[GSoC 2011] Student looking for a little advice

Threads related to Google Summer of Code
Post Reply
Chris_Neal
Gnoblar
Posts: 1
Joined: Tue Mar 22, 2011 10:24 pm

[GSoC 2011] Student looking for a little advice

Post by Chris_Neal » Tue Mar 22, 2011 11:01 pm

Hail the forum!

About me: I'm a student at the University of Pennsylvania, studying Digital Media Design; and I've been tinkering with OGRE for a while. I'm a longtime lurker with an interest in contributing; and as I've got a better knowledge of OpenGL and graphics progamming now than in years past, I have been considering submitting a project for GSoC.

The purpose of this post, really, is to try and get a feel for how large or small a potential proposal should be, and to try and narrow down my interests. I would love to pursue the more novel uses of a deferred rendering solution, such as the recent (very cool) paper on sub-pixel reconstruction anti-aliasing http://anteru.net/projects/research/sub ... ialiasing/ ; both splatted and volume-based GI for realtime, and similar deferred-rendering specific optimizations/features.

Of the items on the help requested page, I found alternative particles rendering, CHC, and dual-quaternions to hold my attention most. I'm curious if others think that the particles rendering would really constitute an entire GSoC - if not, one thought I had involved potentially implementing a more robust set of both that and related items, like volume rendering.

In short: I don't have much DirectX experience, and while I find several suggested projects interesting, I'm a bit wary that my knowledge in some areas might be too low - while at the same time I'm a little unsure how to shape my own interests into a project proposal of an appropriate size and GSoC-worthy goals. What I do have is several years of C++ experience and a understanding of low-level OpenGL that's getting better all the time; and hopefully some good advice from the fine folks on the forum.

~Chris

P.S. After reading the other GSoC posts, I remember reading the advice that I have a look at fixing some outstanding papercuts - hopefully that'll happen later this week when classes quiet up a bit.
0 x

User avatar
Noman
OGRE Retired Team Member
OGRE Retired Team Member
Posts: 714
Joined: Mon Jan 31, 2005 7:21 pm
Location: Israel
Contact:

Re: [GSoC 2011] Student looking for a little advice

Post by Noman » Tue Mar 22, 2011 11:36 pm

Hi Chris, and welcome to the forum.

I'll try to address your problems one at a time
1) You are not limited to the ideas in the help requested page. If you have an idea that you think will contribute to the ogre project (and not just to 1-2 users), feel free to pitch it! If we think its not general enough, we'll let you know. Some other threads talked about continuing the deferred rendering pipeline work, and doing such things as integrating the shader generation with RTSS, or presenting a Unity-like unified shader architecture, where each shader can be written once and used in both the deferred and normal pipelines (this one is hard).

In my opinion, going for RTSS+Deferred shading integration would allow all of the demos that don't use hand-written shaders to run in a deferred pipeline, which would be really cool. (This also goes out to any other students reading this. I think it would be an awesome project).

2) I think the particle rendering is big enough to be a project. There is quite a bit of work to be done to get it to work with both good performance and simple usability. Also, remember that you will be in charge of the documentation and demos, so thats a bit of overhead as well. If you have several years of C++ experience, you should be able to get a rough estimate on the amount of time the tasks will require. Planning early will allow you to refine your suggestion more before the application deadline!

3) I don't think lack of DirectX knowledge is a problem, unless doing a rendersystem GSoC (which you aren't). Most projects will use the RenderSystem in an abstract fashion, so you only need a good understanding of what 3d rendering is, not necessarily knowing a specific low-level API.

Good luck!
0 x

Post Reply