To LUGNET HomepageTo LUGNET News HomepageTo LUGNET Guide Homepage
 Help on Searching
 
Post new message to lugnet.cad.dat.partsOpen lugnet.cad.dat.parts in your NNTP NewsreaderTo LUGNET News Traffic PageSign In (Members)
 CAD / LDraw Files / Parts / 6574
6573  |  6575
Subject: 
Re: Advice on hair pieces
Newsgroups: 
lugnet.cad.dat.parts
Date: 
Wed, 22 Sep 2010 13:44:20 GMT
Viewed: 
22824 times
  
In lugnet.cad.dat.parts, Philippe Hurbain wrote:
Also I suppose it should look out for lines that are totally out of place, >> which it looks like Edger2 does.
Not sure what you mean - it doesn't check if a line has its ends properly ending
on a surface vertex.
(loosely related) I have a pending update for edger2 that check if the angle of
existing condlines matches angle between surfaces. If the condline angle is
flatter Edger2 creates a new condline that matches the surfaces angle. This is
intended to properly create condlines at the edge of primitives (eg. sphere
matching a shrinking cone, such as in this part:
http://www.ldraw.org/cgi-bin/ptdetail.cgi?f=parts/u9029.dat

That's one of my goals with my Edger upgrade as well, which it should do if it
reads the primitives correctly.

What I meant about bad lines was that I noticed your tool does some checking for
spurious lines; that must just be for lines between coplanar surfaces. (I could
see this being a problem for highly detailed parts that are large enough to make
a long curve with a much smaller angle step between though.) I was thinking
specifically of cases where a line doesn't touch a surface vertex at all, like
in LUT's version of my beloved part 6251--there's a bogus line extending to the
bottom of the piece, right at the center of the hole.

The original edger always chooses its control points for conditional lines by
looking at the vertices of the surfaces it's joining. So an edge line that
doesn't do this, that uses control points not touching a surface vertex, would
be considered spurious by the new tool as well and (optionally) discarded.

I use two tools for that. First is Sculptris (http://www.sculptris.com/,
download here:
http://download.pixologic01.com/download.php?f=Sculptris/sculptris.zip), that
allows to sculpt a mesh. The main use is to sharpen the edges where needed
(plastic has always some rounding, and scan adds to that). With a sharp angle,
mesh reduction will create a triangle edge there, and you then may add a fixed
edge line. The other tool is Meshlab (http://meshlab.sourceforge.net/), used to
reduce the size of triangle mesh. After scanning and Sculptris tuning of a
Fabuland head I end up with a +/-200k triangle mesh. Using Meshlab decimation
filter, I reduce this to 3-4k triangles before conversion to LDraw. Sometimes I
do that in two steps, eg. for Dobby's head
(http://www.ldraw.org/cgi-bin/ptdetail.cgi?f=parts/43745.dat). A direct
reduction to 3-4k triangles lost too much details, especially in the eyes area.
So I first reduced to about 12k triangles, selected the areas I wished to
preserve, then reduced again the rest of the mesh.

I'll have to try those out then. Generally I'm really terrible at 3D modeling
and I find most modeling tools (like Blender) too daunting to work with, which
is why I made 6251 and 3898 by eye. (I was doing painstaking work on
reconstructing 6093a for the longest time until LUT finally released their own
and Anathema cleaned it up. That was a huge relief!) But I'm much more
comfortable with utilities for working with an existing file.



Message has 1 Reply:
  Re: Advice on hair pieces
 
(...) So do I... That's why I love Sculptris that has a wonderfully intuitive interface (but it does help to have a reasonably powerful machine). Meshlab is more austere and lacks documentation, but the few needed functions can be conquered without (...) (14 years ago, 22-Sep-10, to lugnet.cad.dat.parts)

Message is in Reply To:
  Re: Advice on hair pieces
 
(...) Not sure what you mean - it doesn't check if a line has its ends properly ending on a surface vertex. (loosely related) I have a pending update for edger2 that check if the angle of existing condlines matches angle between surfaces. If the (...) (14 years ago, 22-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
    

Custom Search

©2005 LUGNET. All rights reserved. - hosted by steinbruch.info GbR