Rendering glitch from 3DM file

Next
 From:  Brian (BWTR)
1361.1 
When trying to render in Carrara I can not seem to cure, after making several modifcations, the visual glitches as shown on the screen shot.
Any help/advice appreciated.
I get the same result with an IWO file save version.

EDITED: 22 Nov 2008 by BWTR

  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
1361.2 In reply to 1361.1 
Hi Brian, right now your hat is made up of several independent surfaces that happen to be sitting right next to each other.

You should select these individual pieces and use Edit/Join to glue them together into a connected piece.

When surfaces are connected it will generate proper meshes that have aligned vertices along shared edges.

When surfaces are not actually connected to each other each one will get meshed individually and the mesh vertices may not perfectly align on the generated meshes, particularly when you have some large and small pieces that meet in a "T" type formation like you have here.

One surface piece may need some tuning, it has a small fold in it where it collapses down to a corner, this is the kind of thing that I was talking about the other day:

So after joining one problem will be removed.

But another problem is in your shapes:





When you zoom in to that corner circled above, you can see that it is not a flat surface as it appears at first, there is a small area where it has a bent and twisting piece in it.

That piece will need to be rebuilt with some cleaner geometry, let me know if you would like some help with that part.

It looks like some of your input curves were at some slightly different elevations with respect to each other, it would be better if these things were all lined up.

- 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:  Brian (BWTR)
1361.3 In reply to 1361.2 
Thaks Michael.
1. I am really concerned at this "different level" thing you suggest. I can not imagine how I would have done that.
2. Not that I got the exact close up view as you, I did find things to adjust. Not happy at the whole exercise that these are not highlighted--as per previous discussion, though.
3. I previously had all those areas quite satisfactorily joined, maybe back in the save before that file I sent.
However, here again, I seem to have very satisfactory joins made of all the offending sections, but, now, see how the file opens in Carrara?

Very odd and frustrating. Sorry to have to ask all these questions.
Brian

EDITED: 22 Nov 2008 by BWTR

  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
1361.4 In reply to 1361.3 
Hi Brian,

> I am really concerned at this "different level" thing you suggest.
> I can not imagine how I would have done that.

It's pretty easy to do if you have dragged pieces around and just kind of drop them in a spot that looks like it really close, instead of making sure you have hit a snap point to know that they are precisely locked together.

Unfortunately it is tough for me to determine exactly what thing you're doing that is not accurate without me watching you through a whole modeling session.

The best advice I can offer is to make sure you hit snap points when positioning and drawing different curves.

When you mirror a curve, don't just draw a mirror line that looks good, make sure you have straight snap turned on and that it has engaged so your mirror line is precisely vertical or horizontal.

When you draw a curve that has open ends which should be vertically aligned, make sure you have snapped both of them on to some vertical guide instead of just placing them freehand. Stuff like that...


> However, here again, I seem to have very satisfactory joins made
> of all the offending sections, but, now, see how the file opens in Carrara?

Is this opening it as a .3dm file in Carrara?

What happens if you export from MoI as an .obj or .lwo format and load that into Carrara instead?

- 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:  Brian (BWTR)
1361.5 In reply to 1361.4 
Michael,
1. I have been, because of previous problems, been extremely careful that I only click on ends etc. I am beginning to believe that MoI does not truly believe what the little notations for clicking say? (I admit to being less than perfect--but!)
2. I have, previously, had an excellent relationship with importing 3Dm, Iges, and Lwo into Carrara--I have reported , to DAZ, that previous problem which you said was a Carraras fault.
3. However, if you note the OBJ import of that file (See the 3DM screen shot), which is identical to the LWO import, and then there seems to be a gross distortion, WEIRD, to the IGES file compared to the 3DM import.

Something aint right in Kallamanzoo!

EDITED: 31 Dec 2008 by BWTR

  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:  Brian (BWTR)
1361.6 In reply to 1361.5 
Michael, I have gone back to an earlier file save where, part at least, of the glitch, occurs.
Now I can not fix this, and even got one of those red squares trying!, so, it seems, maybe all my "fix it" efforts from that file save on previously have not fixed the problem but, maybe, compounded it?

EDITED: 31 Dec 2008 by BWTR

  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
1361.7 In reply to 1361.5 
Hi Brian,

> I am beginning to believe that MoI does not truly
> believe what the little notations for clicking say?

That would be news to me - I definitely put in a lot of effort so that snaps are snapping when they say they are snapping...

Can you please post your updated .3dm model file?

It looks like that is still an unjoined edge problem. If your surfaces have too much of a gap between them right in that area, it is possible for those edges to not get fused together, but still have it all select as one piece because all the other edges did get joined.

Basically like this:



The above can be all joined so that it selects as one piece, but there is still a hole inside, the 2 edges across the hole from each other will not get joined if they are too far apart. Maybe in your case there are different elevations right along that edge?

I don't quite follow what the 2nd problem is on the iges import?

- Michael
Attachments:

  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
1361.8 In reply to 1361.6 
> Now I can not fix this, and even got one of those red squares trying!

It can be hard to fix things up later, it is definitely best to get things snapped together more accurately from the beginning.

Here's the kind of thing I mean - in the BRsSHat11.3dm that you posted earlier, here is a view from the right-hand side:



There is a natural baseline to this object, which I show there with the red line. Now if I zoom in to where that arrow is pointing (the left side edge of your baseline), it doesn't take very much zooming to see this:



The problem here is your baseline is just kind of floating at some arbitrary location, it's not aligned with the grid.

It would have probably been a lot easier to keep the accuracy high if you positioned the model so that this base line was running along the world axis line, like this:



You've got it set up nicely aligned with the other 2 world axis lines, like it is nicely positioned in the top view for the x and y axis lines. But you don't have it lined up nicely in elevation, and that's where you've got the accuracy problems in this case. (several pieces are on different z elevations)


If you do want to have your model at a more arbitrary elevation instead of z=0, then you should probably try to draw a reference line in there early on that spans across the whole width of your model, so it can be used as a common snapping target.

- 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
1361.9 In reply to 1361.6 
I'll see if I can come up with a good suggestion on how to tune this up.

I think this is another situation where instead of making a bunch of little surfaces that kind of work their way around the hole in the middle, it would tend to be better to make one big sheet and then trim it back.

- 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:  Brian (BWTR)
1361.10 In reply to 1361.9 
Thanks as usual Michael, and sorry for putting you to all this work.
Your suggestion that one needs to allign to the grid--for which snappimg to is not (?) required--adds another finness which I had not been aware of.
You have somethng to investigate so I will leave that for the moment.

In respect to that iges file image. The dip in the front should be, as it is, in the front, as with the 3DM file, but the curved up side should be on the right of that---not the left! VERY, VERY wierd!

Don't dispair, what a pleasure to have an app that does not crash!
Brian
  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
1361.11 In reply to 1361.8 
Hi Brian, here is a different method.

In this method, you don't worry so much about the circular outer boundary or inner hole of the hat brim, just try to think of it how it would look as one big rectangular sheet.

So I drew some curves:



Those are then used as the profile shapes for a sweep, the rails are the other 2 ones along the edges, that will create this surface:



Now you have the rounded outline curve that you drew:



Now trim the rectangular sheet using your round outline curve as the cutting object, and throw away the outside part:



Now trim this and the top part against each other, at this point it is good if the top part kind of sticks a little bit down through the bottom sheet, it is easier to make sure that pieces will join up properly if they push through each other and then get trimmed down to the common intersection. That creates this:



Model is attached as BRsSHat_new.zip .


The thing that is nice about this method is that there are fewer final pieces, so there is less worry about alignment between a bunch of small pieces.

Instead of trying to negotiate around a hole bit by bit, it is better to pretend for a minute that the hole doesn't exist, then trim out the hole later on.

Sometimes this is a bit difficult because it is easy to focus so much on the outlines of the final shape, it is easy to be concentrating a lot on using those outlines as direct surface construction profiles instead of as trimming profiles.

- 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:  Brian (BWTR)
1361.12 In reply to 1361.10 
To explain the wierdness maybe?

EDITED: 31 Dec 2008 by BWTR

  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:  Brian (BWTR)
1361.13 In reply to 1361.12 
Michael. Thank you for the zip file.
BUT, who, really, would have that "obtuse?" work flow pattern in mind?
Sorry if I am being niggly, just trying to input a thought process that maybe be applicable to us lesser mortals!
  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
1361.14 In reply to 1361.12 
> To explain the wierdness maybe?

I think I see - I guess Carrara decided to apply a mirror transform to the object for some reason? I don't know why though.

If you do a mirror in Carrara around the x axis, it looks like it would fix it.

- 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
1361.15 In reply to 1361.13 
> BUT, who, really, would have that "obtuse?" work flow pattern in mind?
> Sorry if I am being niggly, just trying to input a thought process that
> maybe be applicable to us lesser mortals!

There are a lot of benefits to working in larger sheets like that though, it is a lot easier to adjust the shape and have a smooth flowing result. If you have a ton of little fragments it can be hard to get them to connect up to one another smoothly, it is easy to have creases where they touch each other.

I guess it is just something that gets more natural the longer you do NURBS modeling.

It's kind of a warning sign trying to negotiate around something that seems like a hole instead of trimming it out as a hole.


Also, imagine that you were going to make that hat in real life - would you start with the little pieces you originally had and try to stitch them together, or would you start with a larger piece and then cut it with scissors?

- 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:  Brian (BWTR)
1361.16 In reply to 1361.15 
No to worry, but this was my starting point on that hat rim.
Most from there, other than I thought logical, was trying to make good glitches.
Brian

EDITED: 31 Dec 2008 by BWTR

  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
1361.17 In reply to 1361.16 
That was a good starting point - actually by using the Planar command you create a large plane that is trimmed by those curves. It is quite similar to the kind of build-large-sheet-then-trim-back method that I showed above.

- 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:  WillBellJr
1361.18 
Wasn't there a script posted a while back that enabled you to highlight "open" edges or something like that?

It made a good tool for checking your surfaces to see if they were closed as you expected - actually I think it was perhaps as simple as turning on points? (Or perhaps selecting certain points?) Don't remember now...

-Will
  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:  Frenchy Pilou (PILOU)
1361.19 In reply to 1361.18 
http://kyticka.webzdarma.cz/3d/moi/
there is Select all curve closed here
---
Pilou
Is beautiful that please without concept!
My Gallery
  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