DescentBB

Over a year I started to create a single player mission featuring an outdoor enviroment with a large canyon and a skybox. You could see a small part of that level in my D2X-XL video clip. Recently I have picked up work on that mission again. Here's a few pics:

Improved environment box, added bots:



A new, subterranean part of the mine:

That looks really cool!
Whats the cube count sofar?





Grrrr....
That reminds me......
Looking at that lvl makes me think that I really need to finish my Monsterball level!

About 2600 cubes. I added equipment generators to D2X-XL and DLE-XP recently, and have placed some in that level - you'll need them.

I have some really cool ideas for that level, but it will take me a while just to implement them, not to speak about polishing everything.

More pics ... this sure is a loooooooooooooooooooong chimney:

The inside looks nice, the outside needs retexturing or something.

I don't think so. And a screenshot never looks as good as the actual game screen.

Maybe it's the lighting. D1/2 doesn't do so well at fullbright.

It's an outside area, you know. It has rocks, and sun light, and light from lava ... and, errrm, rocks. Lots of rocks.

Don't Suck.

I don't know, the fullbright bothers me in the first shot. But between the lava and the sky, that's kind of what you'd expect from Descent. I don't know. I just wish there would be some kind of shadow between the sky and the lava along the outer walls.

That is a really long chimney.

Methinks someone should hack the old D1 \"escape sequence\" code back into D2. Actually, I wonder if that's actually in the code. It was there for the D2Demo...

It is not the exit, and the chimney is so long to give the players some headache when they try to pass it ...

Did I give away too much already?

I don't think the exterior area on the pic looks so bad, but I know it from inside the game.

Anyway, what you are seeing on the screenies are only pretty small parts of the entire level.

I repeat my above statement then. I have high expectations for this.

The entire project has about 7000 segments so far, and I estimate that I will need at least another 1000 to finish it. This is due to the size of the level as well as the degree of detail I have put into it. I have created several separate parts of the mission I had initially planned to put together in the end. Theoretically I could allow for D2X-XL and DLE-XP to handle 8000 or even 10000 segments, but this would be rather impractical, as levels that large will cause extremely long loading times and probably run very sluggish on slower computers. The final mission will therefore consist of several levels logically linked together. This will fit nicely into the intended flow of the entire mission, and an advantage of that approach is that I can insert briefings for each part of the mission that will focus on the particular part of the whole and challenge it offers. Other than almost all other multi-level missions (at least those I know), the levels will directly geometrically connected, so they will offer a playing experience as seamless as possible. A slight disadvantage is that I will have to make the level transitions work with Descent's rather historical end level - fresh start in next level scheme, and cannot have 'global' events (like a switch thrown in one part of the level changing something far away). This isn't hard to work around though. Most of what I have built so far has turned out about what I had planned it to be, and I think I have quite a few good looking or even impressive structures, so I am pretty satisfied. There's still about 1.5 to 2 areas of the level to be built and a lot of detail work to be done.

...oh dear.

I'd better be careful then.

3000 to 4000 segs should still be bearable. Speed in game depends greatly on the level structure - a single room with high detail and lots of visible rooms can slow down things, while a huge level with lots of less detailled rooms should work fine, as long as they aren't all visible.

You can considerably decrease loading times by using '-enable_shadows 0' and turning off OpenGL style lighting (which you do not necessarily need for level having colored lighting already anyway).

What's the limit for a single level - and do you plan to change it? I've just hit 4000 segments...

6000. But if I consider how long you are already working on your level I wonder whether I should turn it back to 4000 ... or you will never finish it (well, lets add a room here, and how about some detail there ... I wonder whether a new area between these two places would look good ... )

Since I have split my mission into several parts, the biggest level has around 3500 segments or so.

lol... Yeah I understand what you mean - most of the problem for me has been not enough straight time spent working on it, usually caused by having something else to do or not really being sure what to do next (or both)...

6000 should be enough, but I will have to stop thinking about whether a D3-like part of the level would fit or work or whatever because I don't think I'd have room for it.

When I build a level, I usually wait for some cool idea to pop up in my imagination. If I get one, I build a level just containing that structure. Once I have a few, I try to put them together in some logical way. Often the requirement for a plausible transition spawns new ideas or makes me create nice details. I am still having unimplemented ideas from two or three years ago in my mind I will now put into this new mission of mine. Some other components were resting on my HD for a year or so, with me just having the intention to get them into the mission, but not being sure how to do it.

Geez, if this was an up-to-date game engine with a really good level editor, this mission would blow everything away you've ever seen ... I really really hope CD will see the light of day and the tools will be as good as has been rumored.

An update:

Geometry and texturing are complete (maybe a few minor tweaks required, but no big efforts any more). The entire mission has 5 distinct levels with a total of about 13.350 segments. Doors, walls and triggers as well as robot and equipment makers are in place. The briefing including images is done and just needs some minor polishing (screen layout). Most robots have been placed, too. I also have taken Testi's criticism about full bright areas back to the levels in question and tried to bring some more shadows in. All that is left to be done is some play testing, making sure there's enough stuff to suffiencently equip all players in coop games, and ... publish the mission.

I hope I have been able to create a few areas that will make your jaws drop. The screen shots do not give you more than a taste: The real good stuff has been kept secret.

Some people at the Descentforum.de are claiming my outdoor areas were pathetic and that they had seen stuff like this done better in other Descent 1/2 levels ... ... these guys are totally clueless. I doubt there is a good Descent 2 mission I don't know, and none of them comes even close to what I have built. There's just a tiny little more detail I can create with a segment limit of 6000, hires texture support and environment boxes ... Ofc, this is not the Unreal 3 engine. But for what it is I think I have achieved something.

You will see.

There have been very VERY few levels in D1 or D2 with outside simulated areas. Mostly due to the systme hit when it was done. With todays machines that's not an issue. On top of that, I can't recall one that I've seen that was very convicing or even attractive.

Likewise; I've seen \"outdoor\" areas, but they all looked even worse than that. The stuff in those screenshots at least holds up to Unreal (1)'s calibre...

Sirius wrote:... they all looked even worse than that.

Ouch - now that was a compliment ...

I kind of think Unreal 1 is all you can hope for in the Descent engine. Doing any kind of outdoors is challenging to say the least in these types of engines.

Diedel wrote:

Sirius wrote:... they all looked even worse than that.

Ouch - now that was a compliment ...

I think this is called a left handed compliment.

Well, Diedel - I'd be kind of pushing it to call it \"good\", and lying to call it \"great\" - but I don't hold it against you, because I'd be shocked to see anyone do any impressive outdoor scenes in the Descent 2 engine. It's not designed for it, and doesn't have the textures or lighting for it.

Suffice it to say, it's at least creative and realistic enough not to make it look like you shouldn't have bothered - which is more than I can say for the \"deformed square-celled grid\" approach I've seen before.

Sirius wrote:Suffice it to say, it's at least creative and realistic enough not to make it look like you shouldn't have bothered - which is more than I can say for the "deformed square-celled grid" approach I've seen before.

Yeah.
Its alot better looking now then before.

I've play tested the first part of the mission and the outdoor look is awesome!

If you want to playtest it?
Sign up here.

A still image just doesn't do the level justice. It looks completely different when you're moving and changing perspective in it.

Yeah, I'm definitely going to play through it once it's done (unless you really need more playtesters ). Outside is an interesting touch; inside looks pretty cool in general (puts some third-party D3 levels to shame, which is kind of funny).

If I don't start to hear something from my level testers soon, I think I may actually need someone who is more up to the job ...

wow, we have lives. But seriously, once this week is over, I'll have gobs of free time to blast through this.

Who was talking to you? You aren't a level tester so far, and you shouldn't have this mission.

Fact is that the level testers were eager to get this mission. Now the weekend is over, and I haven't gotten one useful feedback except a D2X-XL bug report from Aus-RED-5 from them.

true, I'm not, but I've expressed extreme interest to be one. So not sure why you're trying to insult me that way. Furthermore, I was saying that has it occured to you that these people may have 'lives' and don't have 24/7 to work on your mission for testing. Just be patient, good things happen to those who wait. And I did state earlier that once this week was over a good chunk of time would open up for me should you be willing to have me as a tester.

Is it done yet?

Why don't you take a look at my web site and check?

Yes, it's done, playtested, fixed, tweaked and ready for playing.