Subject:
|
Re: Advice on hair pieces
|
Newsgroups:
|
lugnet.cad.dat.parts
|
Date:
|
Tue, 21 Sep 2010 04:56:27 GMT
|
Viewed:
|
20031 times
|
| |
|
|
In lugnet.cad.dat.parts, Philippe Hurbain wrote:
> > Heh, I would've thought my idea of a utility was more radical.
> On the opposite! splitting stupidly into triangles is radical! The method you
> suggest is much more sophisticated and interesting ;o)
Well, at least splitting into triangles at the middle would produce a fairly
clean result in that it would eliminate artifacts, but at the cost of creating
extra triangles.
I've been working on my adjustment tool and my initial results have been quite
promising. (Every time I write one of these things I start over in a new
language. Unlike with Edger I'm doing things in C++ this time, so I've had to
reinvent the wheel a bit, creating all the classes for vectors and matrices.) I
actually have something that will do the adjustments now, but I'm trying to
expand it to also include an upgraded version of Edger.
The quad adjustment for this piece showed that out of five quads out of
alignment, it was able to fix four of them within the bounds of round-off error
(that is, it didn't deviate the vertices from outside of what they could have
been before stl2dat rounded them off). The fifth one it was not able to
reconcile under any circumstances, so I made the tool split those cases along
the edge where the angle between the two triangles would be smallest.
My Edger upgrade should hopefully take subparts and primitives into account
properly. The difficulty there has been coming up with a good structure for
loading the parts recursively. The final result though should be very helpful to
me for redoing TLG parts.
> Yep. have a look at discussion here:
> http://www.ldraw.org/cgi-bin/ptdetail.cgi?f=parts/2543.dat
That's a relief. Anathema has made other quality headgear parts so if he's doing
manual adjustments then I feel comfortable doing the same. Just as a heads-up by
the way, your last comment there deleted your certify vote.
On a related subject if this particular part goes well for me I'd like to do
similar cleanup on 30475. When Mdublade submitted it he apparently only turned
the hair into the correct orientation, but no primitive substitution was done on
the interior and while the quads are all coplanar within 3°, they were obviously
never cleaned up further to a tighter 1° spec. I didn't check out whether the
interior is too small for the head, but I'd bet that's the case. Since I have
that part I'm positive I can fix the interior. I assume the usual etiquette
there is to contact the person first where possible. I know that resubmitting
just to improve coplanar tolerance when it's "good enough" is frowned upon, but
it'd be worth cleaning up the other issues.
> Otherwise it would be nice to have a tool to adjust vertices automatically... It
> would be helpful too for 3D scanned parts to join scanned sections to primitive
> substituted ones (see
> http://www.ldraw.org/cgi-bin/ptdetail.cgi?f=parts/30164.dat)
I can't tell you how many times I've wished I had your scanner, but I have no
head for mechanics. Some parts just haven't been released to us yet, like the
still-mysterious x79.
I'm not sure there's a very ideal solution for joining scanned sections to
primitives except for looking for a certain vertex tolerance and making
adjustments where possible. In the case of part 44358, when I joined the
interior to the cylindrical indentation in the middle, I actually added 4
triangles because I needed to go from 12 points to 16. If not for the problem of
12-point circles I think some kind of limited primitive detection and automatic
substitution would be feasible. As it is I think detecting places where a simple
set of primitives could go is quite doable as long as there are hard lines
around the points in question.
|
|
Message has 2 Replies:
 | | Re: Advice on hair pieces
|
| (...) I am short of time at the moment for a more detailed answer, but are you aware of Edger2? ((URL) The code is a mess, but it does a pretty good job... Philo (14 years ago, 21-Sep-10, to lugnet.cad.dat.parts)
| | | Re: Advice on hair pieces
|
| (...) It would be interesting to have a configurable threshold, a 0.1ldu move is not visible and might bring more quads to flat state. (...) Thanks - fortunately it was certified by two other peoples ;o) (...) Probably... (...) There is a special (...) (14 years ago, 21-Sep-10, to lugnet.cad.dat.parts)
|
Message is in Reply To:
| | Re: Advice on hair pieces
|
| (...) On the opposite! splitting stupidly into triangles is radical! The method you suggest is much more sophisticated and interesting ;o) (...) Yep. have a look at discussion here: (URL) it would be nice to have a tool to adjust vertices (...) (14 years ago, 17-Sep-10, to lugnet.cad.dat.parts)
|
12 Messages in This Thread:
  
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
