So here’s my issue. I have a geosphere that has a material on it. The material has a tessellation map on it, which is fed drawing information from my blueprint. The drawing function is nearly identical to the content example using HUD, a brush material and a 2D render target. It is absolutely fully functional with ONE hiccup. While the UV coordinates are mapped correctly, and the correlation to the sphere work perfectly - the point on the sphere on which the “drawing of a hill” meets the end of the 2D render target’s UV position doesn’t carry over. To put it bluntly, the radius of my brush doesn’t cross correctly from (0.999, 0.999) to (0.0, 0.0). Until my mouse crosses that line an edge form on my sphere.
SO… how do I carry over pixel information in the form of UV information for each pixel within the brush’s radius on the 2D render target from within a blueprint?
Well, actually I tried that. Here’s a bit of an issue.
The way users (in any 3d application) use a mouse has a certain distance between space when the mouse is moved even a little. This changes depending on the depth, but from 2d to 3d some places the mouse is moved to ‘skips’ or ‘jumps’ to the end position you eventually leave your mouse at. Like rocks skipping over a lake, eventually the distance shrinks and the rock/mouse comes to its final resting place. The issue with the seaming isn’t actually to do with drawing or the texture - but house the UV information is relayed from the world vs how mouse 2d information is relayed to the user.
What I did was create a 3 by 3 grid of UV collision detection (rather than just the one) around the point clicked. I made some logic to wrap the blocks that pass a value of 1.0 back to 0.0. There is a distance value between the points… imagine one block in the center 8 blocks around it. The distance pushes the 8 blocks out from the click location and performs the same task as the center block. This method doesn’t work. I can’t make tiny details and sometimes the aforementioned problem still happens.
