title.
#Do overlay cameras receive GI + Light probe info?
broken flame
In what context? Or what do you mean by "info"
lets say i have a main camera and a viewmodel camera, if i then have mixed lighting from the perspective of the main camera the non static objects should be affected by lighting changes caused by the light probes, but from the perspective of the overlay camera its as if any lighting changes caused by the light probes dont actually appear.
storm zealot
lets say i have a main camera and a viewmodel camera, if i then have mixed light...
Light probes are per scene, so they do affect objects rendered by the overlay camera
Not sure if there's exceptions when they would not
Realtime shadows from the mixed lights wouldn't carry over, as shadowcasting is per-camera
broken flame
not sure what causes this then
viewmodel cam is on the left regular cam is on the right
the viewmodel one simply doesnt pick up the shaded area from the lightprobe
@storm zealot
storm zealot
the viewmodel one simply doesnt pick up the shaded area from the lightprobe
You sure the darkness is from the light probe? With mixed lighting it should be a realtime shadow, and looks like it is
broken flame
i mean regardless the issue persists no?
i guess its not more so about my initial question but more about how do i actually solve the viewmodel camera not updating its lighting correctly
storm zealot
i guess its not more so about my initial question but more about how do i actual...
I'm not sure if there is a very practical way to copy the shadows from the world camera to the viewmodel camera
But more importantly it won't fix all your issues
broken flame
is there any better alternative than using a seperate viewmodel camera then
im guessing thats the weakest link here
storm zealot
I prefer the camera relative scale + "squash" method
https://www.youtube.com/watch?v=pM2Lr5roSPc&t=271s
https://youtu.be/11sLIyw0pWQ
This video is for Unreal but it explores all the same techniques and challenges you might face using various methods
And I think ultimately they chose the scale and squash method too
When you "fix" lack of shadowcasting you end up getting issues with intersections (and any connections) between world and view models
This wasn't really an issue in older games where these floaty parallel dimension viewmodel method originates from, when their lighting was much simpler
Usually only based on vague probes
Whereas modern high fidelity games like CoD and BF have much more control over the engine and expertise at their disposal to solve these issues
So I recommend not chasing that exact result
Rather if you don't want to accept compromises in lighting, make the viewmodel more physical and have it dodge out of the way of obstacles more
storm zealot
Maybe try a less extreme FoV difference between the viewmodel and world model too to make them a bit more seamless
broken flame
the metro one is really interesting and i was already thinking of something similliar
storm zealot
It's a different direction in style, but maybe not unwelcome one
I for one don't think the very flat and disconnected viewmodels of modern shooter games feels real or immersive at all