Re: Hobbyist Game Maker
on 10-26-2007 06:11 PM
Lighting: Depends on how you're handling your shape rendering.
Most often I've found that freestanding, non-bsp structures have lighting information baked into either the model, as vertex colors, or into a set of lightmap textures which are blended with the diffuse at render time.
Also, If this is going to be something the character can walk into you're going to want to think about how you're going to handle the collision. Collision meshes seem to work best as purely convex hulls, and given the building in question it would be best broken up into a number of different objects for collision purposes.
IE Pad, stairs, roof, columns, fence section, interior things each as their own object, etc. it will save time on the code end since the player will only collide with one or two simple objects rather than one overly complex object.
Also here's a couple of starting suggestions for your project based on my own experiences of making games on my own (small things), and long term independent team work:
1.) have a plan. Define WHAT you want your game to do. Spend a large amount of time actually documenting (in text) what your game needs to be a game. (make sure it runs on paper. [img]/images/graemlins/wink.gif[/img] )This is mechanics, not assets, assets come later.
2.) Put the art as priority 2. Get your code working first with really ugly things like boxes and spheres with textures on them that nearly hurt your eyes.
Even if your art oozes style and coolness, if it can't move in game, its just a portfolio statue.
3.) after your game is working, make art on an as needed basis. This means more planning. Once you have the mechanics at least to a playable level (pushing the buttons does things that they are supposed to do, the game starts and ends, etc) then art starts to become important. make a list of things you NEED. then a list of things you want. When the NEEDS are met (main character, x number of enemies, boss, playfield, playfield textures.) then go to wants if you still have steam and or time. (that really cool shiny sword that's a rare drop by a rare enemy)
4.) Beware of feature creep. Just because it sounds cool right now, does not make it so. Stick to the documents from part 1.
5.) Don't forget about tools. If you're writing your own game, you're likely going to have to (or want to) write your own tools, at least a level/playfield editor. Having good tools can really save time in the long run, and if you make them re-usable/generic you can build your next project more quickly (welcome to engine design)
oh and lastly: Toss realism. If you're working by yourself and aren't some form of divine being, then go for style. Style tends to work better than realism in terms of audience retention anyway. (cloud strife's hair and sword for example) Plus if you go for a more cartooned look (not specifically cell shaded, but exaggerated proportions etc) it can often save you effort. If you stay with a realisticish environment, but place very tooned characters in it, it will actually make the characters pop a bit more, which can be good for player recognition. (see "who framed roger rabbit?")
now as for art crits:
The character, and house still suffer from a lack of defining form shadow in the texture. They sort of melt into a single visual mass.
Also the character's head seems very odd. The lips and nose are oversized (which is fine), but the overall size of the head hasn't changed, leading the viewer to expect the eyes at about where the bottom of the glasses are. A solution would either be to elongate the head vertically, or bring the base of the glasses at least half way down the length of the nose. Speaking of which, there are an awful lot of polygons in the glasses for very little effect. its mainly the bevel you used through the center.
not too shabby work overall, especially for just starting out.