Skip to content

Instantly share code, notes, and snippets.

@aras-p
Created December 23, 2015 07:19
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save aras-p/4683338f9cd64068cf6a to your computer and use it in GitHub Desktop.
Save aras-p/4683338f9cd64068cf6a to your computer and use it in GitHub Desktop.
different lighting in 5.2 vs 5.3
So, https://twitter.com/gav_clayton/status/679532625321590784
I'm discussing it with people who actually did Standard shader changes in 5.3, but they are all aways for the holidays right now :)
To me it seems that the change is *not* caused by the GGX specular, but rather by the change to how we treat "smoothness" of the Standard
shader. I'm told it matches much more closely with Substance Painter and other tools now, but that does mean that for fully rough
surfaces the specular is more visible or something.
I think it might take until after holidays to figure out whether our 5.3 changes are actually correct or what, but in the meantime...
It seems that in your particular scene using Standard shader is basically an overkill? All the materials I see are fully rough, and
fully non-metals. It's almost as if using Legacy Shaders/Diffuse would be both simpler and work much faster (due to much simpler
lighting calculations). Heck, even using that and your 3-light rig might be overkill, I'd probably just have a custom shader that does
all 3 directional lights in a single pass, evaluating them per-vertex.
@Interkarma
Copy link

Thank you Aras for taking the time to investigate. I replied to you on Twitter but thought I should respond here also so gist isn't left hanging.

Use of Standard shader is for modders to create HQ material mods (some of which already exist) and to support fully procedural GI if that ever eventuates. Please consider that Daggerfall Tools for Unity is itself an open source platform for developers to remix and extend. I want it to grow along with the Unity engine and take advantage of what Unity offers. It's not so much overkill as an investment in the engine's features and future.

On Twitter, you suggested I try a non-zero smoothness to bring the result back towards 5.2. This does help, but only ever-so-slightly. The surface still retains the specular-like shine and colours feel flat and overexposed. Anyway, with a bit of luck this is something the Standard shader folk can answer once they return from holidays. :)

I hope you have a wonderful Christmas and new year, and wish you all the best for 2016!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment