![Embarassed :oops:](./images/smilies/icon_redface.gif)
I have quickly done a model (a 2001 like space station) in 3 resolutions:
![Image](http://nho.ohn.free.fr/celestia/preview/MultiRes.gif)
Number of poly:
Low: 7180
Med: 16332
High: 37884
File size:
Low: 156Ko 3ds - 396Ko cmod
Med: 364Ko 3ds - 824Ko cmod
High: 788Ko 3ds - 1.6Mo cmod
A quick test in celestia show the following:
The low model must be used from dot to 8km of distance, the Med from 8 to 1km and the High for the rest.
Meanwhile the modeling, I have noticed the following:
- As I was thinking making 3 models represent much more work in 2 aspects:
The modeling:
The use of a decimation tool is unfortunatly not the ultimate and easiest solution. Lot of meshes have to be treaten with special care, in some case you will have to create 3 versions of the mesh from scratch...
The planning of the modeling:
As the modeling represent an over charge of work, to not lose too much time, you will have to plan carrefully when you will "fork" the work between 2 resolutions. You will have to save all sections/paths used to extrude or lathe, to decimate them instead of the resulting mesh to create other resolution. I'm afraid this will be a bit hard for a majority of person...
Should I put in download only the cmod models or also the 3ds?
*************
Meanwhile I was modeling I also tough in a new point to my idea of levels for meshes:
A special gain can be done when models have intern structure; imagine you are positionned near the tail of the shuttle looking the payload; rendering big part of the FlightDeck/MidDeck is useless. With a good use of levels, we can gain a huge number of poly and in the same time have the exterior in full rendering.
In a model as ISS, this permit to model the interiors without fear...