3D animated films.
lol I did say SSR uses 2D ray-traces, which naturally means they're limited to the 2D screen space. I was citing their example to point out that vector calculations are performed using similar methods to those that are applied to 3D ray-traced reflections. Moreover, 3D ray-traced solutions have existed in non-real-time applications (see above) and also run on the same fundamental algorithms.
So, again, my questions is what do you mean by "in the form that it is today"? How is the underlying algorithmic logic behind ray-tracing in Pixar movies from a decade ago different from that of a real-time application like Metro Exodus?
And yet, developers who strive for such experiences on consoles today are doing just fine. Yes, the high degree of object interactivity and physics in Breath of the Wild doesn't offer a corresponding physically accurate behavior wrt to lighting. But I certainly don't care. If there is a cheap alternative, more power to it. If not, I don't see its value over the lighting that's already there in the game.
Approximated collisions aren't perfect, but they get the job done most of the time. The only other alternative is collisions based on the meshes themselves, which isn't performance friendly.
This draws an interesting parallel with the discussion at hand. Mesh-based collisions are physically accurate, but they're taxing on hardware. Approximated collisions using simpler shapes with lesser vertices are great for performance, but they're not as accurate. However, if the artist baking approximated collisions does a good job with their placement, then the end result is just as pleasing (not visually, obviously
).In the case of Metro Exodus, yes, but artistic quality is dependent on the developer. Remember how CDPR dumbed down the Witcher 2's lighting model so it could run on the the Xbox 360. The end result on the X360 version looked more visually appealing to me. They did a pretty decent job there.
Of course, I don't deny that they have limitations and don't always produce a superior result. It's why devs use a mix of both baked and dynamic lighting.
I acknowledged that it's a great technological advancement at the very start of this discussion and that it's great for the PC space.
Can't have everything on a limited budget.
I listed more complex shaders as something that does and will continue to produce visually pleasing results. You aren't paying attention
At the moment, these examples come to mind.
- Breath of the Wild, a game that takes physics-based puzzle solving and interactivity to another level. Geometry was on the simple side, which is how they pulled it off on mobile hardware. Can't wait for the next Zelda game to continue this trend of emergent game design.
- Dynamic cover in shooters that is both destructible (Battlefield series) and/or reacts to world physics (Uncharted series).
- AI interactions and world simulation in Red Dead Redemption 2.
Lighting doesn't have to look physically accurate for a game's physics to behave in a physically correct manner. You're missing this essential point. So, no, to have better physics and interactivity, you certainly don't need ray-traced lighting. There are other cheaper real-time solutions that can work. lol that Frostbite engine slide is literally talking about graphical features, not gameplay.
Inverse kinematics is most certainly a game changer for the series, as it wasn't there in previous DMC games. It improves playability and traversals over complex geometry. So, yes, why wouldn't it be a game changer? Uncharted 4/LL also make great use of it.
Why did CDPR switch from canned movement to IK for the Witcher 3? Did it improve playability over The Witcher 2? The use of physically correct behavior for gameplay is way way more important than the use of physically correct behavior for graphical features.
https://www.eurogamer.net/articles/digitalfoundry-2014-the-witcher-3-tech-analysis
Last edited:
