Substance Painter | Weird artifacts after baking

CuriousAn124

I've baked my model from High poly in SP, but it has weird artifacts of chamfer in edges.

I've made UV seams in hard edges, but this didn't help. There should be Soft chamfer, but instead there is a weird hard and soft chamfer mix. 

There are some info about model in pinned photos:
1st photo: chamfer artifacts in SP
2nd:  High poly model topology (in blender)
3d: Low Poly (with soft shadows)
4th: UV seams

Do you know how to fix it?

CuriousAn124

And there is some defolt bake settings. (I raised the value of Max rear distance through advice in one tutorial)

Fabi_G

What do the UVs look like? Any particular reason why you don't use hard edges at steep angles along UV seams? Does it improve when you bake with a higher resolution? You could zip the meshes containing this part and attach to a post for people to take a look.

CuriousAn124

Fabi_G said:

What do the UVs look like? Any particular reason why you don't use hard edges at steep angles along UV seams? Does it improve when you bake with a higher resolution? You could zip the meshes containing this part and attach to a post for people to take a look.

Really, my bad, forgot UV screen (attached -- 1st photo).
I didn't use hard edges, cause i want to have soft forms on this part like on my High poly. Or do i must add hard edges in each UV seams? (it's my first bake, don't hit me
Indeed, increasing the quality of the resolution helped, but didn't solve the problem completely. I still see the artifacts wit 1K res (2nd screen), and the borders with 2K is slightly visible (3d one).

P.s: added blend and fbx files in rar

CuriousAn124

I tried to set hard edges along UV seams, didn't help(

Fabi_G

Looking at the model, the angles on the inset are less steep than I thought. In that case, I would not split UVs there, only for the edges with steep angles. I attached a modified lowpoly + OpenGl normalmap (added 2 more subdiv levels to highpoly for bake). Hard edges (require split UVs) can be used to adjust the lowpolys shading so the Normal map has to compensate less (less gradients). Aligning UV borders with the pixel grid will reduce visible artefacts too, especially at low resolutions (https://helpx.adobe.com/substance-3d-bake/common-issues/aliasing-on-uv-seams.html). Maybe some of the threads collected here are an interesting read for you as well: https://polycount.com/discussion/163872/long-running-technical-talk-threads, specifically "You're making me hard. Making sense of hard edges, uvs, normal maps and vertex counts".
Edit: Added some links

CuriousAn124

Fabi_G said:

Looking at the model, the angles on the inset are less steep than I thought. In that case, I would not split UVs there, only for the edges with steep angles. I attached is a modified lowpoly + OpenGl normalmap (added 2 more subdiv levels to highpoly for bake). Hard edges (require split UVs) can be used to adjust the lowpolys shading so the Normal map has to compensate less (less gradients). Aligning UV borders with the pixel grid will reduce visible artefacts too, especially at low resolutions (https://helpx.adobe.com/substance-3d-bake/common-issues/aliasing-on-uv-seams.html). Maybe some of the threads collected here are an interesting read for you as well: https://polycount.com/discussion/163872/long-running-technical-talk-threads, specifically "You're making me hard. Making sense of hard edges, uvs, normal maps and vertex counts".
Edit: Added some links

Yes, that's work, it was very helpful. Thanks a lot