Honours Project Topic Choice

When it came to thinking about what I was going to focus on for my Honours Project, my mind instantly gravitated to graphics - lighting and shadows in particular. This is probably because during third year most of my coursework projects were focused on lighting, shaders and shadows, so I feel that I have more experience in that area than any other. It may also be because I like making things look pretty. Who knows. Either way, I've developed a strange fascination with shadows. Every time I play a game now I look at the shadows and I think 'ooooh nice, I wonder how they did that.' (Except in the case of my crappy old desktop - playing games on that made me think 'oohhh, I wish I could put the graphics high enough for shadows D: '

Anywho, onto business!

1. State the topic of interest that you may wish to take forward into a project.

Like I said, graphics. I'm interested in lighting, shaders, shadows. Especially shadows. I'd like to look at shadow rendering in realtime 3D applications.

2. Indicate the issues surrounding this topic that you may wish to investigate.

I think that the two main issues when it comes to shadow rendering are performance and quality. The calculations required to create soft, real looking shadows often require a lot of processing and cause the application to slow down. But at the same time, realistic looking shadows can add a lot to a 3D scene so we don't want to sacrifice that graphical quality altogether. The question is, I think, is how to find an acceptable balance between the two.

There are a number of methods available for rendering shadows, i.e. shadow mapping, shadow volumes, z buffer shadows (though these are very basic) etc, each with their own pros and cons. Some of these algorithms could even be combined. (I believe there is also a ray tracing technique for shadows but it would not be suitable for real time calculations)

3. Outline your initial ideas about how you would carry out the practical aspects of the project. i.e. What do you envisage yourself doing for the project. 

• First off, I would obviously need to decide on the platform - probably out of DirectX, OpenGL or XNA. (I'm leaning away from OpenGL already though, mainly because I dislike GLSL.)
• Build a framework in which to run the application - it will have to display a 3D scene with static and moving objects, with support for (possibly)multiple shaders.
• Research and implement the different shadow calculating/rendering methods within this framework and be able to switch between them in order to compare both framerate and realism.

***************************

And there you have it. My initial ideas on the honours project. I'm very unsure about it to be honest, I can't decide whether it's too general or too narrow, too simple or too complicated. But I do know that I definitely want to do graphics...that's probably the only thing I'm positive about. 

Anywho, thankies for reading!