I've never used Ogre personally, but I have read great things about it before. One good thing about it is the consistent syntax, so once you learn how to use one or two things, it makes it much easier to learn the rest of it, because they follow similar naming conventions.

I'll put in a vote for Irrlicht as well. Although it's not as feature-rich as some other engines (like Ogre,) it's got a very easy to learn API so you can get up and running really quick.

A 3d engine is just that - an engine for 3d graphics. Games are just one type of program that makes use of 3d graphics. A game engine usually contains a graphics engine, along with other aspects a game would need (input, entities, collision detection, sound, music, user interface, etc).

Only good 3d game engine I know of that is of a reasonable price for something like this is Torque from GarageGames. I have never used it personally, but have heard some good things about it.

I've used the Panda 3D engine for a school project. I wouldn't say it's difficult, quite the contrary. I've found it nice to work with, and although the built-in physics system is not so impressive, there are solutions to that. We used Python with it, which allowed us to develop things quickly. I think it's also possible to use C++ with it though.

Ogre 3D is a rendering engine, but a pretty good one at that. Some of my colleagues used it for a demo and they were pretty happy about it. I'm using it for a personal object for 2D rendering, which also works really fine. You do have to gather some other libraries for the other aspects though.


I haven't tried any other engines so far, but I can recommend another option: modding an existing game. Many shooters nowadays come with a whole set of tools and editors, which allows you to write your own game-code, build your own levels and models, and so on. All on top of a solid toolchain and game engine.

I would also recommend using some off the shelf game technology, with good modding tools.

Your choice would be dependant on the type of game you hope to make. Start asking yourself about what your goal for the project/game, is it a shooter or stategy maybe? or possibly even a 2D platformer? Do you want more than one player?. You can use the answer to these to help determine your choice of technology. You should take some notes while making these choices, for use in your report. Even if you do not use a particular technology its still worth mnentioning it and why it was not suitable.

One think that might be worth taking into consideration is the size of the modding community behind a particular game, it can be very useful to be using a set of technologies that alot of people are familar with rather than some far out tech that no one knows how to use :).

If you select a relatively popular moddable game then pick up the obligatory initial walk through tutorial and start getting your feet wet.

Some might advocate doing loads of design first, and indeed you should take notes of concepts you have, but I think you will find it easier, and less frustrating, if you get in there and get a feel for working with the technology first. 'Unknowns' are horrible when designing; so get in there and find 'stuff' out :)

Goodluck :)
Michael

Edit: Forgot examples (doh), i haven't much experience with any of these but here is the names anyway:

RTS:
Company of Heroes: (Level editing mostly but i think you can make a single player missions no problem as well, few of those and you have a campaign :) )
World in Conflict: Same as above, this is more about multiplayer.

FPS:
Valves Source Engine (half life 2). Seems really easy to use and you can delve very deep into the source code if you want :)

These engines will give you everything you could possibly need for a school project within their respective genre specialities. There is many others though, I sure people will come forward with more.