Monthly Archives: April 2013

Gray Anatomy, with Extra Gingold

Here’s day 4 of progress on the Magic: The Gathering Drake token.

I want the drake’s arms to be a little stringy, like a really jacked 70-year old. In order to do that I’ll have to increase the detail around the muscle edges.

I’m going to model these as an in-scale human arm, and then distort them to turn them into drake wings. I’m trying very hard to model the forms and volumes that are actually there, not the forms and volumes I think should be there. Muscles can be a challenge.

These arms are eventually going to have membranes attached to them, so I’ll have to keep that in mind as I develop the geometry.

I Can’t Print This On My Replicator

Fused deposition modeling (or fused filament fabrication, if you want to avoid a nastygram from the legal eagles at Stratasys) is the process by which Makerbot Replicators, Afinia H-Series printers, Cubify Cube 3Ds, Repraps, Rostocks, and too many homebrew printbots to enumerate turn STL files from electrons into objects.

In FDM/FFF printing, the device lays down a layer of melted plastic, moves its build platform downward a tiny bit, and then repeats the process until the model’s completely built.

I’ve had my Replicator 1 for about a year now and I think it’s safe to say that I’ve put it through its paces. I’m pretty good at 3D printing with this particular FDM printer. I understand and work around its limitations, paying close attention to the 45-degree rule. FDM printers generally can’t print too large an overhang from one layer to the next; gravity interferes and the print fails with often hilarious results.

A few months ago I designed a model that I knew had no real hope of ever printing on an FDM printer. I wanted to see what Shapeways was capable of printing for under 20 bucks, so I built a Voronoi Seej bloxen and sent it over to them. A few days later this showed up in the mail:

You’re welcome to give it a shot yourself: just download the model. I’d be interested to see your results.

A cutaway view demonstrates why the Voronoi bloxen fails on an FDM printer. Most of the model’s OK, but it violates the 45-degree rule in a big way right at the top of the block.

The print at the top of the post was my first try at printing this bloxen on my Replicator. Those tiny Voronoi cells on the bottom of the model make it really hard to keep the bloxen on the platform. If any one of those little filament loops should detach from the platform, it’ll eventually get caught on the print head and catch another loop, and another, and another, and once your model is touched by His Noodly Appendage, it’s done for. Ramen.

Those little loops need to stick fast to the HBP So it’s back to FDM Printing 101: using a raft.

After a new slice, the Replicator had no problem keeping the base of the bloxen on the build platform.

A few minutes later it looks like the Voronoi pattern is holding up well. The mortises on the bottom of the bloxen are printing just fine. I’m starting to think this print might just succeed…

…aaaaand the 45-degree rule rules its ugly head and the print starts to fail as predicted right at the top of the bloxen.

As fails go, this one isn’t spectacular. Enough filament fell into the gaps to provide a scaffold for the top layers. The bloxen is still a fairly solid print, certainly usable in a Seej match.

So. Another failed print. But there’s still hope for those of us who haven’t been able to secure $30 million in venture capital funding. I’ve been corresponding with Dean Piper, inventor of the mUVe 1 printer. Dean took up the challenge of printing the Voronoi bloxen, and his resin-based printer knocked it out of the park:

I feel like home 3D printing is in the VHS vs. Beta stage of its history. On one side we’ve got superior market penetration of FDM printers, but resin-based machines like the mUVe are going to be the bots to watch in the next few months.

Erk. Almost forgot: this post is the latest in my continuing one-man crusade to make “Voronoi” the word of the year for 2013.

MTG Drake Token, Day 3

Got a start on the MTG Drake Token’s pectoral girdle today. It’s got me thinking about adding detail to the muscle mass so that when the model’s finally posed (I’m predicting two weeks from now or so) we’ll be able to see the striations in the pecs.

It’s important not to get too caught up in details at this point; I want to keep the geometry simple for as long as possible. I’ll move on to the deltoids and biceps tomorrow.

#markmywords, there will be no nipples on this model. Scales, maybe, but no nipples.

I’ve got a grab bag of interesting posts coming this week, so don’t touch that dial.

MTG Drake Token, Day 2

I started roughing out the drake’s body this morning. This is a 3D sketch; the idea here is not to get caught up in the geometry, just figure out the defining masses of the object.

The Drake’s going to be something of a cross between a pterodactyl and a snake. I’ve already got a decent snake head built for the Seej Ouroboros Guardian, so that’s a good place to start.

For all its complexity, Maya’s hotbox does not yet have an item under Create->Polygon Primitives->Token->Drake, so I’ve got to extrude some edges, pull some points, and flip some faces until I get this rough model.

Ooo. That wing’s going to be tough to print. I’ll be reaching deep into the Well of Cantrips to pull that one off.

It’s a good idea to proceed very slowly at this point in the model, just so you can get a fine appreciation for all the problems you’re going to encounter later.

But my mon Zheng don’t shiv. We’ll get this figured out.

Magic: The Gathering Drake Token, rough sketch

My series on the design of my Magic: The Gathering Beast Token was well-received, so I’m going to do it again with this Drake token. People seem to enjoy watching the growth of a model from idea to 3D-printed object.

If you’re awake at the right time, you can follow the 3d modelling process in quasi-realtime by peering into the Scrying Pool in this blog’s sidebar. I hacked together some automated screenshot/image processing/uploading code that runs whenever I’m working a new 3D model.

I’m usually working around 6AM EST during the week for an hour or so, and longer on the weekends. Now that I’m thinking about it, I might add a webcam script so that test prints get uploaded, too. Note to self: increase stimulant intake.

I keep a few older screencaps in the scrying pool in case you miss the live show, too.

Like all of my programming projects, the Scrying Pool’s a spit-and-baling wire operation, so if it breaks just chalk it up to a passing astral dreadnought.

Back to drakes: The other day, intrepid user @shower_tweets asked if I’d take a model request. Sure, I said. It’ll be fun, I said.

@shower_tweets wants a Drake, and a Drake @shower_tweets shall have. This is going to be a tricky model, because Magic: The Gathering Drakes are basically just flying snakes with wings and hindlimbs.

Getting a single-print, no support pose for a flamboyant model like this is going to be a challenge, and like all good art challenges it begins with a sketch. Here I can start working out the pose, thinking about how gravity’s going to affect the model during the print.

Even though these are literally one-minute gesture drawings, these sketches are also useful to help me think about what features I want to emphasize in the model later.

This sketch would never print on an FDM printer. Way too many overhangs, especially in the wings. Curse you, wings.

But it emphasizes the drake’s keel and the pectoral muscle attachments for the wings: it’s very important to avoid Art Major Biology, as far as I’m concerned.

SO! Stay tuned or RSS’d or Grindr’d or whatever it is the kids are using to follow a blog these days. The drake is started and I’ll be posting updates until it’s done.

The Quickstone Challenge


It’s been some time since I posted a new Seej model. I got tied up with work for couple of months and then I got on to making Magic: The Gathering Tokens, and new Seej work kind of fell by the wayside.

Plus I’ve been working on getting the Forge going, and the Scrying Pool, and I’ve been dusting off my long-unused PHP chops, and also managed to develop cubital tunnel syndrome in my left arm to match the carpal tunnel in my right.

So here’s a quickie as I dip my toes back into the moat. I’ve designed the Quickstone bloxen with paper-thin walls and internal supports to print as rapidly as possible.

Look at the red bloxen in the photo above and you’ll see the backlit mortises self-inverting to become tenons.

My best print time on a Replicator 1 with default firmware and ReplicatorG is 27 minutes.

There have got to be übernerds out there with pimped out RepRaps who can smoke that. I’m particularly interested to see how quickly someone with a b9Creator, Form1, or mUVe can do it. The internal supports on the Quickstone are built for the limitations of an FDM printer, but could probably be deleted if you didn’t have to worry so much about gravity during the printing process.

Can you knock Lao Zheng off the mountain?

First, visit The Forge and download the Quickstone bloxen. (It’s buried under Forge -> Seej Models -> Fortifications.) Then email me with the deets on your slicer, your printer, and your results. Pics or it didn’t happen.

Here are the settings I used:

Replicator1 Dual (using single head)
Infill: 0%
Layer Height: .3mm
Shells: 1
Feedrate: 75 mm/s
Travel Feedrate: 75mm/s
Print Temperature: 240°C
HBP Temp: 110°C
ABS Natural

A few rules:

  • We’re on the honor system here. No fibbing.
  • Cram as many 1:1 bloxen into your build area as you like, and measure your time in bloxen per minute.
  • Slicing time is not included in your total.
  • The bloxen must print in one piece.
  • The bloxen must be usable in a Seej match. If it can’t withstand a firm squeeze, it’s no good.
  • you can modify the bloxen’s geometry, but under the terms of the CC license you must make derivatives available to others. I can host your model in The Forge, or I’ll link to the site of your choice.

Have at thee!

Announcing: The Forge

The Forge is now in live Beta at FORGE.ZHENG3.COM.

I’ve been eyeball-deep in HTML, CSS, Python, and WordPress for the last week, huffing and puffing as I try to wrangle The Forge into existence. There are going to be some bumps and bruises along the way, but I think I have a functional 3D model repository up and running. If you happen to find a rough spot, please email me or mention it in the comments. I’ll get some electric Bondo and gaffer tape on it ASAP.

I’ve got some fun plans for The Forge, but I’m also looking forward to a return to modeling my series of Magic: The Gathering tokens. Still got a whole mess of those to do. Stay tuned, freaks.

Quercus Leafreader, Squirrel Mage

When widow Winter refuses to yield to youngling Spring, ragged claws clutch-raking tattered tailfur.

When the kits, mewling, lick the last milk and teats sag empty in the frozen morn.

When fox and owl scour the ground and starving we fear to leave the nest.

When these are the times, the Leafreader comes and leads us to snow-forgotten treasure. Not a feast, but Enough.

And we live.

Quercus is the latest in my series of Magic: The Gathering Creature Tokens.

When your squirrels start getting up to +3/+3 and higher, you’ll need a model that conveys natural majesty. Enter Quercus Leafreader, Squirrel Mage. He holds aloft his Slimesinger (Sam the Slimesinger, if you must know) and dowses for acorns for the Scurry’s good.


Beast Token Fail


Failure of a Magic: The Gathering Beast Token. Off to the scrap bucket with you, buddy.

This model has become something of a 3D printing torture test for printer companies attempting to show off their new hardware’s abilities. Afinia managed to do it, and while this print failed above on a Replicator1, I’ve gotten dozens of these guys printed since. It’s like a beast farm ovah heah.

So if you think your kit’s up to the challenge, Download the model, give it a go, and tweet or email it to me. I’ll post it here.

こんにちは、日本! ここでは、このモデルをダウンロードしてください!

Magic: The Gathering Soldier Token


I’m still working my way through the list of Magic: The Gathering creature tokens. This time it’s back to White with the Soldier token.

This model is distributed under the Attribution-ShareAlike 3.0 US license. It’s also available as a 3D print on Etsy.

I’ve worked as a game industry artist for getting on ten years now (note to 16-year-old-self: WIN) and I frequently get the same question from relatives of non-industry kids who want to get into the business.

What should I study if I want to get a job as game artist? This question comes up at the dentist’s office more frequently than you’d think.

The short answer is: study art and, to a lesser extent, math. Study drawing first, then sculpture, then painting. Get at least a B in an art history class so you’ll have an inkling of how to evaluate your own work. Have a decent grasp of geometry and dabble at least a bit in C++ (or at least Python) so you know what the programmers are talking about.

Worry about learning software later.

But if there’s one sub-subject I’ve found incredibly useful over my career, it’s the study of armor. I can’t count the number of times I’ve had to create a suit of armor for a job. Or just a helmet. Or a dwarf rocking a hauberk. Or a Marine from 2125 with her ballistic combat gear.

If you’re a fantasy nerd you’ll probably absorb a fair amount of knowledge by osmosis, but be careful about learning from fantasy art. An awful lot of non-functional armor has been rendered by artists who haven’t hit the books.

Years ago before I retired from combat with the Society For Creative Anachronism (I kept losing dental work from being whacked in the helmet with rattan cudgels), I spent weeks building a lorica segmentada out of sole leather and 3/4″ copper pound rivets.

Wearing, tweaking, wearing, tweaking, and then fighting in that armor taught me volumes about pinch points, flexibility, and ventilation.

I wrestled with this model for a while, going back and forth between what’s historically accurate and what’s cool, all the while hindered by what’s possible with the current generation of 3D printers. I put in a fair amount of detail that I can’t print yet, especially in the skirt.


And of course I’m limited by the 3D printing overhang rule, so the pose is fairly static. And that sword is crazy large but that’s about the smallest it can get and still print reliably.

Unfortunately it’ll be a while until consumer-level 3d printing technology catches up with modeling and scanning software. I’m looking to create miniatures that are at least of the quality one can get from a large-scale foundry like Games Workshop. The resolution’s almost adequate on the Replicator 1, but the lack of ability to print overhangs is a huge limiting factor.

The Form1, and up-and-comers B9Creator and mUVe are looking like serious contenders for my next upgrade, since they appear to have a higher resolution and their prints are mostly unfettered by gravity.