I found the use of bump and displacement maps to simulate texture pretty fascinating, and looked more into them. Bump mapping creates the illusion of texture by changing light and dark values based on the height of a certain point. However, bump mapping doesn’t affect the actual surface of the object, leaving its silhouette and shadow unchanged. The illusion that bump maps create also breaks if the object is viewed from an angle that exposes its flatness. We’ve developed improved forms of bump mapping like normal and parallax mapping. Normal maps, which have more or less replaced bump maps in the industry, alter light and dark values based on the point’s direction as well as its height. Normal maps generally depict shading more realistically than bump maps. Parallax maps displace individual pixels to offset the texture map, emphasizing height differences in objects like brick walls. Meanwhile, displacement mapping changes the actual geometry of the object by editing its mesh. Displacement maps use significantly more computational power than bump maps, so they’re generally reserved for large details. Bump maps are often used for fine details such as skin pores, where editing a mesh to account for each individual pore would take unreasonably long.
anavmehta12
My intuition with how bump mapping works is that bump mapping doesn't actually change the geometry of the object just slightly adjusts the normal based off the values of each pixel altering how light reflects off the points. So higher peaks are brighter and whiter and darker areas are parts of the surface that are indented.
theflyingpie
I noticed that in the bump mapping render, the edges of the ball are still appear smooth like the image on the far left. This seems consistent with the fact that bump mapping only perturbs the direction of light reflecting off the object rather than changing the shape of the object itself.
jerrymby
Why are the displacement mapping more difficult? Is it because the lighting calculation is way messier than just adjust the normal direction?
I found the use of bump and displacement maps to simulate texture pretty fascinating, and looked more into them. Bump mapping creates the illusion of texture by changing light and dark values based on the height of a certain point. However, bump mapping doesn’t affect the actual surface of the object, leaving its silhouette and shadow unchanged. The illusion that bump maps create also breaks if the object is viewed from an angle that exposes its flatness. We’ve developed improved forms of bump mapping like normal and parallax mapping. Normal maps, which have more or less replaced bump maps in the industry, alter light and dark values based on the point’s direction as well as its height. Normal maps generally depict shading more realistically than bump maps. Parallax maps displace individual pixels to offset the texture map, emphasizing height differences in objects like brick walls. Meanwhile, displacement mapping changes the actual geometry of the object by editing its mesh. Displacement maps use significantly more computational power than bump maps, so they’re generally reserved for large details. Bump maps are often used for fine details such as skin pores, where editing a mesh to account for each individual pore would take unreasonably long.
My intuition with how bump mapping works is that bump mapping doesn't actually change the geometry of the object just slightly adjusts the normal based off the values of each pixel altering how light reflects off the points. So higher peaks are brighter and whiter and darker areas are parts of the surface that are indented.
I noticed that in the bump mapping render, the edges of the ball are still appear smooth like the image on the far left. This seems consistent with the fact that bump mapping only perturbs the direction of light reflecting off the object rather than changing the shape of the object itself.
Why are the displacement mapping more difficult? Is it because the lighting calculation is way messier than just adjust the normal direction?