Hey guys,
I'm going from a high rez sculpt in zbrush (no UV's) to a low rez mesh.
How would I go about getting the poly paint data from a high rez mesh (no UV's) to the low rez mesh (with UV's)?
I have tried to do a method that involves XNormal.
- Taking the high rez (polypainted)
- Making a blank texture map for it
- Generating GUVTiles
- Hitting New from Poly paint (poly paint is scene using GUVTiles)
- Taking the low rez and the high rez into XNormals
- Baking out the base texture from the high rez to the low rez
This works good, HOWEVER -- I am getting artifacting across my model (because of the GUVTiles being constructed of lots of little...well...tiles), and it's not as clean as I hoped.
What is a high poly paint to low rez workflow? Thanks!
Replies
make sure to set a good ray distance or setup a cage too.
edit:passerby beat me to it
My model is 54 Million polys and trying to export the model as a whole crashes my system -- How would I go about poly painting a model that has many different subtools, and then taking it and baking it using Xnormals onto a retopoed mesh of my model?
If I exported the model in "groups" (so I can actually get it out of zbrush) -- when baking it using XNormals -- will I just have to do a couple of bakes and then take them into Photoshop, and combine them?
so just bring them all into xnormal
or try using decimation master.
xnormal dosnt need to draw the model, so as long as you got enough ram, xnormal should be able to handle it all.
When using decimation master on a subtool that has poly paint information on it, doesn't that destroy the poly paint because it's based on verts?
Thanks again!
I did a small test where I took a poly painted sphere -- then decimated it -- the poly paint just disappeared. Is it hiding somewhere or am I missing something?
Thanks for the replies!
Hmmmmm
http://www.polycount.com/forum/showthread.php?t=84917&page=2
-- I didn't give my high rez generated UV's or bake the poly paint to the texture, or have DM preserve the UV's.
..........NOW I KNOW !
Thanks
Hmph.
Well isn't this an interesting sequence of events.
*UPDATE -- Baller, seemed to keep everything in place thar.
1) first model your lowres cage and do the uvs.
2) now define two or more regions of your model based on uv seams.
easy way to do this is just to assign a different shader to each region. you don't need to cut the model, just define some regions where the uv tiles from each region connect.
(example) region one is all the uv tiles that make up the head and chest, region 2 is the hips and legs etc.
3) in zbrush cut your highres model into pieces that match your lowres regions but overshoot the lowres region a little so the lowres is smaller then the highres. so along the border of each lowres region there is still some surface from the highres.
4) export each lowres and highres region and extract each set of maps.
edit: generally you can just use your full lowres without cutting it.
5) now you can composite all your region maps in photoshop and have your full lowres model map.
just make sure you keep all your seting the same when extracting the maps for each region.
and if you have many different subtools you can generally use the same concept. just find places where you can group objects instead of tiles.