PDA

View Full Version : Modeling a digital sculpture


silversilk
04-12-2004, 11:59 PM
I have a real model made in a small scale of a aculpture that I have to model in 3D Max ( I'm relativily new to the program, coming from the architecture 3D world in Form-Z ). I cut the model into sections ranging from 25 to 25 cms:
http://www.angelfire.com/art2/laender/foto1.jpg http://www.angelfire.com/art2/laender/foto2.jpg http://www.angelfire.com/art2/laender/foto3.jpg http://www.angelfire.com/art2/laender/foto4.jpg

silversilk
04-13-2004, 12:00 AM
them I scanned these sections and mounted them in Max:
http://www.angelfire.com/art2/laender/imagem1.jpg http://www.angelfire.com/art2/laender/imagem2.jpg http://www.angelfire.com/art2/laender/imagem3.jpg

silversilk
04-13-2004, 12:01 AM
http://www.angelfire.com/art2/laender/imagem4.jpg http://www.angelfire.com/art2/laender/imagem5.jpg

So my question is, how could I model this sculpture? Should I use loft tool? Or there's a way I can connect the sections ( that I will draw with NURBS curve using each section pict as reference ), make a mesh surface an them adjust the organic characteristic till I get close to the real model?

I mean, all I have are these images ( besides the real model that is already cut in sections ). How could I model it?


I would really aprecciate any help from you guys. But please remember that I'm a little bit new to this kind of modeling so I ask you to be patient with my ignorance for some techinical language you may use. :blush:

mushroomgod
04-13-2004, 12:30 AM
its a fairly organic object.....Id model it using subds/meshsmooth, I think that would be ideal :thumbsup:

mushroomgod
04-13-2004, 12:34 AM
please remember that I'm a little bit new to this kind of modeling so I ask you to be patient with my ignorance for some techinical language you may use.

there are plenty of tutorials on poly/subd/meshsmooth modeling here.. most of its for modeling characters and the like but should be fine for what you need..

http://www.cgtalk.com/showthread.php?s=&threadid=7228

silversilk
04-13-2004, 04:21 AM
Well, that would go totally in the opposite side from what I was imagining would be the way to do it. As I really don't know the software capabilities I thought that the section approach would work ( after all, that's how we model in real world with its real scale ). For that reason it makes me think: modeling with subds and them meshsmooth will give me the precision I need to show with my final model? I thought about using the sections because that ( in an architect mind ) would be the most precise way to do it. I mean, I think in this subd/meshsmooth proccess is more like keep pulling some control points till one get closer to the final result.

Please don't get me wrong. I really want to learn this in the best way possible, no matter which one this is. What I'm concern is about the precision. I'm still thinking that using the sections in some way will give me the precision I need but please, correct me if you guys think that I'm way paranoied with these sections. :thumbsup:

Thanks for all your patient again.

tresD

silversilk
04-14-2004, 07:12 PM
72 views and only one answer? Please people, if you could help me on that method I would really aprecciate. Maybe some model that has the same characteristics as this one.
Thx again :)

azazel
04-14-2004, 07:40 PM
I think you could do it using sections and subds - create the rough shape based on the sections you got (use splines, then the Surface modifier), then refine it further after converting to editable poly (and rendered as subd).

vef
04-14-2004, 07:43 PM
I agree with MG, this model can be modeled with SDS very easily :)

Backenbotten
04-16-2004, 11:34 AM
If similarity to your physical model is important, letting someone do a 3D-scan would be the most accurate and probably fastest option. Then, if you wanted an easily editable model instead of the mesh mess from the 3D-scan, you could build a splines/surface tools or nurbs model, using the 3D-scan as snapping reference and visual reference. You could also make a polygon/subdiv surf model, and see how that compares to your original and NURBS model. If absolute precision is not necessary, and if you became fast with SSS, that could be an exceptionally fast way of doing future work...

Good luck

CGTalk Moderation
01-18-2006, 01:00 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.