Its been a while since my latest tutorial and I have to disappoint you right away: This one won't have any pictures.
In fact, I fried my mainboard and I am typing this one on my netbook.
So lets jjump right in, shall we?
In this tutorial I will demonstrate 10 tasks you should perform if you are using FPSC and 10 mistakes you really dont want to do.
Lets start with the more positive tipps:
I mentioned this already on various occasions but I'm hinting to it again: There are extra texture maps for your stock scifi lamps in your texturebank. These have all sorts of different colors and make a great addition to your entitybank. If there is one thing we aren't having too many of already, it are lightsources and lampmodels. To get these working:
Open your texturebank and check the name of the textures.
Copy the FPE Files of the lamps in your entitybank and rewrite the texturepath within. (You can do this using the Windows editor).
You are now having a large number of extra lightsources...remember: Making copys with different scale values amps up you mapping possibilites even more. And that gets us to the next point in our list.
Copy the fpe of whatever models you like and change the scale values. If there is no “scale” value in your fpe, just add one
(type in “scale = ??? (100 is default)”). This should be performed on the architecture props first as these are most important. Try not to save your textures as a png with too high quality settings. These can be up to 20 mb in size and cramp up your memory cap. Remember to save in a well optimised format and use a .dds converter that isnt complete bogus.
Same goes for Soundfiles. A music file should be no larger than 10 megabytes. Use OGG or MPEG3 formatted WAV files. (I recommend wavepad)
Try to use shaders only on metalic objects and segments (unless you really know what you are doing). FPSC does not have a built in materialeditor and you might just end up cramping up your memorycap and drowning your framerate if you drench everything in shader effects. Your game can also end up looking awfully artifical and overshaded.
You can find seriously great public domain music on wikimedia commons...check it out!
We are all really jaded with modernday's borderline realistic videogames like Battlefield, Call of duty and the likes...to ground yourself back in FPSC reality and what you can achieve in its map editor: Install some oldys again: Get out Project IGI, Return to castle wolfenstein and other classics...you can acutally achieve this quality. I see too many new users trying to get somewhere close to the things they see on their 360...forget about that!
Also: Boxes and barrels only where there really could be boxes and barrles. For instance: nobody has some gasoline barrels in his bathroom.
If you write your game's storyline, I do suggest getting some inspiration form games like Penumbra, Amnesia and Outcry...all of these games play in a single location. Remember to think that you want to make an entertaining short game with simple mechanics, not a world war epic. In fact: If you get any inspiration from Call of Duty Modern Warfare, you are doing it terribly wrong. FPSC can not handle any massive locations and warscenes. Nor can you, as a single developer without any experience create such games and design so many different locations. You have to really get into your setting...”feel” the environment and get it into your editor. Make sure that your level has gameplay of at least 4 minutes. (you will most likely not make it any longer without making it really....really boring)
Give your files singlelined names. Name a model “ww2_tank” instead of “ww2 tank”, If you ever import your modelcollection in another engine...you will thank me.
Do not use triggerzones only. Get yourself some little transparent boxes you can use as triggerzones aswell. Imagine all the precise perfect spawning you can do if your triggerobject isnt the size of an entire segmentblock
If you havent thought about this already, I think this point is the most useful I made in this tutorial.Alright, I hope I haven't bored you to death... lets get to some...common mistakes and things I learned over the years:
Do not try to design your maps in different software (for example: Milkshape). The only way you could import it is as a bunch of single meshes and you will struggle to get the lightmapping to work correctly on it, to get the collision not completely terrible and of course to assign the soundmaterials. (stepsounds). You will also get bad culling.
Do not completely cramp up your memory cap in your level. It is likely that this will cause slowdowns, freezing loadingscreen or crashes in the builded game. Also: keep in mind that as soon as your level is done, you do not have added weapons, sounds, scripts, enemies and HUDs yet...these will also take up memoryspace and can just cause your level not to build anymore.
Do not try to sell a bear's pelt if you haven't yet shot the bear. Do not try to sell a game you haven't made yet. Also, and keep this in mind, under no circumstances shall you contact more experienced people to make your game for you if you havent the skill and absolutely no experience. If you think you have a great game idea nobody else thought of yet and if you believe that you a far more creative than anybody else: Nobody cares, we all are. Its a huge difference between having an idea and creating it. If you have the fantasy and the skill... go for it, if not: Leave us alone. You wonder why so many games are alike? Well, that has a lot of reasons. Especially in the commercial sector. If you want to make a commercial game....lets roll! Start right now with a good concept. Keep in mind to use only stuff you have a license for and contact publisher only ...and only if you at least have a demo ready (demo...not beta demo). Also... try to get in touch with lower publishers that aren't that popular ( you wont get it on steam...). Or, if you only do it for the fun anyway... distribute it yourself like I do with Psishock. Also, if you just cramp a few modelpacks together in the editor, you havent made a commercial quality game.
Do not make the maps before the gameplay, try to take a “level by level” approach.
Even though I suggest a “level by level” approach: Start with level 2 or 3. Remember to make level 1 at the end. (because you will learn while making your game and you should do the first level with your highest skill... because the player will get his first impression and decide wether or not to keep playing in level 1)
Do not use models from different packs that dont mix well. This should be self explanatory to most of you...but some are … less talented and do things like mixing modelpack 4 and zombie apocalypse characters... or task force 341 with stock ww2 officers which is...awful. It is better to have less models that fit well than a lot of meshes that dont work together at all. If you think about mixing modelpack 6 and 10 weapons you arent reading this tutorial anyway... If 2 meshes from 2 different authors dont really look good in the same scene, make sure to adjust them visually with some good retextures.
Billboard plants dont work well as static meshes... keep that in mind especially if you use high resolution lightmapping.
Do not use small props with a polycount above 1000. Its just a waste of polys and a simplified mesh has the same impact on the player. Remember to keep a good framerate...nobody likes lagging.
You shouldnt try to make remakes of older FPS titles...a fpsc game usally takes 30 minutes to 1 hour... a commercial fps release far far longer... you are also likely to get in trouble due to license inflictions or even ripped models if your game gets attention. It is also likely that your remake ends up far worse than the original,
Dont just work alone...keep checking the forums. To this day, there are people coming across FPSC videos on youtube wondering how to get hands on the gun... its really sad to see that some people actually use the software without even checking what other people do with it. There are even some dudes that bother me with emails and just refuse to register here... for whatever reason. The key to success in whatever you try to do is: Do your own thing, share and check out what your fellow developers do. Read, adapt, create.
Okay doke, thats enough gibberish for this tutorial...have a nice evening and: Merry Christmas everybody.
...and yes...I forgot to add episode 3 to the title. You know...gamedevelopers have big problems with episode 3 =P
the formatting didnt transfer from openoffice...sorry for the huge textblocks.
Keine Kommentare:
Kommentar veröffentlichen