For a while I am looking for some similar to Zbrush 2,5D approach solution that would be allowing to work nondestructively and resolution independent.
I mean with resolution independent layers having depth info for proper Z blending to each other.
Photoshop smart object and linking: Works not that bad but to get proper Z blending you have to work in 16 bit making Photoshop crawl like a snail , linked layers or not + a huge stack of groups, linked side files and blending modes.
Substance Designer: a lot of restrictions, no ready Z blending solution, no crop no well controllable size/stretch , a lot of pain in the ass with a huge node network for simple and routine operations.
Fusion: not very texture work oriented, aliased mask after depth blending. Free version 2048 resolution restriction.
Blender: makes anti aliased mask after depth blending in z combine node but other than that inconvenient like a hell.
3dcoat: does proper z combine in between layers for its depth but not for color/spec info. No anything "smart" and resolution independent.
Vector soft: no 16 bit support for depth info.
Would be appreciated if somebody point if I am missing something and share his/her experience
Replies
I mean a layer of bumpy ground for example, then you are adding a layer of water and could move that water plane back and forth in depth z direction blending with a ground the way ground would sink behind or bumps would appear gradually.
Or maybe having a layer of rocks add another sand one the way the sand would be sitting in deeper parts only.
Surprisingly we could do so in real time shaders but I don't know a texture soft where I could do the same easily.
I need a kind of tool to quickly compose pre-rendered fragments from a library in some simple manner with resulting SSAO and cavity mask and without all those Fusion or Blender composing quirks and aliasing problems.
And re-compose, re-scale quickly when necessary to fix repeating or add another fragments.
By default, height is in additive mode but you can change the blending mode to normal to "replace" the height and the resulting normal map.
I mean, I think it works for your use case.
Also that trick would work only with a flat water plane . Right? What If the second layer have its own bumps (sand example) ?
That will work for any material. We are still missing a min/max blending mode to make this easier and remove the need for a mask but it should already work.
Still I don't understand how would I move or scale fragments non-destructively having them "smart" and connected to original hi res sources . This really necessary when you try to do repeating texture and not just for specific unique UV
Also that height mask + levels doesn't work as proper Z combine if the second layer is not just flat or I am still missing something.
I am dreaming about a kind of vector based soft with nice bitmap support + 16 bit for depth and something like Z combine node in Blender.
Would Blender composing tools have a convenient transform gizmo + a few features from vector soft I wouldn't look any further