Unreal Engine 4's post on this
https://www.unrealengine.com/en-US/blog/unreal-engine-4-supp...
Announcing Microsoft DirectX Raytracing
111–120 of 235 posts
Re: Announcing Microsoft DirectX Raytracing
#112Earlier quoted context omitted.
This doesn't even look like it runs at a full 60 frames per second unless they are speeding it up in the editing.
I haven't seen evidence one way or the other, but it's specifically labeled "real-time GPU raytracing." Do you have evidence it doesn't run at 60 fps?
Well, the video is clearly dropping frames so obviously not maintaining 60fps...
Re: Announcing Microsoft DirectX Raytracing
#113Earlier quoted context omitted.
Even in AAA games of today with amazing graphics, I still see polygons in cylindrical, conical and round objects (I'm not talking about raytracing but regular rendering). Everything looks so realistic, but a well or bucket [1] or some other cylindrical object you encounter looks not round but faceted, breaking the suspension of disbelief. Raytracing may bring more roundness, but why not quadratic surfaces or some oth…
That's just because the industry is stuck using triangles in the art pipeline (and don't put in as much effort into the low-polys as they used to). If they were willing to use some variation of patch-based pipeline so that they could utilize the tessellation hardware that has existed for years, it wouldn't be a problem. I've heard artists are scared of it due to how amazingly awful some editors managed to make the NU…
> I've heard artists are scared of it due to how amazingly awful some editors managed to make the NURBS editing experience, but with a decent editor it really wouldn't be a problem.
That's simply not true. I started getting into (high end) 3D in the early 90's. I've probably used every NURB or higher order surface modeling tool under the sun since then.
In fact, there are many amazing modeling tools for working with NURBs or other higher order surfaces.
They just all don't even come remotely close to polygon modeling when precision is secondary and workflow/ease of use is paramount.
So the answer in the VFX world, for years, has been subdivision surfaces. They have almost all of the good of bi-cubic patch modeling and almost all of the good of polygon modeling. What's more, existing polygon modeling tools can easily be upgraded to subdivision surface modelers by merely enforcing 2-manifold topology and adding the ability to display an [approximation to] the limit surface in real time.
Some of the schemes have nice properties. For example, after one step of Catmull-Clark, the entire surface consists of quads. And when treating each local grid of quads as the control polyhedron of a cubic b-spline patch (not a NURB but an UNRB, a uniform, non rational b-spline patch), the surface of the patch is equal to the limit surface that would be obtained by the subdivision scheme.
Re: Announcing Microsoft DirectX Raytracing
#114https://www.youtube.com/watch?v=LXo0WdlELJk I'm pleased to see that the tech video includes spherical mirrors. As I understand it, all raytracing demos are obliged by universal law to include at least three reflective balls in any promotion of the technology. One day, someone will figure out a game where these super-shiny ball bearings are a critical part of the gameplay, and at that point, raytracing will finally ta…
Even in AAA games of today with amazing graphics, I still see polygons in cylindrical, conical and round objects (I'm not talking about raytracing but regular rendering). Everything looks so realistic, but a well or bucket [1] or some other cylindrical object you encounter looks not round but faceted, breaking the suspension of disbelief. Raytracing may bring more roundness, but why not quadratic surfaces or some oth…
would you prefer world's greatest virtual concrete slab? https://techreport.com/review/21404/crysis-2-tessellation-to...
Re: Announcing Microsoft DirectX Raytracing
#115Earlier quoted context omitted.
I don't know why you're getting downvotes. DirectX is harmful for consumers and only serves to lock in games and users to Microsoft Windows.
Microsoft is the dominant platform for games. What financial gain would they derive from pursuing a non lock in strategy?
Sometimes I just wish Valve* would pop up and announce: 'oh so HL3 is coming out soon, and it will be a Steam Linux exclusive for the first 6 months.' While not likely, something like that would be a really interesting gauge of how likely genuine Linux adoption could be...
* I use Valve due to their apparent Linux push with SteamOS (what's the go with that btw!?)
Re: Announcing Microsoft DirectX Raytracing
#116Earlier quoted context omitted.
Even in AAA games of today with amazing graphics, I still see polygons in cylindrical, conical and round objects (I'm not talking about raytracing but regular rendering). Everything looks so realistic, but a well or bucket [1] or some other cylindrical object you encounter looks not round but faceted, breaking the suspension of disbelief. Raytracing may bring more roundness, but why not quadratic surfaces or some oth…
(I do geometry but I'm not in the gaming field) I'd guess that this is because the fastest way to compute a precise silhouette of an arbitrary shape (like quadratic surfaces) is to do a ray test for every pixel that might hit the surface. You could project the surface onto the image plane to get boundary curves, but (a) determining the curves that make up that silhouette is a complicated geometric process that's pron…
> By the way, modelling everything as an analytic surface rather than as a discretized polygon mesh is called "Boundary Representation" or b-rep
I thought that was iso-geometric analysis? What is the difference?
Re: Announcing Microsoft DirectX Raytracing
#117https://www.youtube.com/watch?v=LXo0WdlELJk I'm pleased to see that the tech video includes spherical mirrors. As I understand it, all raytracing demos are obliged by universal law to include at least three reflective balls in any promotion of the technology. One day, someone will figure out a game where these super-shiny ball bearings are a critical part of the gameplay, and at that point, raytracing will finally ta…
My understanding is that rasterization really made portal hard to make. They had to jump through a bunch of hoops to make it all work. Meanwhile I understand that adding portals to raytrace quake was like an afternoon's exercise. I'm also annoyed that there's a map in overwatch where the corner mirrors don't actually reflect anything :P I think generally with many of the tricks that have come out of SIGGRAPH for the…
> Not to sound too crazy but I also have a strong suspicion that if you tied a physics engine to a raytracer, you might get your physics 'for free' which could be a big deal for pushing forward more realistic physics too.
Why is that?
Re: Announcing Microsoft DirectX Raytracing
#118Earlier quoted context omitted.
In the days of ML based AI, seeing raytracing + revolution is both refreshing and funny :)
Actually, raytracing is a domain with strong potential for AI, as there is a lot of work using ML models to de-noise ray-traced scenes in real time. Edit: Yeah, NVidia actually just announced a big project on that exact subject: https://www.youtube.com/watch?v=jkhBlmKtEAk
Pixar is also using Machine Learning to denoise path traced images: http://graphics.pixar.com/library/MLDenoisingB/paper.pdf
Re: Announcing Microsoft DirectX Raytracing
#119The only issue is that a lot of this is already possible in the current renderers at a reduced cost. - Soft shadows work in WebGL fairly well using PCSS: https://clara.io/player/v2/8f49e7c3-7c5e-43f0-a09c-33a55bb1b... - Translucency can be faked effectively using a few different methods. https://clara.io/view/5c7d28c0-91d7-4432-a131-3e6fd657a042 - Screen space ambient occlusion, if using SAO, is amazing. SAO test: ht…
> a lot of this is already possible in the current renderers at a reduced cost.
Maybe this misses the point of standardizing an API? Isn't there considerable value in having a simple framework that can achieve all those effects, and be shared and communicated easily with others? I'm sure a lot of studios would welcome faster dev times, fewer authoring tools, and fewer geometric limits for their rendering effects.
Also worth considering is that all those specific tricks are very limited, either in applicability or in quality or both. Screen space reflections only work on flat surfaces. PCSS still needs a hand-tuned shadow map. DOF has problems at image space contact points. Etc. Let's not oversell today's workarounds as so good they should be preserved forever.
> even if you have a better way of doing things, if it takes more CPU/GPU cycles that a hack that achieves basically the same quality, it will not be adopted. It is that simple.
That's true for any one specific game on any one specific platform. OTOH, that framing misses the trend: there has been consistent and very strong pressure towards higher fidelity, more realism, better physics, faster processors, higher resolution, and overall larger numbers of everything, ever since the first video game ever made. Ray tracing might not be adopted today, but one thing I will guarantee is that next year the CPU & GPU cycles used for rendering will exceed today, and it'll be even more the year after that. When I think about that, I feel like seeing ray tracing take over in games is inevitable.
Re: Announcing Microsoft DirectX Raytracing
#120Earlier quoted context omitted.
I don't think pure raytracing and ray traced global illumination are going to be adopted by many games if there is still such a giant problem with noise. That is far from a solved problem, even in multi-hour per from visual effects renders.
This paper shows pretty good results from essentially throwing a RNN at it: [1] http://research.nvidia.com/sites/default/files/publications/...
Where will the 1000 frame fly through and the 10 noisy frames that they use come from? I find that paper very disingenuous when they don't count these things in their filtering time.
Not only that, but they compare everything with "1 sample per pixel" knowing that the SURE filter uses per pixel statistics of variance that require more than one sample per to start working.