View Full Version : bad surfaces help
andrewjohn81 05-25-2005, 07:49 PM I know how to create a bad surface, but I don't know how to do the opposite.
If you have two surfaces near eachother that you want joined you can force the join. This is great for sending information to a cnc machine or something like that, but horrible if you need to translate the data to max or something like that.
The way you force the join is by doing a _JoinEdge command and selecting the two edges you want joined. It will come up with an error that they are too far apart and asks you if you want to adjust the tollerance to join them.
After you do that, you will notice that it creates a surface that looks kinda funny, but it's sometimes acceptable. If you did it on purpose, notice how the isolines don't quite touch the edge. Sometimes the _SelBad surfaces still doesn't even select those surfaces.
Rebuild doesn't fix the problem, so I'm curious how to accomplish this.
Is there a way to remove the adjusted tollerance or something like that?
| |
elagman
05-26-2005, 01:32 AM
Im not a big fan of the join edge command. You might have better luck with the merge surface, or match surface tools. If you wouldn't mind posting your model I could give it a shot to see if I can get you what you are after. Also an explanation of how I got there.
andrewjohn81
05-26-2005, 11:45 AM
It's not my model. I wouldn't use the joineEdge commmand either. I was just telling how to do that to see if people knew how to undo the damage to the model. I'm getting these models from other designers, but to export them to max you need a clean model so I have to clean them up. That and get rid of the b-side surfaces that aren't necessary.
I couldn't post the models anyway. These aren't my personal models. They are the models I have to work on at my job.
they other thing you can do if joinedge doesn't work is explode any surfaces that are polysurfaces and rebuildsrf them so that matching edges have the same point count. If you match up the points by snapping them in with OsnapPT on, they'll match up perfectly... that's only if Joinedge doesn't work and as a last resort :D
I don't like matchsrf or mergesrf, as they both tend to change the shape of a surface radically.
Teyon
05-26-2005, 12:23 PM
In my experience, the two don't change the shape of surfaces radically when you use guard knots to protect the surface area outside of the edges to be merged/matched. HOWEVER, the function does re-arrange the U and V directions of the surface after it's done. An odd quirk I don't think they ever corrected.
andrewjohn81
05-26-2005, 01:57 PM
well, none of this still fixes the problem though.
If I try and rebuild the surface, since one of the edges was dramatically altered, the shape doesn't retrim properly. If I get another button today I'll post an image. i think we are done with the rhino side of things though, for this project unless one of the other designers brings us a last minute change to something.
jbe3d
06-03-2005, 08:12 AM
Tolerances, tighten 'em up when modeling! Then loosen them to join. I initially set my absolute tolerance to something like .0001, then if I'm doing Booleans or joining things up, I'll drop a decimal place to .001 if necessary. It's like magic.
JimCarruthers
06-03-2005, 10:45 AM
Hmm, now why isn't using joinEdge good enough for exporting for rendering, aside from the surface normals being messed up. In that case you can just slack off the file tolerance until regular join works--no matter how screwed up the geometry is, as long as the mesh looks okay--or there are some new things in the "Bonus Tools" for unifying mesh normals.
If you want to actually "fix" models made with too loose tolerance, then you've got to MatchSrf things up, or extend and trim.
CGTalk Moderation
06-03-2005, 10:45 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.
vBulletin v3.0.5, Copyright ©2000-2008, Jelsoft Enterprises Ltd.