*@Mathieu Quiblier* Yes, I usually make my own base textures. That gives me the option to use all of the information in between and have exactly the features that I need and most of the time it does not take so long to set them up. Only noise textures are really hard to recreate, so I use those directly, I talk about that in the chapter about noises. There are exceptions though, for example the magic texture is really nice to create cheap and simple textile patterns.

*@Huân Lê-Vương* The problem is that your geometry does not have any actual curvature. To make the most of the pointiness feature, you should add a bevel to the edges, for example with a modifier. You can take a look at the objects in the example scene.

*@Simon Thommes* thanks.

*@Abhirama Gopala Dasa* I am not exactly sure I understand you correctly. The workflow examples are meant to support the main content of the chapters to develop a deeper understanding of the topics that have been touched and draw connections. I go over how to use smooth minimum and maximum to combine value maps in chapter 2.4 about texture composition.

*@Huân Lê-Vương* The lenght operation simply gives you the length of the vector that you plug in. Mathematically that is sqrt(x^2+y^2+z^2), but here that does not really matter. What is important is only that for a coordinate map that gives you the distance from the center, or a radial gradient from the inside out.

*@Simon Thommes* ok, thanks!

(deleted)

(deleted, i know how)

*@Michael Schwarz* When you scale down the cube in object mode, that means that the object coordiantes are squashed or stretched along with it. Sometimes that is useful, but here you want the coordinates to be unaffected from scale. To achieve that you can either apply the scale of the object by pressing CTRL+A or scaling the cube down in edit mode in the first place.

*@Simon Thommes* Thank you very much! I just overlooked the scaling factors for x-y-z I had for my board.

@jamesrossbond You can use the same node, there is no difference in how it works. The reason I am using seperate nodes is for node-tree readability. It is more structured when logically disconnected parts of the shader use separate coordinate inputs, that also allows to change the used coordinates individually later on.

@Leslie Solorzano My guess would be that the object you are using has inverted normals. Have you tried your material on a simple default cube? Otherwise you can also compare the node setup with the one from the example scene.

@Simon Thommes The cube looks the same, I actually checked the normals by showing face orinetation and they were all blue. I imported the material from the .blend that you provide and it looks different, I don't know if it has to do with the size or shape of mi object. But it doesn't matter, the important ting was to learn the method

@Luis Saybe Correct, I forgot to mention that in the video as I simply intended to use the example geometry. Generally it is always a good idea to do scaling operations like this in edit mode and keep the object scale for operations that don't regard the shape of the object itself.