Building modeled, Now time to UV Map and Unwrap...Please help!

[HawkPoint3D]

 

Here are a few pics of the modeling I just completed in MAX for preparation into UNREAL.

I honestly don't know where to start in UVing this building. Should I break this into sections i.e each of the angular wings of the building and the center section then?

Advice would be greatly appreciated and I don't mind sharing the model either!

Dave

[Lewis Garrison]

How will you be using it in Unreal? If its from the scale of a person then yes I'd break it up into much smaller sections than you have shown to keep the size of your light maps down. If you're still learning things out I'd utilize the Datasmith workflow to import things. This will assist you in quickly going back and forth and figuring out what works best from Max to unreal.

[jesusgomez2]

Hi,

 

That's simple geometry, the unwrap UVW of 3dsmax will do a decent job automatically (by parts, roof, floors, walls...)

 

As Lewis said, you can also use Datasmith for the complete model, and fix by hand the uv map errors after the light backing in unreal if needed.

[HawkPoint3D]

I know it's simple geometry. I modeled it that way and took meticulous care to only model polys that would be visible. I am not allowing the user to go inside as there is really not much to see in the building to make it worth that effort. I planned the units right off the start in Max to go straight into UE without issues using centimeters. I have already tested just the geometry importing into UE via FBX and it comes in perfectly. my biggest grip is the UV mapping which is always the case. I still have not got my head wrapped around the whole lightmap issues and how to decide how to break up the model if at all. I have seen entire models completely unwrapped onto one channel and the models look astounding. I dunno how they are doing that. This is my struggle.

[HawkPoint3D]

Hmmm...I believe I just found a Gem of a MAX Script! https://www.youtube.com/watch?v=qx8GGcthkxM&t=15s

 

Mapping seems to be going ok:

Edited June 27, 2019 by HawkPoint3D

[Francisco Penaloza]

Sadly the answer to this is not simple, and this is because the way game engines work. We as Archviz artist forgot much of optimization a long time ago, our machines and software can handle most of our stuff even if it is messy. Game engines not so much, now technology is advancing and things are a little easier now than just a couple years ago.

 

Data Smith can do most of the heavy load for you actually. It will just crunch over your scene and you get it ready to bake in no time.

If you want selective adjustment, then unwrapping plugin will help you or the good ole 3D Max unwraper can work just as fine.

 

Having said that, you need to figure out how big your scene will be at the end, what's your final frame rate output, it is just for images?? it is a real-time demo or video animations? this will dictate your average frame rate limits, and then you can decide were to adjust your scene.

If that building will be seen from outside only, but a human scale then you could try to collapse all the windows in a single element, group columns and simplify roof. How big each chunk should be?? well, that depends of how large your scene is, if you are planing to do detailed landscaping and cars, and butterflies, then you need to optimize more.

 

For example, if all the walls are a single object with a4K texture, this is a one draw call, but then if you are only showcasing one side of the building, still will load the whole thing. If the building is divided for each single elements, wall, glass pannels mullions each picket fence, then that is many draw calls per frame, but if you are only creating stills, then this is not a problem.

 

Because the time for me is very important, I usually just export with Data Smith very close the way I receive the model, then I check how the performance goes and then I refine from there.

 

Cheers.