Timeline for Can someone explain dual contouring?
Current License: CC BY-SA 3.0
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 3, 2021 at 19:36 | comment | added | ma1169 | @Somnium assuming you still need it, if I am not mistaken to generate a normal, do this for nx= (Density(x + 1, y, z) - Density(x - 1, y, z)) / 2 ,ny= (Density(x, y+1, z) - Density(x, y - 1, z)) / 2 , nz= (Density(x, y, z+1) - Density(x, y , z-1)) /2 an explanation to why this work here | |
| Jun 3, 2021 at 16:17 | comment | added | ma1169 | I had hard time understanding what jmegaffin meant. the way I understood it is this, it's the ratio between two vertices of an edge with sign change based on their density, so if the density at end A of the edge is 0.2 and at end B is -0.3, p would be placed 40 percent of the way from A to B. | |
| Mar 22, 2018 at 16:21 | comment | added | Somnium | How do you generate normals at quad vertices? | |
| Sep 30, 2014 at 14:21 | vote | accept | Soapy | ||
| Sep 22, 2014 at 17:47 | comment | added | jmegaffin | Not every edge is going to have an intersection, so you aren't necessarily solving the QEF for 12 planes. Other than that you've got it! | |
| Sep 22, 2014 at 17:12 | comment | added | Soapy | So basically for each cube I calculate the p and n for each 12 edges, and then run the QEF for each cell edges p and n into x, which in your example is at the centre of a voxel/cell to start with? Then if four cells share the same edge, I create a quad connecting each four cells x? And is the resultant quad my polygonal data? | |
| Sep 22, 2014 at 14:50 | comment | added | jmegaffin | p is found by finding where the weighted average of the two densities along the edge is zero. Then you calculate n by taking the gradient of your density function at p. | |
| Sep 22, 2014 at 14:43 | comment | added | jmegaffin | x is the vertex position. For each intersection point you have p (the position) and n (the normal), which make up the planes that I was talking about. | |
| Sep 22, 2014 at 11:47 | comment | added | Soapy | So lets say I have a cell/voxel that I know exhibits a sign change (i.e a case like MC), I have ran a noise function for each 8 corners of the cell to find its density. What I'm having trouble understanding is from this how do I find the variables x, n and p of the the QEF? | |
| Sep 21, 2014 at 10:16 | history | edited | jmegaffin | CC BY-SA 3.0 | deleted 10 characters in body |
| Sep 21, 2014 at 10:08 | history | answered | jmegaffin | CC BY-SA 3.0 |