Red Blob Games: blog posts

In the last blog post I wrote about how I wanted to test the parameters of msdfgen, which generates multi-signed distance fields for fonts and other shapes. While testing the emrange parameter, I found lots of bugs in my renderer.

I also wanted to try out msdfgen’s new asymmetric range parameter, aemrange. The distance field goes both inside and outside the font edge, but if I want outlines, I want it to go pretty far outside the font. I made a visualization to show the range:

From this we can see that the interior has far fewer contour lines than the exterior. That’s because fonts are relatively thin. If it were a large object like a solid circle, it would have many contour lines on the interior.

In the last few posts I have shown some of the experiments I did with font rendering. Those experiments were all in the renderer. I’m using msdfgen-atlas to generate the textures used by the renderer, and I wanted to experiment with msdfgen’s parameters. Instead of generating new font data and then reloading the browser, I decided to try “headless” rendering controlled by a shell script.

Over the last few posts I wrote about things I did to improve font quality, such as antialiasing and combining distance fields to merge outlines and halos. But I want to “pop up the stack” a bit and talk about one of the bigger goals for this project. I want to render text in styles that I’ve seen in maps, both online and offline, both fantasy and real. In particular, I want to apply spacing, rotation, and curvature to the labels.

Learning about font rendering, I was looking at text closely last time, and I noticed another issue. The shadows of each letter overlap the previous letter. That’s because I’m drawing one letter at a time. So for example in the fl, I draw the f’s letter, outline, and shadow, then I draw l’s letter, outline, and shadow. So l’s shadow is drawn on top of f’s letter.

Last time I was looking at letter spacing with my renderer to see how it compared to Google Chrome on Mac. But while doing that I noticed that their antialiasing looked nicer than mine. So I tweaked parameters, including antialias edge width, gamma, and threshold bias.

My summer project is to work on labels for maps. In the previous post I described how I created outlines, and how I had a bug in the rendering. While looking closely at text to fix that bug, I noticed in one of my tests that the k and s seemed too close together. The h and e seemed a little too far apart.

In the previous post I introduced my summer project, to render labels on maps. As part of this, I want to be able to draw outlines, halos, and drop shadows. Signed distance field fonts are well suited for this. The basic use is to consider the signed distances -1 to 0 to be “inside” (filled) and signed distances 0 to +1 to be “outside” (transparent).

To draw an outline, we can add another range, maybe 0.0 to +0.2:

My friend L recently mentioned that he hadn’t seen any blog posts from me. It’s true, I haven’t posted for a while. Earlier this year I had explored signed distance field fonts, in particular using Viktor Chlumský’s multi-channel distance fields. I really enjoyed the many experiments I did, and I learned a lot. I had intended to use it in a real game project but the timing wasn’t right. So I put it away. Then before I got started on a new project, life happened. I had to take a break and attend to other things. I haven’t been blogging because I haven’t had projects to blog about.

Then in July I started thinking about other places I could apply the font code. I thought back to my 2019 blog post about map annotations, where I observed that even small amounts of text added to a procedurally generated map could make the map much more interesting. I started dreaming up a new map generator. Broadly, the tasks would be along these four categories:

1. Procedurally generate a map that has interesting “point” features including chasms, volcanos, waterfalls, and towns.
2. Identify large scale “area” features on a map such as peninsulas, bays, and mountain ranges.
3. Generate names for both point and area features, tied into geography, history, cultures, etc.
4. Place labels on the map corresponding to these features.

I had previously blogged about text effects, and how I accidentally discovered that I could alter the personality of the font by rendering a distorted shape. At the time, I was focused on text effects and didn’t want to get distracted by this discovery. So I wrote it down for later.

Well, later came, and I decided to explore it:

I have several pages that are unfinished because I can’t find an explanation I’m happy with. Sometimes while trying to come up with an explanation, I realize I don’t actually understand the topic as well as I thought! One of these topics is heuristics for the A* algorithm.While trying to understand the topic better, I came up with this example:

 

You may know me for my interactive tutorials. But before that, I was writing visual but non-interactive tutorials. In particular, there wasn’t a lot of information about A* on the web, so I decided to collect all my notes about pathfinding together in one place in the 1990s. But then in the 2010s I started making interactive pages. The newer pages are narrower in scope; I covered a broader set of topics on the older pages. I maintain both sets now.

Over the years people have asked me about “flow field pathfinding”. I felt like the early papers about it conflated the flow fields with hierarchical pathfinding, but I wasn’t sure, and I didn’t want to write about it until I was sure.

On my pages I often want to be able to move an object around in a diagram using the mouse or touch. Last year I spent some time learning about browser mouse+touch events, and wrote a page about event handlers for dragging objects around. I hadn’t realized it at the time, but it was only half the solution.

Earlier this year I was trying to improve font rendering in some of my C++ projects, and that led me down a rabbit hole of learning signed distance field (SDF) font rendering. I wanted to try out the SDF fonts in a real project. I occasionally help with Galactic Assault Squad (GAS), especially “engine” code, so for Week 6 I decided to try SDF fonts there.

In Week 4 of the year, I tried out various ways of using distances in signed distance field fonts. In Week 5 I wanted to do something different. I decided to explore what I could do treating each character as its own sprite and then applying sprite antimation. The week turned out to be fun but I didn’t learn as much as I hoped I would. To start, I copied the code from the previous week so that I would have a working program right away. Then I removed things I didn’t care about this week and added new code. This is like forking a project but for my weekly experiments I tend to copy the code instead of forking.

During the initial experiments I made an interesting discovery:

This is an ordinary font made more interesting by using a non-rectangular sprite.

In the last post I described how I fell into the font rendering rabbit hole. I try to put some time limits on each topic — otherwise I would explore forever! I try to pick a theme each week:

• Week 3 was the basics: SDF, MSDF, atlas, shader
• Week 4 was effects on individual glyphs: outlines, shadows, glow, bevel
• Week 5 was effects on how glyphs move: sink, rise, bold, slant, rotate, wavy, bounce, warp

The key idea I wanted to explore this week is that a distance field font can be thought of as contour lines:

When rendering a font normally like I did last week, I considered distance < 0 ? "white" : "transparent". But there are so many more things to do with this distance!

Each week I pick one or two things to work on. In week 2 of this year, I decided I should update my “hello world” OpenGL+Emscripten code from 2015. It’s boilerplate I use occasionally in other projects. It wasn’t compiling anymore, and I wanted to fix that as well as several other things.

One of the unsolved issues in that starter code was that the fonts never looked good. I was using the stb_truetype library, and got this output:

Look at how H and e are too far apart, and S and D are too close together. I tried various things but couldn’t figure it out, so I had two workarounds at the time:

1. Use Omar Cornut’s Dear ImGui, which has nice looking font rendering.
2. Use a monospace font to hide the spacing problems.

For some time now I’ve been unhappy with how much more friction there is when posting to my blog than posting to twitter. I keep wanting to blog more but I don’t. Part of the problem is content. I did blog more in 2018, when I was working on projects that had more to share. Part of the problem is expectations. On Twitter it’s expected that I write very little (280 character limit, up to 4 images or 1 animation). That constraint makes it easier to post. On my blog I tend to write longer more involved posts. But part of the problem is the process. My posts on Twitter take a lot less effort than my posts on Blogger. The Blogger UI got a lot worse in 2020, but it was already inconvenient for me.

So I’ve been thinking about what I actually want, and what I want isn’t Twitter or Blogger. I want something much closer to Hugo or Jekyll — a static site generator. I want to be able to save a file and have it become a blog post. I want to be able to grep over my existing files. I want to be able to write a perl script to fix something across pages. I decided a goal in 2024 is to switch from Blogspot to a static site generator.