David in Action (figures)

[hal]

Hello Davidians,

a new scanned model is born. Hell Spawn is 18 cm tall, from McFarlane action figure collection ( http://www.spawn.com/toys/product.aspx?product=1680 ) and is under construction (the cloak is still missing): here the first images about it.

Every pieces (head, body, arms, legs and gun) were scanned with David Laser Scanner and after rebuild the original pose with 3D editing software.

Scanning (every pieces is fused from 8/10 scans), cleaning, fusion and editing = around 6 hours of works.

In the second image you can see how David Laser Scanner do a good job with details detection.

Soon as possible the compleate figure with mantle (in the back we can see the hole for it) and the movie and the acrobat document for the wiki.

Hoping that this other contribute can show to all how David is the best solution for its price, greetings,

Mattia

[Sven]

Thanks Mattia!

[SAVstudio]

jeeeeeeez, Mattia....

once again awesome work!

Steve

[jharrison]

Mattia, you truly have some of the nicest scans I have seen on these forums. Truly great work!

What do you plan on doing with these models from here? Did you capture texture?

[hal]

Hello,

I've no a plan for these scans, they are only a test ... and an example for showing how much details David can capture

Obviously, scanning each part separately, the work is long, but for me is a pleasure.

No texture for this model. I have a B/W ccd camera, then the texture must be capture in a second time, with a photocamera.

My works are only little pieces of the "table of experiments" of this software. I work with passion to contribute to build a "background visual reference" for the newbie of David

Mattia

[hal]

Hi,

the Hell Spawn scanning work is complete

In the first image we can see the depth image without any kind of smoothing, only one interpolation.

In the second image now is visible the mantle: the main problem with this object is the dimension (around 19 cm tall) and its changeability (it's made with soft plastic). I can scan it only using the opportunity to move the background panels.

The problem of dark materials (from dark red to black) was avoid with a thin layer of baby powder, easy to wash.

The third image is for your fun

Mattia

Edit: in the Wiki Gallery you can see the 360° turntable movie

[MagWeb]

Hi Mattia,


great work as usual ! Bravissimo!

I´m very impressed, foremost the cape is excellent for that´s realy difficult. To scan and to align and to fuse.

To align thin objects brings the difficulty that David´s Shapefusion automatics , searching for similiar topology, suck the inside of the cape towards its outside (or the other way round). Did you align manual or use a reference object?

As one can see from the different Zbrush materials (or marked status?) used for the cape and the body, you haven`t fused all together. Did you use different fusing resolutions?

How many seperate parts owns the sculpture? (Socket, 2shoes, 2hands, gun, body and cape = 8?)

Again, great work
thanks

Gunter

[hal]

Hi,

@ Sven, Steve, JHarrison and Gunter: many thanks. For me is very important your opinion and support.

@ Gunter: the model count 9 separate parts. Each part is fused from 8/10 scans. The final pose was set up using the free moving and rotating tool into David. I've fused the body, the shoes, the hands, gun and head at Poisson 2000, the mantle (two parts) at Poisson 1000 (for 2000 value, time elasped was 5/6 minutes and sometimes, at the end of fusion, I can't see the result on the 3D view, probably a memory or video card problem; I must save the "blind" model and only after a David restart, alla was show again).
About alignment of the mantle, it was a problematic thing: the inside surface was aligned manually, using some scanned parts of outside surface as reference. But the problems are still there: you can see many holes in the mantle surface

Thanks to all,

Mattia

[MagWeb]

Hi again,

I see, but anyhow, the style of the statue allows those holes.

What's about aligning
1. outside meshes and combine them
2. hiding outside
3. aligning inside meshes and combine them
4. aligning inside to outside manually.

Aligning the wings of my Dragon meshes, I noticed that somehow, if you first fuse simple and then poisson you might get less holes. Don´t know how but it worked (BTW@Sven: Maybe that`s caused by the "memory effect" of a previous fusion too?)

Ciao
Gunter

[Bongobat]

Just wanted to say WOW as usual your scans are incredible. Good work Hal.

[hal]

Hello,

@ Bongobat: thanks.

@ MagWeb: thanks for the fusion tips, but now I'm already on a new scan-project. If you want to try or find others way or simply test others parameters, I've uploaded the matle scans (8 separated scans) in this winrar archive: http://www.filefactory.com/file/940fe3/ ... Mantle_rar
(maybe in near future I can change this online service).

As you wrote, the objects like a mantle or a wing (I've the same problem with this dragon model http://www.david-laserscanner.com/forum ... c&start=15 ) are problematic and if you need a perfect scan, a reference object is required.
But these works are for fun and for me the manual alignment of the two inside surface is ok. This is the way I've ran:
- auto aligning the outside meshes of the mantle;
- using the little pieces scanned inside, visible on the outside pieces, as visual reference;
- rotating and moving, by hand, the first inside mesh;
- auto aligment (raw and fine) of the second inside portion of the mantle.

But you can have a better idea with the attached file. I know, and excuse me if they are bad scans, that more and better scans can resolve this problem, but for my initial purpose they are enough.

Have a good time with David,

Mattia

[MagWeb]

Hi Mattia,

thanks for the files.
I´m gonna look at them when there ´s a little spare time.

Gunter

[hal]

Hello,

here an other model 6 inch tall. It's a work in progress (the head and the weapon are still missing). In the third image please look at the polygon density of the armor on the shoulder: there are many triangles (enough to determinate a well detailed mesh) but pay attention: the object was half-decimated! you are watching the 50% of original mesh density!

I think that David can calculate (during scanning and fusion) an impressive number of dots.

Mattia

[Sven]

Again great work, Mattia!
Thank you for showing us again what can be done with DAVID and a good setup. A few years ago I would not have thought that such resolution and details would be possible...
The images show a textured model, right? Could you please show an untextured version of the 3rd image?
Thanks

[hal]

Hello and thanks,

I know that my scans can be repetitive (same models tipology) but I want to finish an other 2 action figures: the dragon, started time ago, and a new Spawn model. Please be patient with me: when I start a project, I can't stop myself

My goal , now, is make a test with similar object tipology but with many differences: color, shape, material, size...

As you know, my hardware setup isn't cheap, but not so expensive (less of 1000 €). And at this price don't exist an other 3D scanner that can do the same work. I've seen (and some friends of mine confirm my impression) many models scanned with the nexengine scanner (1995 €) and the quality is definitely worse.

About your last question, Sven: the model is untextured (as the previous action figures scanned). You can see only the geometry. But if you are talking about the green material, I can change it (maybe you want a David 3D view grab image?) but the result is the same... I've applied only a Zbrush shader, no texture. You can check the real, and coloured model, here http://www.spawn.com/toys/product.aspx?product=2154 and do a comparison between real and scanned.

Soon as possible, I want to do a comparison, taking a photo of real object and making a rendering of 3d model in the same position.

But maybe I've done a mistake and I haven't understand correctly your last question... sorry.

Mattia