Sign in to follow this  
Drats

Water - Masters Thesis

Recommended Posts

Hi this is a rather vague question but hopefully I can get a few pointers. I am doing my Masters thesis in computer games and I intend to do some sort of water implementation as the subject. I have looked around the other work that has been done in the area, but I have not been able to hit upon my exact topic till now. so I know I want to do a water effect but how exactly do I convert it into a masters level project, so that it is something fairly new. I have looked at navier stokes equations, perlin noise etc but still don't know what my project should exactly be about. Can anyone give me any guidelines please.

Share this post


Link to post
Share on other sites
The first step is to do a literature review of the subject to find out the work that has already been done. Sounds like you have started on that, but you are not done. Generally you know you are done with your review when you have traced all of the literature back to a couple of source papers (probably written in the 60s and 70s). You will also start to notice a lot of "cross-pollination" among the more recent papers. Most papers have a "future work" section at the end. That is a good place to get some ideas on how to make your unique contribution, which is the ultimate goal of your thesis. After you read enough papers you will see some patterns emerge. Most of the papers will point to the same unsolved problems, or they solve the same problem but approach it in different ways. You can then decide if you want to go for one of the unsolved problems, or try for a new solution to a previously solved problem. Both have their challenges. If you go for the unsolved problem, you may not find a solution. You can still have a good paper if you rigorously document all the things you tried and show why it failed (heck you may prove the problem is computationally unsolvable, bonus points!). If you go for the new solution to the solved problem, look at the existing literature. A good paper will identify the strengths and weaknesses of the proposed solution. You may try to find a solution that targets the weaknesses of the existing solutions. Good luck, and make sure you talk to your advisor. They have already been around the block on this, so you should be able to get some guidance.

Share this post


Link to post
Share on other sites
I don't think a masters' thesis needs to be original work. I have seen some that survey many techniques, provide implementations, and then do a pro/con often backed up with statistical analysis (e.g., framerates, memory footprint, etc).

So you might include some old methods, and then look at what is being used in games today--displacement mapping, normal mapping, reflection/refraction approximations, etc.

You might look at methods based on physics and also methods that are not based on physics but look convincing anyway.



Share this post


Link to post
Share on other sites
I guess it depends on the school. I have read a number of Master's theses and they all at least attempt to provide a unique solution to a well defined problem. The research does not have to be world changing, ground breaking work, but it should be original, otherwise why bother? Remember, you will have to defend your work and convince a bunch of crusty ol' professors that they should confer upon you the title of "Master of Something". Your life will be much easier if you can stand up and say "here is a problem, here are the published solutions to this problem, here is my solution, and here is how it compares to the others and why it is better for some case/worse for some other case" rather than "I implemented a bunch of papers I found and here are some charts, I also brought doughnuts". Bottom line, you should talk to your advisor.

Share this post


Link to post
Share on other sites
I always thought that solving the wave equation in real-time on a displacement map of a water surface texture, and rendering it with full reflections & refractions could be rather impressive. I haven't looked very hard, but never encountered a realistic real-time interactive (shock-free) water simulation.

Share this post


Link to post
Share on other sites
"Novel" new implementations of existing algorithms tends to be looked upon with reasonable regard from my experience.

I'm sure plenty of people have published CPU-based water renderers over the years, but how many are implemented using GPGPU techniques? I'd imagine a review of existing CPU techniques and how they can be mapped to GPU implementations might be a plausible topic...

I agree with CodeMunkie's opening post - once you've done the literature review you'll have a very good idea of what sort of things have been tried and what hasn't.

Jack

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this