Cheers my Unreal Friend,
I’ll try to answer in order:
1.a._You can snap the objects to a grid, or simply press the “End” key on your keyboard to snap them to the first thing the object hits (could be the floor, or something else). I think you’re aiming for the “End” key but let me know if you’d like to know how to change the snap to grid settings. I suppose you could have your pivot to the bottom, but I do not think that is the typical use of the pivot. I think the typical use of the pivot is to change how the object pivots, or perhaps a symmetry copy and paste thing. I honestly have not used the pivot point much, but know you may need to use it to re-center meshes that you import and have funky pivot points.
1.b._Hm, I do not understand your problem exactly since I tend not to put objects within other objects. Your use case may differ so I’m not suggesting you shouldn’t. Perhaps you meant you put components within components or attached them? I’m not sure if this is a lighting related question or an object placement and collision question. Maybe you can give a use case or explain the question in a different way.
2_Instancing an object reduces the processing required because it processes it as one object (to my knowledge). So if you change one of those instanced objects it will change all the objects. I think of them as merely references and do not have ‘personality’ of their own. It is NOT the same as duplicating an object which would make each object have its own ‘personality’ and EACH object need to be updated or can be uniquely different.
For example if I changed the color of a green cube to be a purple cube and it was an instance, then all my green cubes that were a part of that instance become purple. Whereas if they simply were duplicated I only changed one green cube to purple.
3_Your pillar with separate objects? You might want to explore using one Actor/Object and adding components to it. This sounds like another efficiency question to me. I think that the result would be the same under general circumstance, however with one object I do not think the game engine would do as much processing. Whether that be because of lighting, collision, physics, or level-of-detail (LOD)
To put my logic out there, this would be because even though you matched the polygon triangles, the amount of objects would cause the engine to do that much more work (unless you put effort in to disable collisions, disable shadows…but then I wonder what may have been the purposes to begin with). Each use case is different and honestly if it is game-breaking to make it into one object for some reason then you shouldn’t do that - but that’s my opinion.
4_This HDR light in the sky is something I cannot quite talk to, but using the sky sphere you can adjust the clouds a little, the light direction, the time of day, as well as some other do-dads. I honestly think a sky sphere gives your project a nice personality and atmosphere - but I’m not familiar with everything it encompasses. You’ll want to adjust the Sky Light in your level, and probably play with point lights that have large distances, if not unbound.
I’ll try to answer any questions you might have, but don’t count on me. I actually asked a question before finding this question myself - paying it forward.
Happy Friday,
SmashRash