they can still look close during gameplay, when all the stars are aligned
One superb piece of shadowing I saw last night (up to Chapter 8) was soft shadowing when climbing a tower. It was very realistic with softness and darkness adjusting based on distance of different limbs to the surface. So as the knee got closer to the wall, the shadow under it got tighter and darker, but it wasn't just a uniform shadow change. Very good, and confusing as to why that shadow was so awesome! Is there a special Nate Climbing Shadow Solver?U4's SSAO is one of the simplest and cheapest this gen...
Some kind of mellow screen-space AO, plus some objects (notably characters) have simplified capsule representations used for occlusion. For environmental lighting, a simple sort of cone trace is used to test the extent to which a incoming light is being occluded by nearby capsules.anybody know what they were using for TLOU last gen?
Sounds like the capsule occlusion.One superb piece of shadowing I saw last night (up to Chapter 8) was soft shadowing when climbing a tower. It was very realistic with softness and darkness adjusting based on distance of different limbs to the surface. So as the knee got closer to the wall, the shadow under it got tighter and darker, but it wasn't just a uniform shadow change. Very good, and confusing as to why that shadow was so awesome! Is there a special Nate Climbing Shadow Solver?
Yep. First seen in The Order, out of the games I've played, I'm sure it was used elsewhere.Some kind of mellow screen-space AO, plus some objects (notably characters) have simplified capsule representations used for occlusion. For environmental lighting, a simple sort of cone trace is used to test the extent to which a incoming light is being occluded by nearby capsules.
Sounds like the capsule occlusion.
That, as other's have said, is the result of the character's capsule proxies. It is not exclusive to nate climbing, nor to that tower. It is literaly everywhere in the game, you just had not noticed.One superb piece of shadowing I saw last night (up to Chapter 8) was soft shadowing when climbing a tower. It was very realistic with softness and darkness adjusting based on distance of different limbs to the surface. So as the knee got closer to the wall, the shadow under it got tighter and darker, but it wasn't just a uniform shadow change. Very good, and confusing as to why that shadow was so awesome! Is there a special Nate Climbing Shadow Solver?
Also on the subject of animation that everyone's raving about, it's still very computer gamey. Characters walking on the spot, sliding, glitchy transitions from one position to another aligned with the scenery features. It's improved on previous UC's from having more variety etc., but it still looks fundamentally the same at is core. Given the general enthusiasm for the animation I was expecting more. Heck, even the cutscene presentation is mediocre in results.
Details and vistas are definite pluses. I can agree with on those points.
Here some closeup details of the fabrics on main characters. Using 4k high-res capture.
Using Disney Diffuse Model, added special cheap Sub-surface Scatter. We also added fabric micro details, small wrinkles and aging.
For Specular models, we are using kajiya-kay model for silk fabric, and modified the fabric model from Ready at Dawn for cotton, wool, ext...
Models of the characters showing in the post are done by our talent Character team.
Real-time rendered in game.
If you guys want to know more details, I will give a talk at Siggraph this year July 2016.
Here some closeup details of the hair on main characters. Using 4k high-res capture.
Here are some features we used to create the volumetric look of the hair, realtime in-game:
1. Using baked shadow map and wrap diffuse to fake the shadow. It helps us to get the feeling of depth and volume of the hair.
2. Scatter is really important for blonde hair to look right. It’s because the lights scatter between hair strands.
We can’t afford ray-tracing, of course, so we approached the similar result by wrapping the lights and offsetting the dot(N,V).
3. We used typical kayjiya-kay specular model for the specular of the hair. We also put hair id maps to masked out the offset value of the spec.
Models of the characters showing in the post are done by talent Character team.
Frank Tzeng worked on Drake, Sully and Nadine, Soa Lee worked on Elena, Colin Worked on Sam, Jaehoon worked on Rafe.
If you guys want to know more details, I will give a talk at Siggraph this year July 2016.
That kind of goes without saying when a game is using a baked lightmap, though. Even if you go back to the 90s; Quake 2 used a radiosity solver, for instance.Also the game has tons of globally lit areas, the lighting is static though, but it is definitely using global illumination.
Not necessarily. Baked lightmaps can pose challenges with memory management, and incident lighting data being baked doesn't mean that what the game does with it is cheap. Running complex lighting data through fancy reflection functions can be very expensive.It's definitely a cheap solution
There will always be more possible coherent samples to use in the result for a single scene state with baked lighting systems.there will always be more possible detail to extract with baked lighting systems.