• .
  • .
  • .

Project Pygmalion

As our work on Etch has been wrapping up, I’ve had a bit more time to put into the voxel-art denizens of Zpoc. You may have seen their current state in the “Early Engine Work” post, but with a little love and more than a little practice, they’ve come quite a ways.

Here are a few renditions of “Zpoc Girl” – note the better eyes, a taller body (it looks better in the Iso view as well) and the cowboy boots.

In fact, here’s Zpoc Girl in skirt. The interesting thing about Voxel Modeling at this resolution is the amount of detail you get to save into to the world. In Zpoc, the people you interact with won’t just be a silouhette-like mesh model, but can have all the internals you’d expect:

As you can see, this person is made of bone and stomach and lungs and brains and flesh.

The world itself is to be fully destructible as well, so the walls get the same treatment (siding, studs and insulation, drywall).

I don’t know about you, but this is what I’ve been wanting from a Zombie Survival game for most of my life. I want to see them coming through the walls, breaking through the floors, dropping through ceilings. I want to know that the brick house is a better choice than the wood and vinyl-siding beside it; I want to break off the bottom steps of a stairwell, and chop my way through a bedroom wall with a fireax to get away from the horde just outside my door …

For anyone else that feels this way, welcome home. Sign up for a chance to run our earliest Alphas.


Behind the Curtain

Content in ZPOC is based off of a voxel – a cube of single color much like a pixel. 32 of these in each direction make up a Vlock. Thousands of these Vlocks will eventually make up every game object – from the character you move to the brick wall against which you smashed the skull of yet another zombie. Bridging that gap, however, has proven to be quite tricky.

Version 1 of the Voxmill Editor... crude, yet functional.

Building a single Vlock is easy – we’re using a program called Qubicle. It’s nice when designing a single piece of a wall or a roof tile… but not so nice if trying to build, say, an entire wall with windows, curtains and the occasional door. So a new tool was created for the designers (see picture), allowing them to click-place Vlocks onto a three-dimensional map-type layout. It worked, though it didn’t really save any output. Why, you ask? Because we never really discussed how to get from Vlock to map…
Continue reading


VoxMill Engine – Early Vertex Buffer Test Videos

Here are a few videos of our early prototypes of the engine we’re developing for ZPOC. It is a voxel engine, though these voxels are at much higher resolutions that many other engines out there. If you look closely at the first video, you’ll notice the floors are grated floors over pipes: quite a feat considering how many voxels have to be drawn to create those. The framerates are actually about 20 higher than shown, as fraps is eating most of the frame.

A look at grated floors and extensive voxel modeling.

Another quick shot, working with lighting and render optimizations.

A first look at the BEPU Physics integration.


The Sirens’ Song

Devilsong was our Dream Build Play 2011 submission. We decided to enter this contest fairly spontaneously, signing up on May 16th … less than 30 days before the June 14th submission date. I still have the email, which starts and ends like this:

“– you MUST REGISTER TODAY.

No exceptions, if ands or buts … it takes 3 minutes. If you don’t do it tonight, you won’t be able to be on the team. Sorry for the rush.”

With no code written and no time to lose, we decided to do what any sane trio of coders would do – we went 3D multiplayer. We grabbed a couple friends who “do art” and Devilsong was born.

Devilsong is a throw-back of some of our favorite childhood gaming memories: Xwing vs. Tie Fighter, Wing Commander, FreeSpace and Jumpgate. It’s unfortunate, but since we’ve grown up, the space combat shooter has been sadly (and sorely) neglected. We didn’t have a lot of time, but we wanted to capture the essence of the core fun of those games – the pulse–pounding adrenaline rush as your squad and your opponents twisted through turns and spirals, cutting engines and firing missiles, hoping to out–manuever each other through gunfire and asteroids. We had great talks about a bigger game, one involving fleet actions, squad tactics and more dynamic mulitplayer … but for the contest we had to stay focused. We picked one map, an asteroid field, and two races (ship types in this case).

The lead up to DBP2011 was legend, the stories of which involve supportive significant others, long hours after work, weekends, rising significant other aggro, lots of caffeine, some sick time at work – an engagement proposal in the middle of the final weekend crunch, a new job interview delayed due to temporary burn out, bugfixing 2 hours before submission deadline and, of course, upload issues. It was an amazing experience and I don’t think anyone on the team would trade it in. What came out of it wasn’t just an incredible game, it was also a finished and published product. Actually finishing a game, getting it out there, making it to publish … is unbelievably difficult. Achieving that goal was a mental and spiritual milestone that helped propel us further and push us on to our next project: Etch.