A screenshot of the scenery (as of Mar 23, 2008) viewed from about 6500 feet above ground in very early development.

What is version 3.x of The Supernatural Olympics like?

1 What a 2.5D game is

2 Why go with 3D?

3 The advantages of using 3D rendering

3.1 Realism advantages

• The flat ground can be made hilly. Currently, the ground beyond the pathway is flat, perfectly flat. That's unrealistic. By adding hills here, the quality of realism can be greatly improved. The case with the Northern Mebna Hills and Urusu Hills/Mountains are not what I'm referring to. Think about the city area, or Urusu Valley. You don't find cities that are all at the same elevation all the time. By raising it above the 0-point to match that of the pathway, it's much more realistic. Currently, there's a sudden drop beyond the pathway. With a high quality setting, these hills will be very smooth, but a low quality setting will cause them to have somewhat sharp angles.
• More realistic clouds. With the current, planned setup, the cumulus humilis clouds are in bands and only appear within that band. With 3D, instead of needing a set of nearly 200 clouds, I'd only need around 30 or so varieties and they can be placed absolutely anywhere which provides great realism. In addition, instead of needing about 8 MB of data, it's barely 40 KB for the shapes alone and another 20 KB for the texture.
• Far more varied background with terrain. The current design has the Northern Mebna Hills loop indefinitely, along with the Urusu Mountains and other objects, even if you're in the lakes, the desert, or high in the mountains. With the current design, the lakes cannot go any more than 750 feet into the background, but that's not very realistic. Imagine if, say, Stugu Lake went back not 750 feet, but 3 miles, considering it's over 9 miles wide as is. Lake Windias is the same, only nearly 11 miles wide. Now, let's say you're in Quredda Desert. Look 750 feet into the background and you see grassy hills. Huh? This won't be the case with this advantage. Also, take the Mebna Mountains and Stugu Mountains. When you're up in the mountains, you'd expect to see a lot of mountains. This could be done in 2D, but you're looking at a memory usage that even today's 8 GB systems can't take. With 3D, just throw on a texture, which takes barely 3 MB or so, depending on the size I use, and you've got the whole mountain range. This way, when you're up in the mountains, you'll see other mountains up close. This would greatly enhance the realism beyond my original visions.
• Highly realistic buildings. Currently, the draw count of 2000 per frame is causing a lot of problems. To add the roofs, walls, and other aspects of the buildings would easily exceed these, even if I used quality 50. By using 3D, I won't have this problem at all and on top of that, they could be made far, far more realistic than the currently are.
• Realistic cars and car-movement. The current plans for the cars involve them going along the road exactly from west to east all at the same elevation. To hide the cars, the landscape blocks them or a tunnel is used where the landscape still blocks them. By using 3D, I don't have to do this that much. Instead, there can be curves in the road where the cars just follow the curves. A tunnel will still be needed for the pathway to go over (getting hit by a car would cause problems), but not in the way I had originally planned. This can still be done in 2D, but you're looking at abrupt changes and a lot more memory being used (let alone reduced motive).
• Instant fog intensity adjustments. When you change the visibility, fog has to be recalculated. To do that, the scenery objects must be reloaded entirely and recalculate the fog, which can take about a second or two (visibility is set in the menus and thus won't affect gameplay, even if it took 20 seconds, the game would still be playable). In addition, as height increases, visibility often increases with it. Look at Earth from space - you can easily see rich detail.
• Actual scenery underwater. Currently, as planned, you'll only see the pathway when underwater, due to low visibility or something. How would a 500-foot visibility sound where you can see much more of the bottom of a lake? It's worth it!
• Dynamic shops. Work hours are usually 9 AM to 5 PM, with some variations such as 8 AM to 10 PM. By using 3D, dynamic shops are much more likely and easier to do. In turn, rush hour traffic becomes possible as well, just something else to enhance the realism. Even then, shops change their signs every now and then as well.
• Realistic neon lights. When these lights are on, they emit their own light. When they are off, they are affected by the lighting of the surrounding area. Using 3D just makes this easier to do.

3.2 Quality advantages

• Smooth movement of background objects without antialiasing. With 2D, a whole panel moves one whole pixel at a time causing a sudden "bump" effect. If you've watched my animated GIFs, you should be aware of this. With 3D, the "bump" effect is no longer present. The objects will move very smoothly, even their textures. This is because of the way 3D is rendered - plotting vertices, connecting faces, and drawing textures on them.
• The scenery seamlessly blends - no gaps between hills. If you looked at the screenshots or videos, you often see gaps between the hills that match the grass. This is odd and a limitation of 2D. If I made each hill as an individual object, this gap would be less noticeable, but still present. By using 3D, this will no longer be a problem and the heights of each valleys can be of varying elevations instead of always being the zero point. Everything would be individualized, but all interconnected.
• Little memory usage for trees, grass blades, water waves, and other repetitive objects. Currently, for the forest, I'm thinking of 19 different sizes for the trees, but at 70-100 feet tall for my target species (the Patula Pine, for example, is one I use as a basis) and the tiny distance used, this presents a severe memory-related problem if I took the 2D route. With 3D, I just need something for the branches, the trunk (for what little is visible), and other minor details for textures, which shouldn't take up much more than 1 MB. Then I could just scale and place them accordingly without further memory usage. This makes for a much more realistic forest. A high quality setting would allow for a realistic shape. A low quality setting would allow for the trees to be flat pieces that join up, a common method used. For grass, this is just much faster and simpler to do. Water waves are also somewhat similar.
• Ultra-fine rotations. Currently, gliding snaps to multiples of 15°, but how would you like multiples of 0.00005° or so? By using 3D, this is what I get. The rotations from gliding, float-running, and whatever else there might be would be almost effortless: "Character.Tilt += 0.1*Time;".
• Reduce polygon count, not variety. All the houses are the same and buildings lack variety. That's the lowest quality setting in the 2D version. The 3D version, since memory is no longer a problem, will, for low quality, only have a reduced polygon count, for improving speed, rather than reducing the variety. This way, you'll see the variety as if it was the highest quality setting in the 2D version, as planned anyway.

3.3 Development advantages

• Easy-to-make and smooth-animating character. This is the primary thing holding me back with 2D. The character, as planned, uses a fixed number of frames and the frames abruptly change. Now imagine smooth animation, 60+ fps smooth, and with little in the way of effort to making it. All I need to do is form the basic shape, then move each part by merely rotating them. Instead of an hour for each frame, it's maybe 3 minutes. This would significantly improve my motive for working on the character and improve its quality overall by a very large amount. It would also take up a lot less memory as well - just 200 KB or so instead of 3+MB.
• Automatic/dynamic lighting. It can be done with 2D, but unless you want 2 fps for even small changes, you're not going to get it. With 3D, lighting is automatically done. With Gamestudio, when I used it, I had a hard time getting it to work right because I either got gouraud shading (which is fine for smooth surfaces) or odd effects based on gouraud shading where a square had two shades of lighting used instead of 1. By using dynamic lighting, a day/night cycle becomes available so you can play around during both day and night.
• Automatic shadows. When an object blocks light from reaching another object, a shadow is formed. While it can be done in 2D quite easily, it's the draw count that limits this a lot.
• Optional decals can be more realistic and easier to make. Sailboats, airplanes, hot-air balloons, watchers, and other such objects could be made much more realistic and easier to make. Instead of sailboats going west and east repeatedly, they can go anywhere which is very difficult to do in 2D without using a ton of memory.
• Special objects will have easier collision detection. Remember the welcome sign? The top is supposed to be curved. Currently, I don't have a reliable way to judge the angle of the sign for collision-detection, but with 3D, this would be easier to do.
• Easier, more accurate special effects. When you're 10+ miles above the ground, you'd expect to see Earth's curvature and other objects beyond it. Effects like the flare effect and splashes are also much more realistic and easy to do.
• Cities in Urusu Valley are much easier to do. In Urusu Valley, located before the Urusu Mountains, there are lakes and forests, but no cities, as of my current plans. Cities are difficult to do, but, by using 3D, they are much easier and more likely to be added.
• Overhanging cliffs are likely easier to do. For a while, I've had the idea of using a vertex-based system for the landscape. I'm seeing problems with collision detection using the 2D approaches I've come up with. With 3D, it's much easier, supposedly. For background scenery, it's much easier yet.
• Multiplayer support is easier to do. With the current design, if I do add multiplayer support, the character for all players would be on the pathway in the same spot. It would be much more realistic to have the other players be at different distances, and by using 3D, this is easily possible without using any more memory. Otherwise, I'd have to make the character 5 times over (one for the main and 4 for extra players).

3.4 Other advantages

• Zoom in and out. Want to look at the scenery as if you were looking through binoculars? With 2D, this is extremely difficult to do, though still possible. With 3D, it's a snap and can be done in real time with essentially no effect on the frame rate. When more objects come into view (zooming out), obviously, the frame rate will decrease, but when fewer objects come into view (zooming in), the frame rate will increase.
• Highly dynamic minimap. The minimap in the corner, showing the lay of the land the character runs on, currently, is planned to use fixed zoom settings, which is a major memory hog. Providing a dynamic zoom setting has the disadvantage of a major impact on the frame rate. By using 3D, this minimap will not only be faster to render, the movement would also be much smoother as well. For 20 polygons or so, it should be very easy to draw....

4 Problems and set backs for the conversion