Sorry your browser is not supported!

Well I'm soon to order DB Pro to try my hand at game making. Right now I've begun modelling, and working on the concept. I'm wanting to do a futuristic shoot'em up RPG. The modelling so far is going great, and I'm catching on quick.

Anyway just trying to decide the best way to go about things as I'm modelling. Should I go for lot's of content and a full dynamic world. Where models will have to be kept on the decent, but nothing fancy side of things. Or kick tail models and vastly reduce the number of objects that can be around at one time without bombing a decent system.

For example I can spend an hour or so and 3D model an AK-1000 (future version) so that when I render it looks surreal. Or I can spend about 10 minutes modelling a rough model, but then photo texture it. Where when rendered it looks like what it is, but nor surreal. Yet ingame would look decent, and really light on the polycount.

Because I mean the character would have so many objects just on his person. Plus all the other characters he see's on the streets, and in places. Not to mention the streets themselves, buildings, neon signs, etc.

Basically I've got a big notepad and I'm just doodling in idea's for what I want. For example a character can't just drop an AK-1000 in his inventory. He'll have to carry it, or sling it over his shoulder. Or if he's wearing a really big trench coat he could sling it under there. Yet if it has a folding stock then he could drop it in a briefcase, backpack, or under a jacket. He'll have his car keys in a pocket, maybe a pack of smokes, and a lighter. Basically I want a realistic inventory system. None of that encumbrance stuff. Like Morrowind where I'm walking around carrying 3 spears, and they magically fit into his pocket or something.

I also want an investigative AI that will go behind the player. Say he commits a crime. If he gets out before the police arrive or gets away "clean". Then the crime is still reported, and the AI will gather all evidence from the scene. Such as suveilance video, casings, blood, fingerprints, etc. So if he leaves too much evidence the cops will be looking for him. Or if he busted for something else small they will get him for the other crime.

There will be more than crime in the game though. The player will be able to have a paying job. Ok I'm not going to type it all in here as I'm quickly filling a notebook up. 50% of it will probably never make it in game.

I just wanted people's opinion on content vs eye candy. I like to use Morrowind as an example as it's like a medival version of what I want to do (although it's much simpler). For example if you get more than 40 AI roving around at any time it bogs the machine down. Less if they're actually doing something like fighting. Yet the game graphically is pretty great. Yet the average weapon in Morrowing is 500 poly's roughly. Now I can model and skin a weapon for morrowind that looks like it's 1,000 poly's. Yet it's only 198 poly's. It's mainly due to the texturing which lies to the eye. You see curves when the model has none, etc. Yet the more complex the object say a person vs a sword. I can still keep the poly's low but if you get close you can tell. Although to me it still looks good. I've modelled a good bit for Morrowing and there's a TOC coming out that will contain a good number of my models.

Anyway I'm rambling again as I do. Anyway just looking for opinions. Seems some people care fore eye candy more than they do content, and others are the opposite.
I'm just not really sure yet how many poly's DB Pro can really handle of a decent machine without going into slow motion.

well, the difference between 198 and a 500 poly model lies in the lighting. Besides, morrowind had pretty crappy resolution textures for the weapons and the objects in general. Or just about everything. I wish they would have released a highres replacement package for the current textures. They look crappy. There are few individual objects that are textured nicely, but most of the detail is still in the polygons.

and in morrowind you could carry a lot more than just 3 spears, I could carry hundreds of those. And I was a wood elf. (it wasn't so hard to get strenght to something like 110, even if I played a sneaky thief who ventured into archery and alchemy...or almost anything really.)

Also, I think current top-end cards prefer more polygons rather than bigger textures. But that doesn't mean that you couldn't use good textures. Also, Normal map texturing is worth checking out. (using hi-poly model to generate a normal map that affects the lighting of the lowpoly model.)

Sounds eerily like the idea I posted in one of Las6's threads. It's cool though, my idea has been living in my head for the last 20 years, I've practically went through every facet already .

Personally, I like to see a lot of characters on screen at once than a few high poly characters, it's the immersion factor that you want more than gorgeous models. Perhaps you could have a few key character models that you spend more time on, like maybe a cute girl model, so you'd notice a cute girl in the crowd of low poly models.


Van-B

That sounds like a hefty project.... is it your first?

he inventory could be sorted in a similar way to games like Commandos 2, System Shock 2, Deus Ex(I think) and various others - Have a grid inventory, and each object has a set amount of space(An assault rifle, for example, could take up 3 grid squares horizontally). Therefore you can only take an item if you have enough room - it sort of turns into a mini puzzle trying to rearrange items so they all fit. You could then change that grid say, if the player buys a bag or a long coat, add a few more squares to the grid.

Definatly build your game around content. 100% The graphics will be welcomed but not needed if the content is solid. Make your characters look good, don't get me wrong, but keep the detail on the spots that will get the most attention from the eye. Like on the cute girl model suggested by Van B, the attention will probably be on hair, breasts and legs. Keep the poly count low, but maybe give a little more poly budget to those areas.

If your main character were to have hair that was a defining feature for the charater (cloud from ff7) then spare some polys there. But keep the bottom of the feet as few polys as possible.

If you are making the game from a 3rd person perspective, you might want to limit how close the camera can get to other actor models. If you're running the game at 800x600 and the model won't get close enough to take up a quarter of the screen, It won't have to be as detailed as if were running it at 1024x768. Also if the character has cigarettes that he pulls out of his pocket, you won't have to make the texture any bigger than what it will show up on screen. If you're running at 800x600 and the cigarettes only take up 50x75 pixels, the texture won't need to be any larger than that (64x64) just to keep things simple. I'd rather see a semi smurfy looking guy with lots of thoughtful knick knacks than a detailed guy with just his gun out and a utility belt.

I don't know how well you can implement the AI portions of your game without using a dll. Basic languages and AI aren't friends. Just remember that CONTENT IS THE game.
Good luck

This will be my first attempt with Dark Basic Pro. However I am familier with 3D modelling. http://www.cuneo.us/tesmw/images/Swift22/scimitar.jpg
As well as the basic programming language. Visual basic, access, and scripting. I've written software code before yet it was data transfer software between a few networks. Well it did more than transfer data, but needless to say was not very graphic intensive. In fact it had no graphics. They clicked a button in an Access program, and a few moments later about 40 of their documents were updated country wide between 6 networks in factories.
I've written other stuff as well, but that was probably my biggest as it was about $25,000,000 worth of transactions.

Anyway for the AI part...well I shouldn't have said AI. I wouldn't actually try to make the AI walk around looking for clues.
More like where they crime was commited would be a sector (city block) on the map. When the cops showed up they would stand around, walk around looking busy etc. Yet a routine would run that would check the area for any player left objects. Depending on the distance of the evidence from the central crime scene would determine it's chance of being discovered. It would then drop the "evidence" data into a folder, and the evidence itself would be removed from the scene via script and put into an evidence container at the police department.
Then say a security camera looked dead at the player while commiting or in the crime scene area. The cops would then be "looking" for the player.
Thus if the player came across a cop he might be arrested. Then let's say he has in his inventory a weapon matching the caliber used in the crime. Along with something from the crime scene like a disk or a security guards pistol. It would then connect the player to the crime, and he would be really arrested. If not it would simply jump forward to the next day when he gets released.
If connected to the crime he would have several choices. Pleading guilty, pleading innocent, hiring a lawyer.
If he pleads guilty he gets a reduced sentence or maybe parole. Innocent he could get off clean, or get more time depending on the ammount of evidence. If he hires a lawyer ($$$) it will improve his innocent plea chances.
Maybe some more options like "Bribe judge" etc. Yet he would simply click on his choice, and then it would tell him what happened.
If he gets convicted it would then simply jump forward to him getting out of prison after his sentence. It would be on his record causing all sorts of problems. Probation etc.

It's a lot to bite off I know, but I think I can manage it. Basically there would be two or three main plots all linked together. So no matter what type of person they played it would follow the same story. Just in a different way or on the other side of the fence so to speak.
Plus lot's of side quests or jobs. So they can just do their own thing if they wanted too.

It is a lot though thats why I'm trying to lay it all out on paper first. Get a game plan going and work through it. Rather than getting somewhere and needing to redo something because I want something else. I'll have enough of that without trying to create code, and the game plot at the same time.

Plus I'm going to break it all down by sort of doing a demo and then adding features to the demo. The game is sort of meant to be played on the shady side of things though. Sort of Shadowrun like in many ways.

Of course I realise that link I posted above isn't my best work in 3D. It was something I made for a total conversion for Morrowind along with a bunch of other weapons.
Total time for that sword took me around 10 minutes to model, and texture.
It's also under 200 poly's total, and I used a compressed .tga file for the texture. It's basically a beveled plane, and a cylinder makes up the entire model.

Anyway I know I'm definitely biting off a big chunk. However my current job allows me 8 hours a night where I can model, program, etc. While getting paid for doing so. So basically I can work on it 8 hours a night 5 days a week without any time problems. Even more if I wanted to cut into my own time which I probably will as thats all I do now when I'm at home is model in 3D.

Mod's for Morrowind, skins for IL2 Forgotten Battles, and levels for Soldier of Fortune II.

I might as well put forth all of that effort into my own creation.

*yawn* another RPG...

Warhunter,

You should be nice and a little opened minded. So you don't like RPG's. Thats great, next time don't post a reply, its simply not worth it.

That sounds interesting Zen, any idea when you will have at least some ideas on paper?

Yeah I've been writing up ideas stage by stage in a notebook. Of course I then have to go back and add to other parts as I get other ideas while thinking about another part. You know how it all roles together.

Also it is an RPG, but it's more like the result of Soldier of Fortune II having a 100mph head on collision with Morrowind. lol.

Except in the future, and pretty much based in one vast city. I've also started doodling some things and I've been modelling them as well. Of course I'll have to refine them later, but I can still save myself some time by starting early.

Right now I'm working on character models though. As thats my weakness. People. They're tough for some reason for me, and then I have to model all the gear, armor, and clothing, etc to fit the characters.

When I finish some stuff up I'll post some renderings of them.

I'd go with mid-range polygon detail, and keep the number of objects to only the neccessary amount... lowers development time so you can spend more focusing on coding.

Polygon detail is important for lighting too. So you might want to beef up the detail in the parts of weapon you're actually going to be seeing (the top and back) and leave the other part very low detail .

las6-- did you ever try ATI's Truform with Morrowind? With MW FPS Optimizer it's possible. It beefs up the polygon detail loads, makes it run like a snail, and the lighting looks absolutely wonderful.

I do agree about the textures... but they served their purpose okay. I wasn't too bothered by it.

Morrowind has to be the first game I've seen that had tables and chairs with curved edges, rounded feet, etc. They just went nuts with polygons in it.

How's this for a model for in game?


It's a cross between a AK-74, and a Galil.

Edit: I just stumbled upon a name I think sounds much better than Earth:4020

"GRIM FUTURE"

As I sit here jotting stuff down in my notebook as it comes to me since I have nothing else to do at work. I realise now that I have bitten off more than I can chew. I still think I have an idea for an awesome game. Just that it would probably take a half dozen modellers, a few programmers, and a year or two to create. So I am putting it on hold for now.

Instead I got an idea for another game that I think till rock hard. Which is probably not a whole lot easier, but doesn't involve a library worth of dialogue, and modelling an entire vast city. It may be difficult though to implement the multiplayer setup I want.

The only name I can think of is "Mercenaries" for now anyway. Basically it will be a FPS with some quirks. In single player mode the player will be like the boss. He can hire available Mercs, but has to pay them. Each job he takes will mean getting paid depending on the difficulty, objectives, etc. Then he will have to spend that on equipment, and gear for his team as well as pay their salaries.
He'll also use it to buy or rent a transport for their next mission.
Each mission will begin with sort of a Rainbow Six style strategy session.

Online however it will allow different squads of Merc's to go head to head for cash just like in Single player mode.
And I don't know if it's possible, but I would want a stand alone server programm. So that the players could have "Hard core" characters. Meaning when they die, they lose the character. Yet online from mission to mission their money stays.
A bunch of other stuff to make it semi-rpg. Yet mainly a FPS. Players will be able to spend money to buy just about everything they want or need. Such as weapons, explosives, general gear, body armor, technical equipment, etc. I mean does he want to spend $15 on a good compass or $400 on a GPS system? Buy an AK-47 legally for $300 or go on the black market, and pay $3,000 for one full auto with all the bells and whistles.

Online rewards would be done on a handicaping system. Where it would average one teams rank, and objectives, against the other side. So a high ranking team wouldn't earn as much going up against a bunch of newer players. Yet the newer players could earn a bunch if they win.

Still complex I know, but much simpler than what I was trying to do before.