Re: [rp-ml] dxf Representation

From: Jonathan Chertok (Universal Joint) <>
Date: Mon Apr 27 2009 - 16:17:05 EEST

Hi Markus,All.

I think the question needs clarifying.

Does anyone know whether dxf describes 4 points in space or two sets of
tris for each quad? I've opened the file and I'm not sure whether there
is a source to sort through this.

The lowdown is I have software that is just displaying the quad face
(which could be curved or flat) and does not display this as two tris.
My understanding is that this additional edge can be either "implied" or
"explicit" (in part whether you can see it) and I'm wondering if this
has to do solely with how the software interprets the dxf data, with how
the data was generated originally or both.

The other issue is one of rendering quad faces and I've come to
understand that this is sometimes controlled by the graphics card which
is an issue I'd like to understand better.


development + design + construction

[ blog ] on digital tools, rapid prototyping technologies and the fabrication of free-form structure.

Markus Hitter wrote:
> Hello Jonathan
> Am 26.04.2009 um 17:15 schrieb Jonathan Chertok (Universal Joint):
>> Can anyone help me understand the nature of dxf representation - say
>> vis-a-vis stl? Are triangles definitively described in dxf format?
> I'm pretty sure DXF allows more graphics elements than just triangles.
> STL is triangle-only.
>> Does this depend on the software generating the dxf? On the quality
>> of the code?
> Yes, the quality of the representation of your geometry depends on the
> quality of the software. Neither DXF nor STL make any rules about how
> close a set of triangles fits specific needs.
>> Does all software "pass off" the rendering for this quad to the
>> graphics card and if so can there be a difference between what is
>> described in the geometry and what is shown on my screen?
> Now, triangles sent to a graphics card represent the actual geometry
> often at a very much lower quality than what you'd typically need for
> manufacturing. A graphics card, for example, doesn't care wether the
> edges of two neighbouring triangles fit together or not. It's also
> harmless to send thousands of unconnected primitives to a graphics
> device - an RP machine would likely choke on such a quality and ask
> for a single, closed solid.
> HTH,
> Markus
> - - - - - - - - - - - - - - - - - - -
> Dipl. Ing. Markus Hitter

Received on Mon Apr 27 16:16:53 2009

This archive was generated by hypermail 2.1.8 : Thu Jan 07 2010 - 08:26:36 EET