I have a question for all the game dev's
What do you think about artists using the photoshop normal map plugin to simulate a high poly model. Is this acceptable in the industry? for example a wall texture. Do you
a. Make a really high poly wall and project the normals onto a lower wall.
b. Make a height map in photoshop and use the Nvidia plugin to generate a normal map.
If you've seen the Unreal Tech demo, did epic artists actually make that high poly wall or simply use the nvidia plugin. Same can be said for the textures in doom3 or quake4.
For characters and props, id think that you would get better results by making the high model and projecting the normals. But that would take alot of time, particularly for making high poly props.
What are you thoughts and useful tips when creating normal maps? What method is preferred with the time constraints of production?
Replies
I'm sure epic modeled the walls, i can't see them going cheep and just creating a bump-map and turning that into a normal map...
ya?
Lies. Everyone knows they're cheap SpeedTree-using-Photoshop-filtering-photo-sourcing hacks.
I've been doing some parallax mapping lately, and I'm constantly spending more and more time in Photoshop. 3D is nice for getting the base shape down, something to work off of, but painting is so much faster for the bulk of the work.