3D print, I must be dumb...

[RKapuaala]

Is that the Aureus? If so, what kind of material did you use to print this out?

 

[MAX.I.G]

This is same manufacturer but next generation after Aureus. Our Printer Mini Multilens. This is not desktop printer like Aureus, its bigger and more proffesionl, we use all materials that Envision can offer us. Grey, Green, Orange, wax, and after modification even dental material.

 

[RKapuaala]

Wow,,, incredible tollerances Max. Even there cheapest model the EDU has incredible resolution .001" Between 25 and 100 um in the Dynamic Z axis resolution!
BTW,,,,what is the difference between Dynamic and Linear resolution?

 

[brassnautilus]

the contain a list of 3 coordinates that describe the x (side to side),y(top to bottom),z(front to back) location of a points in a given space with a given axis that begins as 0,0,0.
These coordinates are nothing new. They are Cartesian coordinates that were introduced in the 15th century I believe by the mathematician who argued correctly that the location of any object in space can be described by its distance from a fixed object zero and its distance from 0 sideways, up or down, near or far. It pretty much has been used since then for that purpose.
Thats it in a nut shell.
There are various types of printers, but they use a form of the printed layer some way or another.

2.5D programs don't really generate 3D coordinates. Generally, sculpting with programs that utilizes absolute coordinates can be called drafting. 3D drafting perhaps.
There are also many different approaches to the "printing" process.

 

[RKapuaala]

I'm not following you brassnautilus. There are 3 coordinates in a face set and sometimes an additional value/s to describe some proprietary function. These coordinances describe, height, width, depth,,, how is that 2D?

 

[brassnautilus]

I'm not following you brassnautilus. There are 3 coordinates in a face set and sometimes an additional value/s to describe some proprietary function. These coordinances describe, height, width, depth,,, how is that 2D?

Hi.
In printing (the hardware sector), as you had described, there is only X-Z coordinates. Objects are oriented inside the printer (usually) with their biggest displaced areas arranged parallel to X-Z plane, while the Y axis being the direction the layers would stack onto. In other words, dimensions in Y axis would be measured with height of the stacks, not exactly point coordinates like X and Z.
In sculpting (the software sector), many programs do not utilize 3 dimensional coordinates. They use means other than coordinates to describe the this "missing" information. You could obtain a complete set of information on the 3 dimensional shape by having 2D coordinates and all the surface definitions via derivatives.
Some systems convert everything to absolute coordinates in the end, some don't, each approach has its pros and cons.

 

[RKapuaala]

Brassnautilus, I will concede that all printers print in layers thus making one coordinate an approximation of the original coordinate based on the layer thickness. I have written code for parsers to convert various 3D files for vrml, wavefront object and lightwave files and if I recall correctly sterographic files (almost 20 years ago, so mememory on the last is approximate) many years ago and I know those files are truly 3D with no derived information for any plane. Do you have examples of software that creates 3D files using only 2 dimensions. I would be interested to know.

 

[brassnautilus]

pixologic programs don't have absolute coordinates on the Z axis. It's all derivatives, from shading/highlights. This is why you cannot have axis-aligned grids or size things in those programs, only scaling.

 

[RKapuaala]

I think you are getting texture maps, bump maps, displacement maps and reflections confused with 3D objects. These maps are 2D and all the coordinates are derivatives, not just the z coordinate. Shading and highlights as displayed by a viewer are also derivatives, but they are derived from the absolute coordinates in a 3D file. Keep in mind, the printer does not care about shading or highlights. These are unimportant to printing. I use sculptris and it is a pixologic application. You can move individual points on a polygon any where in the 3 dimensions right where you want them and they pretty much stay.
I have also used more robust 3D applications like Lightwave where you can import the sculptris file as a wavefront object file and select a coordinate and change its numerical value in all 3 dimensions. While precision is constricted there are no derivative coordinates. Export that file back out, import it into Sculptris. Don't change a thing, but save it as another file and the coordinates will remain the same. Keep in mind Shading, textures, reflections, those aren't 3D objects.

 

[RKapuaala]

Those are an applications rendering of a 3D object.

 

[RKapuaala]

BTW, scaling is relative to which program you are using. Some programs keep the original coordinates and insert a scaler into the file while others go through each individual coordinate and changes their numerical values. That process takes longer than just inserting a scaler and while it is better in some instances, it causes problems in others. I have used absolute scaling on some files and ended up having parts that are relatively close create issues for the printer because of occlusions inside the object and manifold surfaces. When that happens I have to repair the object one point or several points at a time, or import it into meshlab and pray that their tools can fix the issues (sometimes it can sometimes it can't)

 

[brassnautilus]

Check your graphix memory usage while working in ZB, should be zero, because the program does not generate actual 3D coordinates (or reposition them when you change a view or modify a surface).
It's a different process than most 3D programs. When printing objects sculpted in those programs, loaders usually translate that 2.5D information into 3D coordinates according to a set condition (ie. perspective and angle of view), and the print result would generally vary from the 3D render you saw in the program. This variation would not had been there if the sculpture information were already mapped in absolute 3D coordinates.

 

[RKapuaala]

Rendering has nothing to do with the file. The file is read and the renderer (which uses graphics memory) creates the appropriate image based on the 3D info in a file. Rendering is always 2D. The other dimension is used to add perspective and depth to the 3D object. Here is an excerpt from a wavefront object file created in sculptris:
v -6.72988 149.312 -5.43594
vn -0.000163388 0.0282487 2.15527e-005
v -6.68962 149.313 -5.83021
vn -0.000225368 0.0282482 5.84211e-005
v -6.5243 149.315 -6.16118
vn -0.000412274 0.0282454 0.000208182
v -6.22552 149.324 -6.42673
vn -0.00106955 0.0282245 0.000498917
v -6.93879 149.307 -8.01792
vn 8.76855e-005 0.028161 -0.00222898
v -6.76728 86.9568 1.26585
Here is the same object scaled in accutrans the same 6 coordinates (note the coordinates above prefaced with vn or the normals)
v -0.451883 10.025674 -0.365001
v -0.44918 10.025742 -0.391474
v -0.438079 10.025876 -0.413697
v -0.418018 10.02648 -0.431528
v -0.465911 10.025339 -0.53837
v -0.454394 5.838784 0.084997
So the same faceset has been modified to a smaller scale and the normals have been stripped from the file and place at the end of the file.
vn -1.0 0.00034 -0.000357
vn -1.0 0 -0.000469
vn -1.0 0.000491 -0.000588
vn -1.0 0 0.000894
vn -1.0 -0.000251 -0.000879
vn -0.999999 0.000346 0.000948
Note that the normals (which are used to determine lighting) have only 2 coordinates. The last coordinate is a derivative.
While the vertices (prefaced with a v) have 3 coordinances. That makes it a 3D file not a 2.5D file. The object rendered on the screen is 2.5D because the depth is the trick of perspective on a 2D surface. There is no 3D on a flat screen. On a printer, there is 3D, but because all printers have a minimum layer thickness the stereolithographic application has to derive the placement of the coordinate for that layer based on its approximate distance from the layer.
In a way a printer is 2.5d, even if you print in layer thicknesses in angstromes because the layered dimension is constrictive. In addition when the printer prints, it must fill in the gaps between points in a 3D file thus it is 2.5 D. The file on the other hand is not. At least the standard files are not, and from what you are telling me about ZB (which i don't own) the 3D file is not on that either, only the textures and lighting.

 

[brassnautilus]

Different programs dude. Try sculptris or zb (or talk to people that use those) and you'll find out they do things the other way around.

 

[RKapuaala]

My mistake, I had to read the specs over again the normals also have 3 coordinates, but they are perpendicular to the surface.

 

[RKapuaala]

I'm using sculptris. And you are confused and you are talking about how the file renders, not how it is created.

 

[RKapuaala]

Here is a file excerpt from a stereolithographic file,,, a file used for 3D printing. Now show me how the vertexs in this file are 2.5? The file was created in Sculptris and exported to an object file and then to stl.
solid lips
facet normal -0.205426 -0.477405 0.854333
outer loop
vertex -0.03234 1.62044 0.143692
vertex -0.032431 1.62088 0.143916
vertex -0.036744 1.62117 0.143041
endloop
endfacet
facet normal -0.186374 -0.286198 0.93987
outer loop
vertex -0.036744 1.62117 0.143041
vertex -0.036739 1.62073 0.142908
vertex -0.03234 1.62044 0.143692
endloop
endfacet
facet normal -0.250974 -0.400704 0.881163
outer loop
vertex -0.032431 1.62088 0.143916
vertex -0.032619 1.62184 0.144299
vertex -0.036743 1.62218 0.143279
endloop
endfacet
facet normal -0.20822 -0.224153 0.95205
outer loop
vertex -0.036743 1.62218 0.143279
vertex -0.036744 1.62117 0.143041
vertex -0.032431 1.62088 0.143916
endloop
endfacet

 

[brassnautilus]

I was never referring to when the file was loaded to printer. That's a conversion, a re-calculated result, like when you output into .obj.
The 3D coordinates were generated by your loader, calculated from the 2.5D output of the program. You were printing an .OBJ equivalent of the original sculpt, this equivalent has absolute coordinates.
That's why you should sometimes notice the ratios (especially on faces) were not exactly like the sculpt after printing, because sculpter was using different perspectives than the loader/compiler.
When a figure was sculpted in a 2.5D programs they are generating 2D images that contained Z axis information via derivatives, there's a simulation of pixel positions using pre-defined perspectives, not 3D space. This information was "translated" by your loader/compiler to the 3D coordinates using another perspective as basis for the calculations, and these perspective differences created visible variation between the sculpt and the print.
In other words, you wouldn't get these variations if the sculpting program was using absolute coordinates (autodesk, solidedge...).
Or you could do what I had suggested earlier for further confirmation. In a system with absolute coordinates we could always plot grids that were aligned with X-Y-Z axises. This wouldn't be possible in a 2.5D program. What those program could do, was to lay a soft gridded skin over the surfaces of objects, showing you all the contour information. These surface definitions are "hard data" in 2.5D programs. They are usually generated in a true 3D program.

 

[RKapuaala]

We are going to have to just agree to disagree. I've looked at the raw data from sculptris and from modelers that let you plot the points with a mouse or by changing the numerical values in a text field, I've also built 3D files in a text editor, so I think I could tell if Sculptris was modifying the 'Z' coordinate. BTW, the Z coordinate differs in different apps. In some apps its the Y coordinate reading x, y, z width, height, depth and in other files its x, z, y width, height, depth.

 

[brassnautilus]

http://www.kickstarter.com/projects/117421627/the-peachy-printer-the-first-100-3d-printer-and-sc