Jump to content

architectural to realtime best practices


nawash
 Share

Recommended Posts

Hello all

I own a small RT3D company and I now work with architects.

I used to work in the video game industry.

Working on architecture realtime 3D presentation is quite exciting for me but I still wonder how I can improve my production process.

For now I get the models from the architects from archicad, autocad or 3DSMAX.

Those models are filled with useless additionnal polygons that kill performance and correct rendering in Unity. (BTW thank you to the guys who added xview in 3DSMAX).

 

For now I have worked with 3D artists that optimize these models under my supervision.

 

This step is horribly long and kills productivity.

 

Do you guys would have recommandations for a nice and smooth production process that would kind of lower the 3D models optimizations hurdle.

 

Recommended formats (sketchup ?) and/or practices ? :confused:

 

I am currently creating an architecture viz for iPad and as you can imagine it requires a lot of optimization.

 

Any hint about a smooth production process ?

 

Thank you.

Link to comment
Share on other sites

I'm still trying to work that out also. One of my the models that I texture baked had 620,798 poly's in it. That was an overall existing school campus, the new site and the new building with some interior work. All texture baked. I was able to bring it into Design Review, Navisworks and Unity with minimal difficulty regarding the transfer of information. However, since there were so many poly's the performance really dragged and sometimes I'd get kicked out.

 

What I'm trying to work out now is optimizing the model in 3DS Max before I export it. I've seen a massive drop in poly's this way. I've also taken out troublesome high poly models that were used for renderings and animations.

Link to comment
Share on other sites

Hi, some time ago I've developed a maxscript for that purpose, basically it finds faces(polygons) that overlaps and deletes face with a lowest surface area or both of them if they are of the same size. For a building with 150k polys it was up to 50% reduction (72k polys) with no visual difference. Well sometimes it takes hour or two to clean up and not all polys gets deleted etc, but still it's a lot faster then do it by hand.

Link to comment
Share on other sites

Model "Quality" these two will help you identify larger model problems.

http://www.ericchadwick.com/examples/provost/byf2.html - Explains some of the causes of what is considered extra geometry by the render engines.

http://www.ericchadwick.com/examples/provost/byf1.html - some ideas on how to lay out the world organization and hierarchy to let you get away with more.

 

Most of the problems i have found come from the process of designing architecture and keeping up with fluid designs while not compromising too much on the above model ideals.

 

it seems like you need to have a wicked diverse skillset to push the envelope (Scalability-wise) with realtime work right now

Link to comment
Share on other sites

  • 3 weeks later...

I am experiencing similar problems with geometry complexity too. Trying to balancing complexity and rendering speed is often a challenging process. We use a Revit-based workflow so the geometry is often harder to work with (all walls are separate objecst, cannot control the number of vertices on curved objects, etc...)

 

Try using PolygonCruncher. It is a very effective tool to reduce the number of polygons of a complex model without losing much details.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...