most computers. But another cells, is the lobby to a company, and has endured 200 years of people fighting in it, and weathering, and not to mention the radio tower that fell through the roof. It has like 300 rubble piles, and then tens if not hundreds of flipped desks, and sooo many items strewn about. This might cause an issue for the people playing it, so I will roombound it. But even splitting the map in half like Bethesda did someplaces helps. Better to have half the stuff at most times, and only all of it in certain places. Plus this probably saved them time. I'm sure they had spent 14 hrs working on the cell, and were tired, so they just made two huge bounds.

Thanks for the answers

I guess it's all a matter of judgment but I've a few indicators from here with what to look for, so much appreciated :foodndrink:

btw that's one hell of a cell Tarrant, I think I saw that screenshot (or one very similar) in another of your threads, always wows me at how massive you've made that place

I had a multicelled version (basically for testing navmesh) and decided that I hated it. When it is one cell, it really feels like one facility. Especially where followers are involved and they are doing their thing within the place (such as running to intercoms and sleeping and so on).