SIGGRAPH 2012
- Thread starter fellix
- Start date
Starts today, right? Here are some of the other trailers
SIGGRAPH 2012 : Computer Animation Festival Trailer
SIGGRAPH 2012 : Emerging Technologies
SIGGRAPH 2012 : Computer Animation Festival Trailer
SIGGRAPH 2012 : Emerging Technologies
For those coming, don't forget this year's Beyond Programmable Shading, which includes an updated version of "5 Major Challenges in Real-Time Rendering" from Johan Andersson, DICE. It's a half day course this year, and as usual we will be posting the slides from the talks.
http://bps12.idav.ucdavis.edu/
http://bps12.idav.ucdavis.edu/
SIGGRAPH 2012 Course: Practical Physically-Based Shading in Film and Game Production slides are up.
http://blog.selfshadow.com/publications/s2012-shading-course/
“CSM Scrolling: An acceleration technique for the rendering of cascaded shadow maps”
http://www.insomniacgames.com/mike-day-siggraph-2012-csm-scrolling/
Beyond Programmable Shading
http://bps12.idav.ucdavis.edu/
http://blog.selfshadow.com/publications/s2012-shading-course/
“CSM Scrolling: An acceleration technique for the rendering of cascaded shadow maps”
http://www.insomniacgames.com/mike-day-siggraph-2012-csm-scrolling/
Beyond Programmable Shading
http://bps12.idav.ucdavis.edu/
Last edited by a moderator:
The slides (with notes) from my presentations are now available:
http://www.humus.name/index.php?page=Articles
http://www.humus.name/index.php?page=Articles
Billy Idol
Legend
Cool, thanks!
Brigade 2 (real time ray tracing engine) demo:
More info here:
http://raytracey.blogspot.co.nz/2012/08/real-time-path-traced-brigade-demo-at.html
More info here:
http://raytracey.blogspot.co.nz/2012/08/real-time-path-traced-brigade-demo-at.html
This example is pathtracing, which involves firing rays in a random direction from where a ray hits a surface (to get global illumination). These directions are chosen randomly (or pseudo-randomly) to prevent even worse sampling artifacts. This randomness means that for two adjacent pixels, a sample for one pixel might hit a light source quickly, and the sample for the other may hit an area of deep shadow, hence the disparity between the two. As the number of samples per pixel increases, the average value of the samples for each pixel approaches the true value and the disparity evens out. It looks worse in motion because of the relatively small number of samples per pixel before the frame is cleared for the new geometry positions, and there isn't any guarantee that a pixel which randomly hit shadow last time won't randomly hit a light this time (or vice versa).
Unreal Engine 4 - The Technology Behind the Elemental Demo
http://www.unrealengine.com/files/misc/The_Technology_Behind_the_Elemental_Demo_16x9_(2).pdf
http://www.unrealengine.com/files/misc/The_Technology_Behind_the_Elemental_Demo_16x9_(2).pdf
Very good presentation. Highly recommended reading for everyone interested in real-time graphics.
Frenetic Pony
Veteran
Great paper, but I wonder if you can't get more coherence, and less revoxelization. At least in terms of re-rendering dynamic object 3 times each frame (even if it is a really low res proxy). I mean, most objects are going to be purely static in nature in relation to themselves. If you could just store the voxel representation of physics object X, say a classic barrel, and just move that representation around in the structure you'd probably get even more speedup.
Apreciated. That is not a bad starting point for next-gen. All lighting in real time (ok, crytek already offers this) with support for multiple shadowd lights thanks to their SVO, and proper GI on everything as a standard. It is a good direction indeed.
Siggraph 2012 talk – Shadows in Games, Practical considerations
http://www.dimension3.sk/2012/09/siggraph-2012-talk-shadows-in-games-practical-considerations/
http://www.dimension3.sk/2012/09/siggraph-2012-talk-shadows-in-games-practical-considerations/
