Sorry your browser is not supported!

To make a believable town or home, you need believable scenery. Never once have I walked down a street with just buildings and empty alleys, or a house someone lived in, with no furnishings. But if I try to make a level that really looks realistic, will it drag down my FPS considerably, or is it okay? Because I don't want to spend a lot of time on just one level when I can't even play it, nor do I want to make a subpar level, thats a waste of time.

I know how you feel - but FPSC doesn't do outdoor scenes believably like other engines...

This is probably the one aspect that FPSC can't be improved in drastically enough, and it's a major reason I moved onto the UDK after two years of using it...

I thought about UDK, but I'm not skilled in any programming language. I'm thinking about learning BASIC going to DarkBASIC Pro.

FPSC is plenty apt to make large outdoor areas if you understand optimization. use entity buildings instead of ones made from segments, and keep detail a little lower than in small cramped indoor areas, and you should be fine.

Thanks, I'll try that

Entity buildings are a poor way of optimising, because it can't be broken down into smaller chunks and rendered on demand...

In the perfect scenario, you should have lots of smaller models that slot together to make a bigger model, i.e., a house. Then rather than the entire house having to be rendered even when only a tiny bit is on screen, only the parts of that house that are visible need to be rendered. Les rendered models means better frame rate...

UDK doesn't need programming, I wanna know who started that rumour - but let's end the UDK talk there.

actually, entity buildings are a great way of optimizing. FPSC won't need to render chunks of a large, well textured box.

imageflock.com/img/1284341166.jpg[/img]

this scene is a perfect example of what outdoor stuff can be achieved in FPSC. it's level 8 of When Worlds Collide, my game. this is how it was done.

all of the buildings on the inside of the 3 detailed ones are entities.

the 3 detailed buildings began as large segment floors with rooftop stuff on them at a high layer.

there are some segment objects blocking the players view at times. this way still not everything is rendered at once.

detail is used sparingly, and only where it counts. attracting the players eye away from ugly, undetailed areas like street level, and the other buildings.

all of the outer walls of the detailed building roofs are large flat plains.

polycount rendered in that picture? 72000. FPSC was running at 40/40 FPS in v116 with fenix mod.

Yeah, let's end it but with a side note:
UDK does require programming if you plan on a commercial release, because none of the included media is allowed to be used.
Therefore, you need to create your own classes and methods for your objects if you can't use theirs, right?
So, unless they have changed the license; UDK does require programming for commercial releases.

I have an outdoor scene in FPSC that takes the whole map and runs it up to the 1.8 gig limit, and it runs fine. (in the 50's for fps)
The only thing I did was to make sure everything was low poly (billboard trees, etc) and limited the number of characters spawned at any one time.

I've attached a sceen shot of my big map, but only one character was spawned at the time.
This is just the graveyard. (the farm is on the other side of the fence with a segmented farmhouse you can enter and an entity shed as well as a field with a crop)
I am going to break the map in half for two levels though instead of one, so I can add more characters and stuff.
It is my cutscenes that have me at the media limit, so that is the real reason for having to split it in half.

If you want heavily populated areas, then I suggest making a lot of corners between buildings etc, instead of straight streets.
That way, the player won't be able to see very far so the characters can be spawned only in the area the player is in.
In other words, have zones that spawn characters when the player enters and then destroys them when they leave.
The player will never know the characters aren't there since they can't see them anyway.
Of course, the zones would actually spawn for areas to be enterd, or adjoining areas, so the player would never see it happen.


So, if you had five zones...

1 - 2 - 3 - 4 - 5

Player start zone and the adjoing zones could be populated at start.

entering zone 1 would populate 2 and destroy 3

entering zone 2 would populate 1 and 3 and destroy 4

entering zone 3 would populate 2 and 4 and destroy 1 and 5

entering zone 4 would populate 3 and 5 and destroy 2

entering zone 5 would populate 4 and destroy 3

Population would have a check to see if it is already populated before spawning, so you wouldn't spawn them twice.
Then your direction you are traveling won't matter.
The zone they are in is already populated from the zone you entered prior to entering that zone, and adjoing zones are populated for the limited view.

It is a make-shift LOD for the characters you could say.
It could be done easily enough be settargetname=x and activatetarget=x for the zone's scripts, and then the characters would be scripted to destroy when activated to do so.
This is basically the same as spawning enemies, except you are spawning these bystanders before the player gets near them as opposed to spawning the enemy right there.

This example basically gives you 60% of your characters spawned at any one time if the zones are evenly populated.
I haven't tried or tested this method before, but it should work.

If you had 10 zones as opposed to 5 in the example, then only 3 of the 10 (30%) would be spawned at any one time.
That would make a big difference as opposed to having them all spawned all of the time.

Sorry, my connection's too unstable to see the pic right now. Anyone in Britain planning to change ISP, do NOT get Sky, they are terrible and it will randomly drop off with no warning, at any time of the day/night...

I'm personally not planning any commercial releases though. Not any self motivated ones at least. Between college and social life, I have too little time to work on it myself, no money to hire the help I need for it, and I simply want to learn the engine because it's already landed me a position helping out a small team hoping to create a certain sequel to a certain FPSC legend...

Fair enough. I'll never make my game dream come true. But I suffer from creative diarrhoea, and ideas for games come and go every week for me, adapting, warping, getting trashed completely. Every one is a learning experience in a new sector of the editor, and I don't feel a single concept is wasted if it helps me get more knowledge of it...