I’ve been working on a ambitious Role-Playing game project since a while now (in fact quite a long while but I’m not in hurry to finish it anytime soon) and while I’m in the process of working out the story, technical details like the combat mechanics, skill system etc. and creating interesting characters I still haven’t made a decision on the type of graphical projection for the game so far. I’ve been thinking about five kinds of projection from the most basic one (2D orthographic) up to full dynamic 3D which would be quite an effort. As my development platform of choice happens to be Flash, the resources in terms of 3D are limited.
So with that in mind I thought it would be good opportunity to introduce some of the most-used projections in computer and video role-playing games to get to know them a little better. This is by no means a complete list of all sorts of projection used in games but I believe these the ones most commonly used for role-playing games.
Orthographic Projection (Top-down View)
Orthographic is pretty much a collective term for all projections that lack perspective (i.e. all 3 axes have the same length) but I’d like to use it here to describe the typical top-down view – often called birds-eye view – used in RPGs.
The top-down view can either be completely flat which is obviously the most easy way to draw graphics or it can give the impression of depth by using an oblique style. Many of the older J-RPGs like Zelda and Final Fantasy are using this method. You see buildings, characters and other objects that point into one direction (usually to the top of the screen) to give the illusion of depth. Often this style is used in a very simple form so that characters are not being able to ‘step behind a wall’, i.e. the collision detection that checks where the characters are allowed to walk adheres directly to the graphic tiles used in the game. Some other games of this style might involve some trickery to offset the line used for collision detection so that characters can walk behind walls and are half-covered by them to add some more dynamic.
The advantages of this projection are clear: it’s the easiest to draw and the easiest to implement. Also the nature of this style makes it easy to use for RPGs that use a character party which needs to be guided around the screen. Selecting party members and using them tactically (positioning, line-of-sight, etc.) is more easily done here.
The disadvantages are that this projection is the most simple looking and most overused one (but then again these are not necessarily disadvantages) and in particular that this projection type can ruin potential immersive atmosphere, i.e. unless you’re using some sort of Fog-of-War the player can see what lies around the next corner. Even if you don’t draw any opponents until they are in the line-of-sight of the player in my opinion this doesn’t convey the same level of immersion as 3D or Flip3D (see below) projection would.
Isometric Projection (Tiled)
Isometric view and (although not technically isometric) it’s cousins trimetric and dimetric are projections that were (and still are) used in a long list of games, mostly strategy-, management- and of course role-playing games. Highly regarded by many as the finest-ever invented graphics style, isometric projection has the advantage over basic top-down view that is adds more depth and at the same time often more dynamic and clarity to the layout of the play field.
I must admit not many RPG titles that are using a tile-based isometric projection are coming to my mind but some good examples from this domain are Fallout 1 & 2 which use trimetric projection (all three angles are different) and the Indie RPG series Avernum.
Advantages of the tiled isometric projection include the easiness of implementation and a big favor for tactical gameplay. It’s easy to see tactical elements in relation to each other and so it comes to no surprise that many strategy games used this style.
The disadvantage I see with isometric graphics is the same as with top-down view projection, the lack of immersive atmosphere (see above) and that isometric graphics (tiles, sprites) are more difficult to design (but assets rendered with a 3D package can help a lot here).
Isometric Projection (Full Backgrounds)
The big contender among isometric engines that used pre-rendered backgrounds instead of tiled graphics was without a doubt the Infinity Engine and with it such classics like Baldur’s Gate, Icewind Dale and Planescape: Torment.
Instead of using tiles, larger backgrounds were modeled, textured and pre-rendered with a 3D graphics application and then used … well, as sort-of large-scale tiles in the game. This technique allowed for a lot more visual detail and variation. Environments that were difficult to draw with a tile-based approach such as for instance a deep, rocky abyss or other intricate landscape were made possible much easier with fully rendered backgrounds. Also many other items like furniture and interior walls could suddenly have a lot more detail. Another advantage that the Infinity Engine introduced was the use of dynamic lighting and acceleration of visual effects by use of DirectX. Full-screen rain or snowstorm? No problem here!
Collision detection and path-finding was solved by using a special bitmap for every background that defined the walk-able areas using different colors.
The big advantage for this technique is of course the gain in graphical detail and that it eliminates some of the problems of a tile-based engine, namely intricacies when dealing with dynamic lighting, path-finding etc.
The disadvantage here is the extra-work of complexity required to implement such a full background rendering engine, the additional task to design pre-rendered background graphics using 3D modeling and that such an engine requires more system resources than a tile-based engine, RAM and CPU-wise.
Now what in the hell is Flip3D? You might hear this term for the first time ever here but Flip3D is what I like to call (for the lack of a better word) the projection style used by many games from the Golden Age of RPGs era that use pseudo 3D images which are drawn with perspective in mind and if you turn in-game to the left or right the whole view is flipped turned by 90° hence the name, uhh, Flip3D (if you know the technically correct term for this type of graphics display technique I’d like to learn it but until then let’s just call it Flip3D).
Countless RPGs utilized this projection, most notably titles like Dungeon Master, Eye of the Beholder, the Ishar series, the Might and Magic series and – although not really a RPG – one of my all-time favorites, Hired Guns. All these games used pre-rendered graphic parts like floors, walls and ceilings that are composed together on the screen to create a room or corridor or even an outside environment.
The advantage is that these graphic parts could be re-used, making it easier to quickly build whole maps. The disadvantage to this was often that in many of the games the maps looked monotonous. In my opinion Hired Guns nailed this problem quite well which is one of the reasons that made it stand out from the others for me. The maps looked detailed and very different from level to level, in the background they’ve used a gradient to represent the dawn on the firmament and there was usually a dark backdrop used for a horizon to give the impression of mountains, landscape or buildings in the far distance. It’s exactly these details that are one of the features that filled the game with atmosphere. What could be there in the far distance? The distance that you were never able to reach because it was not a part of the map but, alas, only a backdrop.
The lack of clear detail opens a lot of freedom to the player’s own imagination and this is one of the points why I highly regard some older games that use simplified graphics where modern games with tons of detailed real-time 3D choke the last bit of imagination from the players mind (not that I’m saying that this is necessarily bad but for me it is gaming on a different mental level).
And with that we have the main advantage of this projection type … used with well-drawn graphics it can add a nice deal of ‘hidden’ atmosphere to the game because you never really see what lies (or lurks) in the distance.
The main disadvantage of this projection from my point of view is that it’s not much in favor for tactical gameplay as long as you want to involve your party into the action. Games with Flip3D projection are necessarily first-person so you don’t see your own character and your other party members are either behind or sideways of you or standing in front, covering a big part of the view field. Guiding and commanding all your party members can become quite a chore here.
3D projection is the current state-of-the-art in computer game graphics and will probably not be replaced so soon (until somebody invents some sort of 4D virtual reality device). This projection allows the highest degree of freedom meaning that any object can be viewed from any angle, at any distance and with arbitrary FOV.
There is virtually no quality loss when zooming in and you can re-use 3D objects for all kinds of displays, for example you could have your once-designed character 3D objects walk around in the game world and at the same time use them on a special inventory window where you drag and drop equipment onto the same used 3D object of their bodies. This allows for ‘create-once, use everywhere’ unlike with 2D assets where you often have to create variations of the same object for use in different views.
You could also use 3D with a fixed camera to simulate a certain type of 2D view and the advantage of 3D objects would still be obvious.
The disadvantage of choosing this projection is that the creation of assets, i.e. 3D models of characters, items and environments is a lot more involved as with 2D graphics. Unless you are some sort of super-developer-designer-machine with 300 years of time up your sleeve you probably need a small team of professional 3D designers who can do this work for you. Ok, I’d say nobody died of trying but recent 3D game graphics have become so professional that it’s not an easy feat to try catching up with it. And who wants to design a game with 3D graphics that look like anno 1996? I’d probably be better off with nice-looking 2D graphics instead!
Then, as you want to create a large world for a RPG you’d have to design many interior/exterior maps that are used for the game world. If you use 3D I doubt you want to design this all manually. What you need is a map editor to create the environments more comfortably and re-use assets quickly etc. Writing such a map editor is already a whole job unto itself and takes a lot of time.
Another complication with 3D graphics is the platform used for the game. I use Flash (or AIR) with use of ActionScript. While I’m convinced that ActionScript can handle all the other projection types, with 3D we’re hitting the limits very quickly. PaperVision 3D, Away3D and the Alternativa Platform all have been used sucessfully for game development already but I have yet to see a good example of an extensive RPG made with one of these and the fact that a few thousand triangles bring the framerate to a screeching halt doesn’t help the case.
Still, any of these 3D engines could be utilized in Flash for a simpler form of 3D graphics style, for example the use of mainly simple primitives with a good job in texture detail work and a combination of 2D sprites. I’m definitely looking further into this possibility.
I’m still in the dark about which projection style I’ll eventually use. I might be creating a first prototype with a very simple and flat top-down view (a la Ultima) to test other components of the game and then later change to something more advanced.
Any of the introduced techniques have their charming sides, for example (as you might have had no troubles noticing) I’m very fond of Hired Guns and it’s dark and futuristic pseudo 3D view. Then again an isometric view could be more suitable for tactical gameplay (which, as I’m planing the game should have quite a few of).
Realtime 3D would be the ultimate thing but I dislike the idea of creating super-simplified 3D objects and Flash currently lacks the rendering power to use higher-end 3D graphics. I might as well be watching if the Flash platform improves in this area and then see what can be done in terms of 3D later.
Was it that? I’m sure I forgot a ton of facts and details here but I would be very interested in what others have to say about this topic or what view styles you’re about to use and why etc.
Further Reading: Significant Bits – A layman’s guide to projection in videogames