DIY and Entry-Level 3D-Scanner Forum

Hi,

I have done a quick tutorial on how to reduce number of polygones on a scanned model using sculptris software
it's a free tiny sculpting software, you can adjust number of polygones with an reduce polygon brush without removing details
you also can add some details with sculpt tools

https://vimeo.com/77226060

cool! Thank you for this tutorial I haven't used sculptris before. What is it capable of? Is it a light version of ZBrush? Can it handle textures?

_________________
Bernhard
www.virtumake.com

it' s a light weight sculpt software, but in alpha version.

I use it to produce quick model, some rough who will be retopologized or put via "GOZ" directly in zbrush

the most valuable thing : it can produce new topology when you sculpt
with the "detail" parameter, you can adjust the number of topology created, it also depend on the size of brush and zoom view

you can paint on model, with some "shader" or matcap, with colors and with sample bitmap, you can use pattern brush to sculpt and paint
you can paint diffuse and bump map

UV's are created automaticaly, with an automatic uv unwrapping... crap unwraping but it's working

if you want i can make more tutorial about using sculptris, and i can show how to retology in 3dsmax and do a render to texture to grab bitmap uv in a more usable way to recompose it in high detail. these technics can be used in blender but i am not too used to it

an image to see what i'am talking about :


i do a retopology , unwarp uv and render to texture very quickly
it can transforme a huge amout of polygones in a clean topology model with clean texture UV

Hi I'm a raw beginner in this field but have an interest in processing scan data for 3d printing, I've a few questions about the benefits or othewise of retopologising the scan mesh. (Hope it's not too dumb a question!)
From what I understand mesh generated from scanning can be messy with high poly counts, if I were to retopologise I would end up with a simplified mesh which would slice well and be well suited to games deployment etc but would have lost essential mesh detail for 3d printing, I would then have to use some form of sculpting tool to work detail into the mesh so it would resemble the original subject.
However I've been reading around and it appears that it is possible extract usable data from a displacement map generated from the uv map of a captured image and project this onto a retopologised mesh, the displacement map actually changing the structure of the underlying mesh, is this correct?
Would this be a viable means of creating better detail meshes from Primesense type depth cameras?

Hi,

Retopology and displace map are not useful for printing i think, due to the fact that 3d printer dont use map to change toplogy and the stl transformation to send mesh to printer reafect topology.

for diplacement map créated using a high mesh on top of a low mesh it's called render to texture procedure. it use the high mesh like a cinema camera projecting detail on uv of low detail mesh.
You can make a low mesh looking more detailed, exactly what they do in game nowadays.

Using high details mesh created by scanner to generate displacement map can result in an ugly displacment map (i think, not proved)
because hight mesh have some "wave" in topology generated by depth camera resolution.

using depth information directly from sensor dont result in great detailed displacement, i have done some quick test and it's... anecdotic



for this test i take image from sensor directly with openNI simpleviewer and put it in gray color in a displacemnt map slot shadder
put this shader on a plane and render.
you can see that details are pretty messy

Hi Singaii, thanks for clearing that up!

@singaii: I am impressed about your knowledge in editing meshes and texturing them. We could learn a lot from you, if you go on with your tutorials. Texture mapping, for example could be very useful for all Kinect and Carmine users, since the RGB Camera is not so good. Could you describe the process of taking pictures with a normal camera and how to map them on a mesh?

Best,
Bernhard

_________________
Bernhard
www.virtumake.com

Yep, I can do some tutorials but only in 3dsmax. Anyway, these technics are easy to adapt from a software to another
the most dificult for me is to speak in english, already in french I find my voice horrible.

I can take a project from the begining to the end with these steps :
1. scan
2. adjustment sculpt
3. retopo (pretty long task so i make it with x4 speed video for some annoying part)
4. uv unvrapping
5. render to texture from a scan with scanner texture
6. adjustment for texture with some high detail photo shooting

i can do it for any model but i think human face is most interesting (but quit a long task)
if you want a quick video i can do it on other model type

Hi singaii，thank you for your introduction about post process of scan.
I also think it is more helpful to virtual reality rather than 3d-printing.
I'm very interested in it and hope to learn a little.
For example, could i adjust the texture of the “pumpkin” scan as following，

pumpkin (click to view in 3D)

_________________
Join the world, Enjoy the life！

Yes, VR is an excellent application of textured objects. I have seen sone videos about the Oculus Rift system and it is just stunning.

I think we would benefit most, if you could show an example of a rather simple object, so we can understand the basic workflow.

Best,

_________________
Bernhard
www.virtumake.com