uncertainty with normal map in large bicels

sbstndc

well I have a friend who sent me this, and the question is why in the normal map the bicel appears inverted, both the hi poly and the low poly are correct, and then in any painting program like sutbance painter you can pitnar without problems. Greetings.

here some captures.

pic1lowpoly surface

pic2 lowpoly normal baked in marmoset, here is the question

pic3 normals lowpoly

pic4 substance painter

iam717

flip y in marmo change setting on normal or object to exported app? either can help i think..i 4get 2much 2remember.

pior

@iam717 : this is unrelated.

@sbstndc : unlike any other type of texture, the colors in a tangent space normal map depend completely on the orientation of the UVs.
Make the following test now : take all your UVs, rotate them by any amount of degrees, and bake again. All the colors (except the default blue) will be different.

If anything, such changes in color reveal that your UV layout is probably not the most consistent, with parts unexpectedly rotated.

Overall to preview this (and control it) apply a checker map with colors and numbers and use it at every step of modeling and unwrapping ; and lay out your UVs in a way that would make sense to someone else texturing it manually in a non-3D painting software.

sbstndc

I understand, I just checked, it was just that, the orientation of the surface in space, well thank you very much for answering @pior .

pior

Anytime !
That preview/render looks great btw

Klunk

being a discrete "one off" (not having to be used else where) It could be handled with object space normals. Just a thought either way is good/should give the same results