Hello, im thinking about how to solve a workflow of scanning a ground (whatever- forest, pavement, cobble stone road etc.) into a 3D mesh with textures using photogrammetry.
I want to take the resulting 3D mesh and generate: diffuse, displacement and normal map. so far so good, but i would like to make ALL the maps (diffuse, normal, displacement, maybe reflection etc.) tileable...
I know how to make a tileable texture in photoshop, but im not sure how to make all 4 maps tileable and still matching each other, i would have to paint/clone/stamp ALL the maps at the same time, but im not sure how to do that and in what program to do that...
Any Ideas how to achieave this?
Replies
Like you say, unfortunately you cant clone many layers at once, I tried recording all my strokes in photoshop but that didnt work so good (photoshop dont record brush size changes) and very slow
The photshop document tend to be quite big so I use colors on code on layers, red= diffuse, blue= normal and so on so I can filter the different texture components easily.
When done, show/filter only diffuse, hide the rest and ctrl+shift+e to merge/copy into a new layer. repeat for rest, normal, disp and so on..
Offset these textures, group and layermask away visible seams.
Be careful rotating the textures around when it comes to the normal map
i must figure out a workflow for this.
There must be a workflow for directly editing 3 or 4 maps at once, to get them tileable.
OR Mari might have such functionality...? (it should be the most advanced painting app as far as i know)
There is a node to automatically make your textures tileable, use it on all your channels with the same settings and you're done.
Or you can paint a simple mask to mask between the original texture and an offseted one and drive the blending of the other maps the same way.
The Situation: We have a Scaned Wall,so we have a HighPoly Wall, and this HighPoly Wall we want to bake on a Plane, but that all Maps are tileable.
Interesting, would love to see an example as well.
Been playing a bit with photogrammetry, but haven't gotten any good results yet. Will try again when it gets cloudy and see if I can setup a test in SD5.