END OF AN ERA, FRACTALFORUMS.COM IS CONTINUED ON FRACTALFORUMS.ORG

It is some kind of plastic composit, not VERY fragile, but I wouldn't like to drop it on the floor...

I am very impressed by your voxel export bib especially the inside box! 

That's right I'm wasting a lot of triangles because of this. On the other hand, there are 2 big benefits:
1 - you can see the inside (assuming you cut the bulb in half)
2 - the amount of matter is significantly reduced, thus the price (or I can make it much bigger for the same price as a small "full" version)


This is what I did, except that I did it directly in M3D in InvMax DE comb mode. Try it for yourself, you see that it is not very satisfactory. On the other hand, if I try to use the Outside+Inside mode (or an equivalent method using InvMax DE comb) then the walls are too thin (and even full of holes after it's been processed by Fiji). So I try to increase DE stop to make them thicker, at the expense of details...

My pleasure. I am convinced that 3D printing is a technology that will explode in the coming years. I bet that by 2025-2030, it will be possible and affordable to have a 3D printer at home, like you have a regular printer. There are already kits to build your own below $3000. Imagine that instead of going to the shop and even ordering online, you will order anything directly on your 3D printer and print it directly at home.

For sure!

Ok, did some tests my self, but to start with the bad news: It seems, that the triangulation of the voxelstack itself avoids an efficient polygon reduction, at least within my first fiji models. I see no way so far, to solve that problem in fiji. Resampling or any other procedure I tried, didn't show a significant improvement. I prepared some snapshots to illustrate the problem.
first the rendering of my chosen cutout left the initial fiji output @4 million polygons - right an optimized one @1 million. as there are big cutting planes, which are easy to optimize, the amount of reduction is not significant for other models


next a closeup of a "boxish" area that shows, the layered structure. here is the probem - a plain or curved surface is no longer plain or curved. I think every optimization algorithm will fail here. left the raw fiji output - right the optimized version


last but not least an example, that it is possible to create optimizable meshes out of 3d Fractals. It shows the mesh of a quaternion created with a 10 year old (!) version of fractal zplot. As long as 3d printers reqire 3d meshes, this is the smartest implementation. reduction rates of 95% and above are possible here. left initial mesh - right optimized mesh

Très cools ces fractales mais un peu cher !
Un julia 4D Quaternionesque serait un modèle intéressant (http://www.subblue.com/blog/2009/9/20/quaternion_julia) .

I just received the 3D printed Mandelbulb from Shapeways and I'm quite positively surprised. The material is a white plastic which is rigid and slightly rough but shines a little in the light. Even if the level of detail on the model seemed quite low, as the object itself is quite small, there is no impression of lack of detail.

Here is a photo and the model. Sorry for the low quality, I will take a photo with a better camera.
http://www.shapeways.com/model/794123/mandelbulb.html

I will probably also order the more detailed version I made afterwards http://www.shapeways.com/model/805955/mandelbulb-8.html
And also all the others

Thanks slon_ru for the links. Have you tried to print your models? Any feedback? So far I am happy with Shapeways, but I might look at these too if they have better prices or different constraints, as the 1 million triangle limit is a real issue.

Some more sites to explore in this 3D printing sites comparison matrix: http://3dprinterhub.com/3d-printer-services