Moi(Rhino)
All  1-6  7-16

Previous
Next
 From:  andras
3066.7 In reply to 3066.6 
Thanks BurrMan and where can I set the tolerance values in rhino or moi? I guess the tolerance settings in moi is fix, not available to set
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  BurrMan
3066.8 In reply to 3066.7 
The tolerance setting is in Rhino under Units I think. If Rhino has a different tolerance setting it can report additional numbers if you have your UI set to show 4 or more decimal places. I could be mistaken though with regard to that seting may have been an issue with another package I was using inconjunction with MoI. I dont remember if the numbers came into MoI off or not.
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
3066.9 In reply to 3066.6 
Hi Andras - do you mean you were expecting a 100% completely identical surface with every single surface control point in the same spots?

There are some things where that will happen like with Extrude or Revolve where the surface construction technique is simple and follows a very standardized pattern, but there will also be several cases where that won't happen.

Certain surfacing commands involve a "refitting" step where new curves are created that follow along the other curves but are reconstructions of them.

That fitting process is not a kind of simple process that is always the same between every fitting algorithm unlike Extrude or Revolve for example.

Network in particular involves a refitting process on every curve.

There are also other differences between some of the other parts of the algorithms, like for example Rhino's Network command kind of pretends that sharp corners in the curves do not exist and kind of sags away from the corners, while MoI's pays attention to sharp corners, stuff like that. For an example of this see the attached NetworkSharpCorners.3dm and zoom in to the corner after you create a Network from these curves in Rhino.

> So, no mix?!

Why do you say this? You can still mix together the results. After things have been calculated they are just all surfaces and you can move the surfaces back and forth between the programs.

- Michael

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
3066.10 In reply to 3066.7 
Hi andras,

> I guess the tolerance settings in moi is fix, not available to set

Currently some pieces are fixed at 0.001 units, but other pieces have a relative tolerance factor that is calculated to be some fraction of the bounding size of the objects involved instead of being only a single global fixed value.

I'm generally moving more and more pieces to the relative tolerance method, that's actually how I'm able to fix the different problems that you recently posted where very small segments with extrude or planar were not working.

Having only one fixed tolerance value like how Rhino is set up tends to cause a lot of problems.

One thing that is really bad with having a settable single tolerance is when you receive a file from someone where they have set the tolerance to some value that you are not expecting, like really looser than you normally use. Then when you try to build a new shape after loading that file, it tends to be very surprising that things are way looser than you normally see. It's a problem that Rhino users run into fairly often when using files sent by other people.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
3066.11 In reply to 3066.8 
Hi Burr, I think I remember what your tolerance issue was in the past - if I remember right your CAM package added the tolerance value to the bounding box values that it reported back to you.

If I remember right, it just calculated the actual bounding size, and then added the tolerance value to either side of that when it told you the sizes. That seemed like a somewhat strange thing to do actually.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
3066.12 In reply to 3066.6 
Hi andras in the attached file you can also see a big difference between a Network in MoI and a NetworkSrf in Rhino due to the different algorithms for corner handling.

So yes, in some things there are fairly different ways of processing things.

- Michael

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  BurrMan
3066.13 In reply to 3066.11 
Yes Michael that was it. I have since found in that other package that if I extract the edges of the solids then measure those, they are exact! There were also some other methods to measure that produced the results I was looking for. The "Entity summary" value was showing the bounds data included. When it really came down to it, the MoI models were exact. I was less informed about what i was looking at and you helped me to understand.
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  andras
3066.14 
> So, no mix?!

>Why do you say this? You can still mix together the results. After things have been calculated they are just all surfaces and you can >move the surfaces back and forth between the programs.


I mean for example of using Network command.
It is only an assumption(!)

Well, in that case If there is an existing surface in the project.(Such as 100 pieces of column on the facade) The surfece's seam curves are fix. And the college ask me to modify the existing surface with some Boolean (or any) but I will receive only the initial curves (formerly the college mentioned that the surface was produced by Network command and me as well make the surface by Moi network tool). In this case I have to know the difference of the two software. And this method is not allowed(!) because I must to ask the whole existing surface not only the curves. Because in the further period we can get many unfitted surface. OK, If we work with metres and there is some milimiter differents is not so big problem but we have to strive accurate results. I say again it does not mean problem but we have to consider it.
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
3066.15 In reply to 3066.14 
Hi andras, well just because you get surfaces with not the exact same control points in them does not mean that they are not both accurate to within a proper tolerance value.

Like I mentioned, just differences in the way fitting works can produce different numbers or arrangements of control points even though the distance in the actual results are both within the same tolerance level.

quote:
And the college ask me to modify the existing surface with some Boolean (or any) but I will receive only the initial curves (formerly the college mentioned that the surface was produced by Network command and me as well make the surface by Moi network tool).


This would seem to be strange to me that they would ask you to modify the surface but not actually include the surface! :)


At any rate, if you are going to be working closely on a project with someone else who is using Rhino, you will probably also want to also have Rhino and you will then have the choice of doing any particular operation that you want either in MoI or Rhino depending on what you need.


There is one thing in MoI's network that I do need to tune up though - currently Network in MoI is refit to a somewhat looser tolerance of only 0.01 units rather than the usual 0.001 units, that's to get a lighter surface without too dense of control points. I do plan on improving this in the future so that the outer edges will get fit to a 0.001 tolerance and only the interior ones will be lighter.


- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
 From:  andras
3066.16 In reply to 3066.15 
OK, I see.
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged
 

Reply to All Reply to All

 

 
 
Show messages: All  1-6  7-16