awesome, it would be worth disabling scene redraw though. it would speed things up massively. Also making sure the modpanel isnt open might speed things up aswell.
Very cool - I have been using the one by Syncviews, it works differently but i think they both have their uses! (Sync's works thanks to a propagated selection idea, meaning that it also helps for meshes that have been edited after subdivision)
Very cool - I have been using the one by Syncviews, it works differently but i think they both have their uses! (Sync's works thanks to a propagated selection idea, meaning that it also helps for meshes that have been edited after subdivision)
this is really a nice script, but lets say you make a box, and then apply 3 levels of turbosmooth. (or how many you want) and then collapse to mesh.
if you apply the script over that mesh, it will result a much smaller box than the original. (check image below) thats because apparently, it wont average the size, it will only create a new unsubdivided mesh fit inside the current dubdivided mesh.
is it possible to make it average the size? perhaps on future releases?
thanks in advance.
this is really a nice script, but lets say you make a box, and then apply 3 levels of turbosmooth. (or how many you want) and then collapse to mesh.
if you apply the script over that mesh, it will result a much smaller box than the original. thats because apparently, it wont average the size, it will only create a new unsubdivided mesh fit inside the current dubdivided mesh. (see image below)
is it possible to make it average the size? perhaps on future releases?
thanks in advance.
fael097 it is definitely possible to make it like the original mesh, but I will have to research it a little, reverse engineer the Catmull-Clark subdivision algorithm.
Right now I'm working in c++ with the SDK to make TurboReduce modifier. By the way I changed the name to turboReduce.
this may be kind of a silly question, but how would this script deal with a mesh that had any triangles? i know zbrush throws a fit when trying to recreate subds if there is a single tri. is this the same?
this may be kind of a silly question, but how would this script deal with a mesh that had any triangles? i know zbrush throws a fit when trying to recreate subds if there is a single tri. is this the same?
If you mean triangle in the base mesh then it does ok (you can take a geosphere subdivide it and then get it back to its original),If you mean tris in the high-res then its not possible
so how precise is this? Would an intricate sub-d object that was accidentally collapsed be rebuilt correctly into the original cage? What happens with co-planar supporting edges?
so how precise is this? Would an intricate sub-d object that was accidentally collapsed be rebuilt correctly into the original cage? What happens with co-planar supporting edges?
Yes and yes. It create a 100% exact genuine replica of the previous subdivision level. Watch sneak peak #2 to see how precise it is, even with coplanar supporting edges.
[ame]http://www.youtube.com/watch?v=RmpE1Q9p_CI[/ame]
awesome. Is this sensitive to things like the order the verts are stored in? For example- could I run this on geometry imported from another package? Also what would happen if you moved a vert and then tried to run the script? Would it give you a new cage with the altered vertex averaged into the sub-d cage or would it just break?
awesome. Is this sensitive to things like the order the verts are stored in? For example- could I run this on geometry imported from another package? Also what would happen if you moved a vert and then tried to run the script? Would it give you a new cage with the altered vertex averaged into the sub-d cage or would it just break?
I'm not aware of it being dependent on vertex order, you can delete parts of the mesh and it still works. Yes you can run it on geometry imported from another package, but you will get accurate results only if the mesh was subdivided with the Catmull-Clark subdivision method (maybe later I will add support for other methods).
If you move vertexes it will give you a cage with the vertexes averaged into the the sub-d cage. But the cage will be inaccurate meaning if you subdivide it it will give a different mesh as the one with moved vertexes.
Replies
Version 1.1 is out with the feature to recreate all subdivision levels at once.Here is a demo http://www.youtube.com/user/obliviboy#p/a/u/0/asfko3kQQqQ
if you apply the script over that mesh, it will result a much smaller box than the original. (check image below) thats because apparently, it wont average the size, it will only create a new unsubdivided mesh fit inside the current dubdivided mesh.
is it possible to make it average the size? perhaps on future releases?
thanks in advance.
if you apply the script over that mesh, it will result a much smaller box than the original. thats because apparently, it wont average the size, it will only create a new unsubdivided mesh fit inside the current dubdivided mesh. (see image below)
is it possible to make it average the size? perhaps on future releases?
thanks in advance.
fael097 it is definitely possible to make it like the original mesh, but I will have to research it a little, reverse engineer the Catmull-Clark subdivision algorithm.
Right now I'm working in c++ with the SDK to make TurboReduce modifier. By the way I changed the name to turboReduce.
I released version 1.2 which is 100 times faster than 1.1 and more bug free. You can get it here http://www.scriptspot.com/3ds-max/scripts/turboreduce
Thanks everybody.
2 times faster than 1.1 and macro available
Here is the changelog for the last two versions:
1.32
->Added an icon
->Added ability to TurboReduce a simple plane
1.31
->Added the SmartClone function
[ame]
edit: Still in development.
so how precise is this? Would an intricate sub-d object that was accidentally collapsed be rebuilt correctly into the original cage? What happens with co-planar supporting edges?
Yes and yes. It create a 100% exact genuine replica of the previous subdivision level. Watch sneak peak #2 to see how precise it is, even with coplanar supporting edges.
[ame]
I'm not aware of it being dependent on vertex order, you can delete parts of the mesh and it still works. Yes you can run it on geometry imported from another package, but you will get accurate results only if the mesh was subdivided with the Catmull-Clark subdivision method (maybe later I will add support for other methods).
If you move vertexes it will give you a cage with the vertexes averaged into the the sub-d cage. But the cage will be inaccurate meaning if you subdivide it it will give a different mesh as the one with moved vertexes.