Jump to content
  • Advertisement
Sign in to follow this  
jammm

DX11 Technical papers on real time rendering

This topic is 1389 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Hey guys,

 

I have to provide a demonstration and give a seminar as a part of a college course this semester. As usual, I chose computer graphics as the domain of my seminar, and decided to implement a rendering technique and show that as a part of the demonstration.

 

I have found a couple rendering techniques interesting, namely SSAO (from Frank Luna's DX11 book) and Bokeh effect ( https://mynameismjp.wordpress.com/2011/02/28/bokeh/ (thanks MJP!) )

 

BUT, for some reason they require students to cite a technical paper as part of my seminar report which should have atleast a smattering of relation with the rendering technique that I'll be implemeting. Now I know what I'm supposed to present, but so far I just can't seem to find any good *related* techical papers which I could use. The main reason for that is that they want us to choose papers from IEEE/ACM dated 2012 onwards. With ACM as an option, I can look at SIGGRAPH papers too. BUT if I could find a paper pre-2012 which is still good, exceptions _could_ be made.

 

tl;dr :

 

Could anyone provide me references to technical papers which are atleast *somehow* related to SSAO and Bokeh (or algorithms used for Bokeh), from IEEE and/or ACM/SIGGRAPH?

 

EDIT: Just to clarify, this isn't research, just a computer graphics seminar where I implement a rendering technique for demonstration.

Edited by jammm

Share this post


Link to post
Share on other sites
Advertisement

wouldn't it make more sense if you find one or more interesting papers from those journals and implement them ?

Taking existing/available implementations, adding a couple of references and submit them as your own research is NOT how you get a degree...

Share this post


Link to post
Share on other sites

wouldn't it make more sense if you find one or more interesting papers from those journals and implement them ?

Taking existing/available implementations, adding a couple of references and submit them as your own research is NOT how you get a degree...

It's not really research, just a presentation. As I said, it's just a seminar. Nowhere in the post have I specified that this is research.

Edited by jammm

Share this post


Link to post
Share on other sites

 

wouldn't it make more sense if you find one or more interesting papers from those journals and implement them ?

Taking existing/available implementations, adding a couple of references and submit them as your own research is NOT how you get a degree...

It's not really research, just a presentation. As I said, it's just a seminar. Nowhere in the post have I specified that this is research.

 

 

The following advice is if you're genuinely interested in learning here, if your priority is to pass the seminar as simply as possible than I guess you disregard it.

 

The point of these exercises is typically to have you research a topic you're unfamiliar with and show that you're capable of picking up knowledge and understand what's going on on your own (by crawling through refs, etc...). The post 2012 criterium is probably there specifically because most pre 2012 algorithms have simple tutorials or at least sample projects available that you could copy, or which do a large part of the "understanding" for you and give it to you in a simpler and more digestible manner, at the cost of depth and accuracy. 

 

That may not seem like a big thing, but if you're unfamiliar with it, there's quite a difference between crawling papers to understand and implement something, and reading a tutorial and implementing what's presented there. But paper crawling is important if you don't want to wait around for others to explain stuff to you when new papers come out. Just as an example, most tutorials you find will go fairly easy on the math and pretty much always sweep something under the rug that's extremely important to get a detailed understanding of the relevant field, as opposed to just that specific algorithm that's presented in a tutorial.

 

Also it's kind of lame to implement something like the basic SSAO algorithm that Crytek came up with when lots of improvements in that area have been made since.

 

But now, some practical help: This thesis from 2013 is a pretty good overview over some relatively recent advances in SSAO (it compares six specific algorithms).

Share this post


Link to post
Share on other sites

Regarding DOF/Bokeh: you should most definitely check out this presentation from CryTek (from 2013), and possibly also this 2014 presentation about COD: AW. In my opinion, the purely gather-based approaches are more scalable in terms of performance, and can give some really great results. The first presentation also has some good references to earlier papers that you can check out.

Share this post


Link to post
Share on other sites

 

 

wouldn't it make more sense if you find one or more interesting papers from those journals and implement them ?

Taking existing/available implementations, adding a couple of references and submit them as your own research is NOT how you get a degree...

It's not really research, just a presentation. As I said, it's just a seminar. Nowhere in the post have I specified that this is research.

 

 

The following advice is if you're genuinely interested in learning here, if your priority is to pass the seminar as simply as possible than I guess you disregard it.

 

The point of these exercises is typically to have you research a topic you're unfamiliar with and show that you're capable of picking up knowledge and understand what's going on on your own (by crawling through refs, etc...). The post 2012 criterium is probably there specifically because most pre 2012 algorithms have simple tutorials or at least sample projects available that you could copy, or which do a large part of the "understanding" for you and give it to you in a simpler and more digestible manner, at the cost of depth and accuracy. 

 

That may not seem like a big thing, but if you're unfamiliar with it, there's quite a difference between crawling papers to understand and implement something, and reading a tutorial and implementing what's presented there. But paper crawling is important if you don't want to wait around for others to explain stuff to you when new papers come out. Just as an example, most tutorials you find will go fairly easy on the math and pretty much always sweep something under the rug that's extremely important to get a detailed understanding of the relevant field, as opposed to just that specific algorithm that's presented in a tutorial.

 

Also it's kind of lame to implement something like the basic SSAO algorithm that Crytek came up with when lots of improvements in that area have been made since.

 

But now, some practical help: This thesis from 2013 is a pretty good overview over some relatively recent advances in SSAO (it compares six specific algorithms).

 

I agree with you, I do intend to learn from implementing these algorithms, but the reason why I'm going for a simple-to-implement algorithm is so that I could first learn how that works, then implement a paper which improves upon that algorithm. Also, I only have a month to prepare for this, so I have to be realistic with my own goals and expectations by the end of the month.

 

Thanks for the thesis link! I'm sure I'll be having a great time going through it, understanding the algorithms and, if possible, implement one of those algorithms as part of my demonstration!

 

 

 

Regarding DOF/Bokeh: you should most definitely check out this presentation from CryTek (from 2013), and possibly also this 2014 presentation about COD: AW. In my opinion, the purely gather-based approaches are more scalable in terms of performance, and can give some really great results. The first presentation also has some good references to earlier papers that you can check out.

Thanks alot for the links! That's some great information right there smile.png

I also found a paper here, which talks about an advancement in DOF/Bokeh, offering better performance than normal non-seperable filtering. It seems to be cited in the Crytek presentation too. The plan is that I'll be learning DOF/Bokeh from your samples, implement them on my engine and try to implement this paper after that. I hope you don't mind me referencing your approach while implementing DOF/Bokeh on my engine.

 

I'll be sure to cite your blog and the paper I found on my report.

Edited by jammm

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

GameDev.net is your game development community. Create an account for your GameDev Portfolio and participate in the largest developer community in the games industry.

Sign me up!