In the Planet Construction Kit, I introduced some 3-D modeling programs, but that section is pretty outdated. What I recommend now (if you can’t afford a pro package) is Blender, which is free and full of features.  But like every other 3-D program, it’s complex and baffling and you can’t really figure it out just by messing with it.  So, I wrote a tutorial that gives the basics of Blender.


It isn’t a full manual… that’d be a book in itself… but you can get pretty far with it.  If people like it, I could add more (I only barely cover UV maps and creating humanoids).

Happy modeling!


The Overwatch World Cup Viewer is great for reviewing World Cup matches.  It’s also great for no-clipping around the world, seeing how the maps are put together and getting views you’re not supposed to be able to see.


For instance, above you can see the entire Nepal map.  All three stages are loaded at the same time, but you can’t see one stage from the next.

And here’s an unusual view of Ilios showing all three stages.  You can see this statue from Ruins; it’s interesting that it actually has a face (and belly button), which you can’t see when playing.

ow ilios 2

If you compare Blizzard World to the map of it, you can see that not everything is actually modeled. There are supposed to be a Spawning Pools Water Park and a Caldeum Market to the right, a Blackrock Mountain to the east, and a pirate ship in the water; none of these exist. But the rest of the park is pretty much all modeled, though only just enough to look OK from a distance:

ow bworld.jpg

The house marked with an asterisk isn’t even on the ground.  Also note the shadowy figures in the foreground… apparently this part of the park is still open, and has visitors. You can see them moving around as you play the map.

There are even cars and riders on the monorail– though they’re rendered as minimally as possible:

ow bworld 2

Here’s an unusual view of Hollywood. I’m really surprised that so much of the city is rendered, even if there’s also a lot of model re-use. You can see the theater where you spawn– the green roof in the middle background– so all of this is off to the left when you exit spawn, so most of it can’t be seen, even as Pharah. It’s interesting that they have enough of a polygon budget that they can model all this– including the backsides of buildings that you absolutely can’t see from the playable area. (And all those pipes and air conditioning ducts and curved roofs are really 3-d modeled.)


(I’m surprised because in Hammer, the level editor for the Valve games, anything you can’t see is scrupulously removed.  If you put a cube in the distance, only 2 or 3 sides will actually exist in the level. Evidently we now have polygons to burn!)

Here’s the theater itself– the green area is the lobby of the theater where you spawn. Behind it, a little disappointingly, there’s just some random tiny buildings; they didn’t block out the actual theater.


I wondered if the Rialto map has all the extra bits required for the Archives event (where you are the Blackwatch team sent to deal with the Talon guy).  Nope.  They obviously re-used a lot of the map, but not the extra parts (like the restaurant).

Finally, here’s something you’ve probably seen, but only while plummeting to your death. It’s the Omnic shantytown located under the King’s Row power plant.

ow kingsrow

This view is looking up toward the power plant. Again, this is suprising in the level of detail. You can see the track for the cart; the bright yellow circular thing just visible above the track is the dynamo (or whatever it is) above the final point.

Hmm, found some figures on the web. Alyx from Half-Life 2 has about 8000 polygons, which was a lot for 2004. (The Combine soldiers have only half that.) By contrast the Overwatch characters have 30,000, not including their weapons. That’s… a lot of polygons. So a few buildings with 100 to 200 polys are nothing to worry about.

(One trick which the game engine probably uses is to load low-poly versions of things that are in the distance. Still, the point is, the polygon budget is mostly thrown at the characters.)

Not content with playing Overwatch, I’ve been watching it– i.e., pro streams and games.

For the World Cup, Blizzard created a separate viewer, which lets you follow any player, and indeed control the camera. This is pretty damn neat, and I hope they’ll implement it for Overwatch League– heck, for any games.

You can also use it to look at the whole map in ways that you can’t when playing. Biggest surprise: the three-stage maps are really one map.  E.g., Lijiang Tower:


I really thought these were separate maps with skyboxes to show the bits of the other stages that you can see.  But with the viewer you can fly from one stage to the next: everything is there, down to the last health pack. Note that you can see some of the player info– the actual gameplay is at Control Center, but we can see Night Market in perfect detail. (And note that the spaceship spawns are there, although this isn’t the current stage.)

Another example: here’s a view of the Busan map showing both the temple and the city. (There’s still some culling that goes on– if you pull back far enough from the city, it disappears.)


And here’s a closeup on the hillside, showing that at this distance from the city, the trees
are just 2-d pictures on flat quads.  (You can see a bit of the city center to the right.)


What about the actual pro play? Well, I really enjoy seeing Space or Emongg play D.Va or Zarya, or Surefour playing anything, or Fareeha playing Pharah. I don’t know if I learn much, but some things amuse me:

  • Pro players still destroy everything in spawn.
  • Space changes his players-to-avoid after almost every match.
  • His ult tracking is amazing.
  • Wait times for Top 500 are terrible: 5 minutes or more.  Nice for streamers: they can look at chat.
  • Top 500 players still complain about unbalanced matches.
  • If someone’s out of position, the callout is e.g. “Zarya feeding.”
  • Surefour sounds infinitely chill.

And speaking of Surefour, if you watch just one pro game, find today’s Canada-France game and watch the Busan map, especially the Meka Base.  He has some game-winning Sombra ults.


As I’ve mentioned before, I’m kind of a graphics snob in video games. I just don’t like the look of Half-Life, Fallout 1/2, or Morrowind. So, I was surprised that Robert Yang managed to convince me that, in at least one case, lo-res is better.


This is a comparison of the original (2006) and a remastered (2017) version of a video game, Final Fantasy XII. And yes, Yang’s contention is that the fuzzy original is better.

(Why the girl is leaning to one side in the remastered version, I don’t know.  It’s distracting, but not the point here.)

His analysis of the “remastering” is helpful:

If I had to guess, the artists probably did this: (1) scale up texture by 200%, (2) increase contrast, (3) desaturated a little for that grayish next-gen feel, (4) apply a sharpen filter, (5) overlay a noisy detail texture on top to try to make the surface look more detailed.

He notes that you can automate this process, so you can handle a whole folder of images in a few minutes.

Now, my first reaction was that I liked the sharper image better. (I’ve never played the game, so I have no nostalgia here to invoke.)  In general, our eyes like sharpness! We can really see the intended patterns; the banner looks ten times better; the leaves are more recognizable.  It’s like putting glasses on!

And none of that is wrong. But look at the things Yang is pointing to: desaturation; the sharpen feature; a noise filter. The way I’d put it: the new image is

  • way too loud– it draws attention to itself, though it’s just a background
  • way too contrasty– if you looked at an actual wall, you wouldn’t be conscious of such a wide tonal range, it would mostly look one color
  • much less warm– look especially at the pavement, which has gone from a warm orange to almost black-and-white
  • too flat; because everything is in focus, it looks like a picture, not a world

You can certainly do realism well, but this realism done badly.

Yang points to another example, a fan remake of Half-Life 2.  I won’t name it, because I’m not going to say anything nice about it and there’s no need to embarrass a hard-working modder. Here’s a comparison.  (The top image is apparently another mod, but much closer to Valve.)


Oh dear. Let’s go over the problems.

  • What the hell is going on in the screenshot? It’s wicked dark.
  • You can barely see what is supposed to be the focus of the scene: Breen and Eli Vance.
  • Contrariwise, the modder has inserted extremely bright lights where they do no good at all. “Here, I really want you to pay attention to this: the floor.”
  • In general, the physical modeling and the lighting effects are far better– e.g. the round hole in the ceiling isn’t an obvious polygon; the lights, like real lights, don’t just light up the air. But all this realism just hides the narrative.  We don’t get an idea of the shape of the area; we can’t see what’s going on; we don’t know where to go next.
  • Why did he blur the red highlights from the windows?  Why did he lose the overhead light? Come to think of it, why don’t those very bright lights actually illuminate anything?
  • Yes, you’ve learned how to do a shiny floor; but what’s the point? All it does is reflect some lights and thus confuse the scene further.  Does the Combine care that much about waxing their floors?
  • What the hell kind an outfit did he put on Mossman?

Not all the images from the mod are this dark, but when they’re not, they’re generally too busy, too desaturated, and less coherent. They look like someone Googled for hi-res versions of every texture in the scene, without any care to making them work together.

Realism is nice, but isn’t an end in itself. You also have to think about consistency of style, and focusing the player’s attention on what is important, and giving them the information they need to follow the story and navigate the world.  The old Valve was very good at this.

As an example of a game that properly shows off the increased realism that’s now possible, I’d name The Witcher 3. I haven’t finished it, but good lord is it gorgeous. And without losing the readability, consistency, and focus that’s needed for a game to work as a game.

I had these drawing studies for my last gods picture and thought they might be an interesting process story.

The nice thing about these gods, Nečeron and Eši, is that they have things they can do. Nečeron is god of craft, so he can be building. Eši is god of art, so she can be doing art. But just that would be a little boring. From somewhere, but undoubtedly influenced by M.C. Escher, came the idea of each creating the structure that’s holding up the other.

Here are some doodles trying to make it work:


Nečeron’s bit is easy: he’s creating whatever Eši is standing on. (It starts as a table.) But what is she painting? Maybe some sort of framework holding up the platform he’s sitting on?  That’s the lower left drawing; it looked cumbersome.  Maybe a ladder (bottom right), but then he only has one hand free to work. Finally I tried a set of stairs, and that worked.

Here’s the second attempt at that:


I decided that the concept worked, but now ran into the next problem: I can’t really draw this scene out of my own brain. The figures here don’t look terrible, but the proportion and placement of the limbs was difficult, and the blobs representing the hands hide the fact that the concept requires four iterations of my personal drawing bugbear: hands holding objects.

(These are sketches, and would certainly have been improved if I kept working on them. But one thing I’ve learned is that poor proportions do not improve by rendering them really well. Better to get the sketch right.)

I tried looking for photos online, but getting these specific poses would be difficult.

Taking reference photos, however, is easy! I have an iPad! Here’s the pictures as they appear in Photoshop, with the sketch done right on top of them.


Who’s the model?  Oh, just some guy who’s available very cheaply.

If you compare this with the previous step, you can find an embarrassing number of errors in the original. E.g. Eši’s legs are way too small, the shoulder facing us is too low, and her neck is not drawn as if we’re looking up at her. Plus I think the final poses are far more dramatic.

I did the final outline over the purple sketch. Then the procedure is: select an area in the outline; fix the selection to make sure it includes everything I want, and fill it in on a separate color layer with a flat color.  Then go over each flat color area and use the airbrush to add shading. The bricks and stairs also get some texturing, added with filters. The jewelry is done on a separate layer with its own drop shadow— a cheap, quick way to add realistic shadows.

The gods aren’t wearing much.  That’s just how gods are, of course. On an operational level, there are two reasons for this (which we can assume are shared to some extent by Almean sculptors and painters).  The lofty level is that I like the human figure and hate to cover it up.  The less lofty reason is… clothes are frigging hard to draw. Figure drawing is hard enough, and clothing requires a whole new set of skills and rules of thumb, and looks terrible when you get it wrong. Plus, these are Caďinorian gods, so they should be wearing Caďinorian robes, which require, like, a black belt in drawing. They’re made of wrinkles. There’s a reason so many superheroes wear leotards: they’re basically drawn on top of the nude figure, with no folds.

The final picture:


Tonight, I like it; in a year, I’m sure it’ll dissatisfy me. Actually, when I look at it, I wonder if the angle of the iPad foreshortened the figures, making their feet proportionately too big. Oh well.



Hat tip to @jwz here. The technical name for these is apparently Image-to-Image Translation with Conditional Adversarial Nets. Here’s a link to a (currently) working interactive toy. It takes a simple drawing and turns it into a rendered nightmare.


Well, that wasn’t too bad, if you don’t look hard at the eyes.  Can it handle blonde hair?


Nnnnnno, I wouldn’t say it can.  OK, got it, dark hair. Maybe a more cartoony image would look better.


Well, maybe we should play to this thing’s strengths.  If I draw a monster, it should produce a monster, right?


I dunno, it kind of turned into Orc Gary. I wonder if I could import him into Skyrim.

So, who’ll be first to produce a graphic novel with this thing?

I have a couple of side projects besides all of the India.  One is Ticai, the game I started working on a few years ago. Here’s what it looks like today.


You can compare this to the last look from… gulp… three years ago here. What’s changed?  A bunch of things:

  • A better skybox, rather than featureless blue.  Still needs to be redone, but at least I know how now.
  • Even more buildings, including the nice round temple on the left.
  •  The cobblestones are bump-mapped, so they don’t quite look like a flat texture slapped down on a flat surface.
  • Previously the streets were modular; I figured it would be easier for Unity to render them if there was only one copy of each unit.  Then I realized that the entire street grid has fewer polygons than a single human model. So now the whole grid is one model. This cleared up a lot of little alignment problems and makes the streets look better. It also allowed me to do things like put the tower on the right on a little hill.
  • The camera stays closer to Ticai. This makes it harder (though not impossible) to see through walls and such, which helps out a lot in some of the smaller spaces.

Unity has been upgraded to version 5.4, which broke a few things.  Most are fixed, but something has changed about the lighting which I haven’t figured out.  Ticai’s clothes don’t look smooth, nor does the round temple.  Unity used to correct for that, and I don’t know how to fix it yet.

There were some major bits of the city that weren’t done yet.  There is a whole underground that was only mocked up; it’s all finished now. I also added an alchemist’s shop:


I like the various jars and things. There’s even a microscope!   Not shown: the alchemist has a rather pretty globe of Almea.

I’m convinced that one reason games are so often late and buggy is that the developers spend half their time redoing things.  You make something quickish just to get it working (possibly learning how to do it at the same time). Then you learn how to do things better, get dissatisfied with what you did, rip it out and redo it.  For instance, Ticai’s feet:


I half-assed her feet the first time… I figured I could suggest her toes using the texture, and it looks bad.  Finally I redid the toes, separating them in the model.  Plus I redid the ankles. Also her eyes: she has eyelids now, and blinks.  Her face still looks kind of weird, though, so I’ll have to work on that.

(The four toes are not a way of saving work: Ticai is Almean, so she really does have just four toes.)

I put the project aside before mostly because I was hung up on the writing side. The game is supposed to be a set of interlocked mysteries, which Ticai solves by running around and talking to people.  I want a really complex conversational engine, where you don’t have four options to choose from, but a hundred or more.  But of course that means a lot of writing, and even more testing, and I haven’t found a way to keep the amount of work under control.

The other project is a new conlang, something at least two or three of you have been waiting for patiently for years.  It’s Hanying, one of the language of the Incatena— in fact, the language of Areopolis, more or less Morgan’s native language.  I said it was “in origin a Chinese-English creole”, and it was… for the first half-century or so of its existence.  But it will be much weirder than that.  E.g., it suffered a series of phonological adaptations to new speakers twice, and it went through both some relexification and decreolization.  By the time it’s done I hope it really looks like something that survived a thousand years of change.

Next Page »