View Full Version : SPEED MODELING CHALLENGE: "Nano!" Deadline: 31 Jan 2016

01-20-2016, 10:14 PM
Deadline: 31 Jan 2016

1. Every challenge specifies a subject for you to model along with a time limit. You should spend no longer than the allocated time modeling. As in most challenges like this, your honesty on modeling times is essential. How you interpret the brief is up to you as long as the final model conforms to what is asked for.
2. Post a wireframe and final render of your model (min 800x600 pixels) in this thread. Note: Some challenges will include final render instructions and time limit restrictions as well. This is up to the judge creating the challenge.
3. Only LightWave is to be used for modeling. However, 3rd party Plugins are allowed.
4. You can enter as many times as you like, posting each in this thread. If you run over the time limit, we encourage you to submit the model, but it won't be subject to the judging.
5. Time spent on texturing, setting up a scene for a render and rendering is not included in the allocated period.
6. Although, not required, it is always nice if the artist provides information on how the model was created, what plugins were used and any special techniques used.
7. The following criteria is used to award points. Modeling is looking for good poly flow. The better it is, the more points awarded. Creativity is how you interpreted the specifications of the challenge and applied them in way not necessarily expected. Expression is how you used the model in the final render/scene that supports the creativity aspect of your submission. Render refers to how well the model, creativity and expression were conveyed.

Criteria Points
Modeling 0 to 5
Creativity 0 to 5
Expression 0 to 5
Render 0 to 5

Challenge: Nano

Anything too small to see with the naked eye or related. This can be bacteria, atoms, skin cells, etc. I'll also expand it to the tools used to manipulate or view this tiny world like microscopes, or petri dishes.

TIME LIMIT: You must complete all geometrical/mesh construction in 120 minutes! We're on the honor system here people.

PRIZE: Winner gets to set the terms of the NEXT Speed Modeling Challenge!

JUDGE: I am the judge.

Bonus points for scientific accuracy and animation.

01-24-2016, 03:53 PM
It's a feather! Total time (Modeling, Scene setup, Rendering): 1:53


Here's the reference:



01-24-2016, 04:10 PM
Very nice!

01-25-2016, 09:51 AM
That looks awesome, MonroePoteet. Took me a second to realize your render wasn't a reference image.

01-25-2016, 05:19 PM
Thanks! I think a lot of it is the scene setup to simulate the Scanning Electron Microscope (SEM). I've attached a really basic "Pollen" scene with approximately the setup I used to help other folks "fake it" like I did. The basics are nodes in the Surface editor using the Occlusion node for Diffuse and the Incidence Angle on a Gradient on Luminosity to highlight the edges. There's a Bump node in there to give the Pollen some texture. It takes some tinkering depending on the model, specifically the Gradient on the Incidence angle. Note that it's the Alpha channel on that Gradient.

Forgot to mention: the Surface is 50% Luminosity and 50% diffuse, with the nodal inputs on top of that. The SEM_Table surface has a similar setup with a tiny Turbulence to give it a brushed look.

The shadows on the Light are turned off, and it's 70% ambient with only 30% Light intensity. I also adjust the Light angle so it's oblique to the camera, or mostly. Radiosity turned on.

The lettering is teeny-tiny and parented to the Camera, with 100% luminosity and no Diffuse. I use a 1% size on the Text creation tool, and then position it a few centimeters in front of the Camera.


Enjoy. BTW, this isn't a real submission. Took about 5 minutes to model using a tesselated sphere / Ball, Bevel and Subdivide. The rest is the scene setup, which I think is important in this type of challenge.


01-26-2016, 11:05 AM
Thanks MonroePoteet!

And I'll try to make a model but not sure I have the time ...

01-26-2016, 06:51 PM
Yeah, I know what you mean. Always something calling my time away from LW (I was doing CNC milling today). No problem.


01-29-2016, 08:11 PM
MonroePoteet --- That has a very "Andromeda Strain" (circa 1971) sense about it. Now I'm dating myself...

01-30-2016, 02:11 AM
To support the contest I threw this together and as you probably can tell, I was all out of inspiration and decided to model the first thing I saw. So.. When you print out a paper from (Linux's) nano you might need one of these:


01-30-2016, 07:32 AM
Nice clean model. How did you get the thickness on the springs for the render? Or is their thinness in the wireframe just an artifact of the ColorPolygons?

FYI, my "inspiration" was to do a Google Image Search on "scanning electron microscope", page through all the pictures of the device itself and find the actual photo pages. Here's the Google search:


Here's a few of the sample images I thought about trying:

Dartmouth's General Index (anything with SEM):http://remf.dartmouth.edu/images/

and where I finally found the feather reference:




01-30-2016, 07:26 PM
132128 132129 132130

Carbon 60 and Hydrogen atom from the viewpoint of standard model (quarks,gluons and electron).

Modeling was achieved during beautiful and symmetric time of 51:15, and 20 minutes of aliging,light/texture/rendering plus some small modeling since some flaws were observed in the output.

p.s: the modeling was finished in less than 20 minutes but it took a long time making electron orbit, rail extrude is really a very bad tool and I had the same difficulty as Bohr,Plank Heisenberg and others showing the trajectory of orbiting electron :))))

p.s.s: RIP Larry Schultz for his nice soccer ball modeling, (I did not review the tuts before or while modeling, I did some experiments and errors to make the bucky-ball) and RIP my neighbor who passed away today sadly. he was an excellent guy and I have rent his house.

- - - Updated - - -


01-30-2016, 07:38 PM
Forgot to say these:
Please don't stop the contest for I am making my animation. Also the ball made by mTp is really cool and it is amazing that he did this in such a short time. I think some jitter and non-uniformness could have made it even better. (non-uniformness seems to be a wrong word, help me find proper expression please).

Thanks in advance.

01-30-2016, 11:02 PM
Nice job, Farhad! The bucky-ball is a good representation of C60 and the quark layout of H looks great.

Yes, rail extrude can be a pain. My experience is that it only really works well if the first segment of the rail is aligned mostly down the Z axis, the shape to be extruded is placed EXACTLY in line with the first segment of the rail in the background layer, and never, I repeat NEVER select Aligned to Start.

RE: the "Pollen" ball, from what I've seen online, a lot of pollen is really symmetric and uniform. For example, in this micrograph of actual pollen from Wikipedia:


"Nonuniformity" is a fine term. Here's the Thesaurus.com page for it, giving some synonyms:


Nice job! Sorry about your neighbor.


01-30-2016, 11:18 PM
I forgot to mention: the trick I used on the Pollen was selecting all Polygons on a tesselated sphere / ball (Segments = 3), then use Bevel to inset and shift them inward twice. Finally, hit TAB to convert to subpatches's, and subdivide Metaform. Here's the entire Modeler command sequence:

SHIFT-O (make a ball)
n (bring up the numeric panel)

Select Tesselated and Segments of 3, leave radii at 50cm
RETURN (commit the ball)
CTRL-H (enter Polygon selection mode)
" (select all polys)
b (invoke Bevel)
n (bring up the Numeric panel)

Enter "-2cm" for Shift and "1cm" for Inset
RETURN (commit the Bevel)
b (invoke Bevel again, polygons created by previous Bevel still selected)
n (invoke Numeric panel)

use same numeric values
RETURN (commit the second Bevel)
/ (unselect polys)
TAB (convert to subpatches)
CTRL-D (invoke subdivide)
RETURN (commit Metaform subdivide with defaults)

That's it. If you wanted a more jittered appearance, I do it on the initial Tesselated ball before doing any of the Beveling.


01-31-2016, 03:27 AM
Nice clean model. How did you get the thickness on the springs for the render? Or is their thinness in the wireframe just an artifact of the ColorPolygons?

Thanks! They are just 2-point-poly-chains made with "Helix". The Helix tool leaves a spiral curve wich I converted to 2ppc and then render edges with negative width. So a little cheat ;). For the wireframe I just went positive on the edge width to show the polychains.
Modeling time well under the hour. Most of the time was finding out that I had converted the spiral curves to 2-point-LWCADcurve-chains which wasn't rendering. Surfacing was not too long either after some great tips from you (mtp). Did do my own version though.

01-31-2016, 07:36 AM
An additional note on Rail Extrude: re-reading my post, I have had success (now) with the rail not just down the Z axis. The key (for me) is to make sure the initial geometry is exactly aligned to the first segment of the rail, and facing the opposite direction from the rail so the new geometry isn't inverted. I get into Polygon selection mode, select the polys of the initial geometry and make sure polygon normals are enabled, then use the orthogonal views (Top, Back, Right) to line up the poly normal with the first segment of the rail. And, again, never, NEVER use Align to Start!

Nice trick on the springs. I'm a little surprised the nodal surface setup (e.g. incidence luminosity) affects the edge rendering. I'll have to give it a try.


01-31-2016, 10:24 AM
LoL, My polygon was facing the same direction with the path and by flipping it everything became normal.

Thanks for those tips mTp and for your nice words, you know that there are a lot of things yet to be improved about rail extrude and similar tools but that would be meaningless now to complain about them since everybody knows about those issues.

My result was first very thick and I took advantage of point normal move, if we only had more proceduralism in modeler :(

01-31-2016, 01:48 PM

This description brought me to make the following,
you see as soon as the molecule starts to vibrate it also starts to glow which is kind of signifier of the IR emission. thanks to the excellency of instance and displacement working perfectly together (which compensates some flaws of the modeler).


02-03-2016, 07:54 AM
Maybe time to wrap this up and get a new challenge going?


02-03-2016, 11:43 AM
Sorry, no entry - moving house. Will be sorted probably for the next but one challenge.

02-04-2016, 09:50 AM
Maybe you guys should vote on who the winner is, so the next challenge can start.

02-04-2016, 10:44 AM
Hey everyone, sorry i've been out of town and had limited computer access.

Modeling: 3/5
Creativity: 4/5
Expression: 5/5
Render: 5/5
Bonus: 1
Total: 18 /20

Modeling: 2/5
Creativity: 3/5
Expression: 3/5
Render: 3/5
Bonus: 0
Total: 11/20

Modeling: 2/5
Creativity: 4/5
Expression: 3/5
Render: 5/5
Bonus: 1
Total: 15/20

That makes MonroePoteet the winner. Nice work everybody!

02-04-2016, 11:32 AM
Gosh, thanks! As Squarewulf said, nice work everyone!


02-04-2016, 01:55 PM
Congrats mtp! Well deserved.

02-04-2016, 01:55 PM
Congratulation mTp, and thank you again for those tips, I think we should have a discussion later somewhere else regarding the problems mentioned.

02-04-2016, 02:53 PM
Thanks again! I posted the new challenge here:


I forgot to put the deadline in the title, and can't edit it. Deadling is 17-Feb-2016, or two weeks. Have fun!