SDF vs. MSDF vs. Slug: GPU Text Rendering (alphapixeldev.com)

132 pointsby ibobev12 hours ago52 comments

psyclyx 5 hours ago

Slug mentioned!

When the Slug patent was released to the public domain, I put together[1] Snail[2], a Slug implementation in Zig.

One thing I found was that it was tough to get small text to look good with some fonts. TrueType fonts often have bytecode that tweaks curve points to better fit the pixel grid at a particular size. Part of the pitch for Slug is that it doesn't require per-size glyph prep, so Slug text is just unhinted.

As monitors have gotten denser, the major font renderers have moved away from bytecode hinting, either toward auto-hinting (ignore the bytecode, look at the outline, and decide what to do) or toward no hinting at all.

Snail has GPU auto-hinting that tries to replicate a lot of that, so I could have hinted text without per-size prep. It precomputes knots at various glyph features, and the shader then moves the knots to stretch/squeeze parts of the glyph. It's not perfect (especially for serif fonts!), benefits from per-font tuning, and focuses on Latin glyphs (pretty much always draws CJK glyphs unhinted because they're too complex for the current shaders to handle). Also, cases where you want hinting and wouldn't be better served by just scissoring prepared bitmaps aren't all that common. It also includes a TrueType VM, for cases where the per-size cost is acceptable (DejaVu Mono looks decent with the auto-hinter, but it has phenomenal hinting bytecode).

It was a lot of fun to put together, I learned a lot, and as far as I know the auto-hinter is unique among Slug implementations. Slug is a very cool algorithm.

[1] through high-effort delegation to Claude [2] https://github.com/psyclyx/snail - includes some diagrams (made with Snail!) that explain most of the prep/rendering of a glyph with Slug

elcritch 3 hours ago

> One thing I found was that it was tough to get small text to look good with some fonts.

Same here. For scaling from medium to large fonts the MSDFs were good. However at smaller "normal" font sizes they just didn't tend to look good even with larger MSDF sizes.

Plus for font rendering the GPU overhead was too large to be worthwhile for normal GUI apps. Font atlas'es are pretty small, and most fonts for GUIs are statically sized.

However I did find MSDF are fantastic for rendering complex SVG paths with lots of benefits [1]! You can use a 64x64 SVG star and scale it up fullscreen with reasonable loss in quality [2] (or use 128x128 for almost lossless images).

I use them as the core of my GUI library to render complex SVG paths for icons and such and compose individual paths on the GPU.

You can use them to implement fair chunks of Lottie animation as well [3]. Though I never got enough into the Lottie work to go beyond basic demos, but you could easily scale MSDFs, add shadows, feathering, outlines, etc at 100+ FPS.

1: https://forum.nim-lang.org/t/14062#85314

2: https://github.com/elcritch/figdraw#msdf-bitmap-based-sdf-re...

3: https://github.com/elcritch/lotty

GuB-42 11 hours ago

I once implemented SDF text rendering. To me, the thing I liked the most about this technique it how easy it is to add effects on top. With a few lines of shader code, I had outlines and softening of the edges (antialiasing). I didn't try the MSDF variant, as I didn't mind corners not being sharp when scaled up, so I don't know if these effects break on MSDF.

Slug doesn't seem to support any of these, it is just "for a given point, am I in or am I out?", but it doesn't tell you by how much, which is great on really high resolution displays and large sizes, but you would lose the ability to do the kind of effects you can do with SDFs, and have to deal with antialiasing separately.

JoshTriplett 10 hours ago

Having a technique that optimizes solely for whole pixels seems relatively reasonable, given current hardware PPIs. Systems that want to handle grayscale could use a different technique for that.

That said, there are interesting effects other than grayscale antialiasing, and I wonder how well slug could handle things like outlines (useful for subtitles over video, to make them readable on any background).

eviks 9 hours ago

I don't follow, current hardware PPIs are generally low, while whole pixels work for high?

JoshTriplett 9 hours ago

> current hardware PPIs are generally low

Not especially, unless you're running something like a high-refresh-rate gaming monitor that's still 1080p or less.

On most current phones, laptops, and monitors, I would expect the difference between grayscale antialiasing and monochrome rendering to be hard to notice.

mncharity 4 hours ago

> unless you're running something like a high-refresh-rate gaming monitor that's still 1080p or less.

Or HMDs. Upcoming high-resolution PICO Space Pro is said to be ~45 pixels/degree max. Versus ~50 ppd for a 1080p laptop. Formerly known as Project Swan, its September global release announcement was postponed until... unclear.

exDM69 7 hours ago

> Slug doesn't seem to support any of these, it is just "for a given point, am I in or am I out?", but it doesn't tell you by how much

It does give you an anti-aliased value between 0 and 1 that estimates how much of a pixel is being covered.

But this is a linear estimate based on horizontal and vertical distance to the Bezier curve. It does not look correct at long distances, which is why you shouldn't use it for outlines, drop shadows or the other cool things you can do with (M)SDF. A single pixel outline works fine but is not really legible with modern display resolutions (very thin lines).

For finding the minimum distance between a quadratic Bezier curve and a point would require solving a 3rd degree polynomial, where Slug's algorithm gets away with solving a quadratic equation per pixel. This is makes a big performance difference.

mattdesl 11 hours ago

I've also been working on a GPU curve renderer, Windfoil, based on a formulation that Fable 5 originally proposed to me during a directed search [1]. It is similar in some ways to Slug, not always as fast, but uses less shader storage (single band instead of two) and produces higher quality anti-aliasing i.e. closer to a box-filtered ground truth.

It may be of interest to some game/graphics devs here...

[1] https://github.com/texel-org/windfoil-algorithm

yeoyeo42 11 hours ago

interesting. how does the AA work in your algorithm?

the reason slug has two bands is precisely because of its anti-aliasing. you having only one implies you do the AA differently, or are less efficient with searches off the main direction.

for readers: the slug shader traces two rays, one horizontally, and one vertically, to find out if a pixel is inside a glyph or outside. one would be enough for pure inside outside, but for anti-aliasing purposes it's also helpful to know how far away you are from the closest edge. but if you have a horizontal ray running parallel to a horizontal glyph edge (not uncommon), the ray glyph intersection will return no value at all. so slug traces two rays and blends between them for anti-aliasing that always works in both cases. the bands mentioned here are just acceleration structures, basically a list of cells where each tells you which parts of a glyph are contained within it.

it's still approximate. a true ground truth would just supersample and do many point-in-glyph tests within one real pixel - at edges some of them would be inside, and some of them would be outside, giving you a smooth value to display depending on the shape of the glyph within the edge pixel.

mattdesl 11 hours ago

Windfoil uses boundary integrals to average winding over a given rectangular footprint (normally one pixel, independently in a fragment shader). For many cases, like typical text glyphs, this will reproduce a box-filtered ground truth exactly, i.e. for those shapes it is not approximating AA like Slug or MSDF. But because it takes the average and then applies a fill rule after, it's not an exact in all cases (fwiw, Slug can produces similar errors with tricky self-crossing curves and such).

jdanford 12 hours ago

Man, I am getting incredibly tired of reading LLM-generated writing

cute_boi 11 hours ago

Author probably didn't even read...

kevin_thibedeau 11 hours ago

I think it's TLDW.

20k 11 hours ago

What signals this as being LLM writing for you? I'm crap at noticing specifics here

nurumaik 10 hours ago

For me it was "head to head" table. Typical complete nonsense llm-generated comparison table. Rest of the content is still quite good tbh, generated or not

JoshTriplett 10 hours ago

Not arguing whether this article is or isn't LLM-generated, but I've seen skewed comparison tables like that in marketing since before LLMs. Pick a bunch of qualities skewed towards the new thing being proposed, highlight all the ways the new thing wins.

poly2it 10 hours ago

> A note on fairness, because the technical reader will ask.

I think it's a mix of human and LLM writing.

nuxi 7 hours ago

Stuff like this is a giveaway for me:

> Drawing it on a GPU, crisply, at any size, under any 3D transform, while the text changes every frame, is not.

> One small texture, resolution independent within reason, one cheap shader.

English is not my native language, so I may be wrong here.

tokenscoper 7 hours ago

For me, it was this part of the paragraph:

> There is no atlas, so a hundred thousand CJK glyphs cost a font's worth of outline data, not an atlas the size of a video. Text can change every frame at no baking cost, which is exactly what you want for live data, user input, and localized content. And it all happens in a single draw with an ordinary fragment shader, no vendor extension required.

Specifically, these sentence structures:

1. ... so <blah> costs <blah>, not <blah>.

2. And <blah>, no <blah>.

3. ... fixed resolution ahead of time, which quietly ...

mbrumlow 5 hours ago

They were the trained on human text right? So maybe that is the way human text as whole reads.

Maybe what we are seeing is us internet dwellers being exposed to wiring styles of humanity we had not been exposed to ?

furyofantares 5 minutes ago

I won't try to convince anyone with specifics but it was definitely LLM generated. I think by someone who knows what topics they want to cover and did have something they wanted to post about.

I learned a few things from it and have some interest in the topic. So this sort of a weird valley where I'm very annoyed by the writing but also getting something out of it.

I have written a blog post about SDFs before, which I tried to author with an LLM helping and the text just came out awful. I think I rewrote my entire post eventually - I am fairly confident at least that none of the LLM's words remained. I mention it because I've already read the SDF section of this blog largely, when an LLM generated mostly the same text 5 months ago.

Ithildin 7 hours ago

Then don't. I'm incredibly tired of people complaining about it. This is the world now. We're going to be interfacing with these models until we take the long nap. A models personality and prose are fine things to comment and give feedback on, but this belly-aching is worse than what it said in the first place because it's not constructive at this point.

tokenscoper 7 hours ago

As much as I'm all for human generated writing, coming across and having to read through mostly-LLM-generated writing is becoming the norm (at least for me, it's true). While I may not necessarily approve of the manner in which this was written, it was still something I wanted to know more about and I read it anyway.

It does bias me negatively towards the author, and makes me a little sad that people are choosing to not put more thought and effort into their writing, but this is a trend that is here to stay.

Making a big fuss about this hasn't gone anywhere from what I can tell, and it's unlikely that the outcome will be any different going forward.

marssaxman 2 hours ago

I wonder if you are more or less tired of reading LLM-generated writing than I am of reading complaints about LLM-generated writing.

YuechenLi 7 hours ago

This article has some inaccurate information since MSDF atlas doesn't have to be baked statically, so the "CJK character means huge atlas" is not really an issue if you can async upload them to the atlas.（Async outline extraction is a bit harder for C libraries, so that's why the pipeline is mostly restricted to atlas upload).

The other thing is MSDF rendering is fairly cheap and can be done on the CPU quite easily without GPU shaders, the atlas generation/upload is the expensive part, but it's a one-time cost per character per font as the atlas is just a normal bitmap texture file, MSDF text have sharp edges at most zoom resolutions, and the small size text is better handled with simple CPU raster anyways.

I really failed to see significant benefit of using Slug over MSDF + raster fallback for small fonts, it's definitely more exact, but I'm not sure if the marginal resolution benefit is worth it over much more complicated GPU dependent rendering, so I'd really want to test it out myself when I have the time over taking the word of an obviously AI written article for it.

decoding 5 hours ago

Yes, there is at least one downside to MSDF among a set of possible downsides (huge atlas, or dynamic rebaking, or limited glyph set), and the article is written as though they all apply, which is a bit misleading. Slug is cool, glad it exists, but it isn't the right fit for my rendering project either.

exDM69 7 hours ago

As hobby project, I wrote a font rasterizer algorithm (two algorithms actually) that produces pixel perfect anti-aliased images (like Slug) using a similar but different algorithm. Rather than using Slug's clever root classifier for Bezier curves, I subdivide Bezier curves into monotonic sections where evaluating the winding number is much simpler and can be done in parallel for a group of pixels. It works nicely on GPU with warp/wave/subgroup operations and CPU using SIMD.

At first glance, subdividing the Bezier curves sounds like a bad idea (more Beziers to rasterize) but it opens doors for some parallelism, and most Bezier curves that appear in fonts are monotonic in the first place (so the increase is very modest). This was inspired by this entertaining but not very serious video about font rasterization [0].

The first parallelism optimization is checking against the curve bounding box vs. a rectangular (in uv-space) region of pixels, and this can quickly determine if the Bezier needs to be evaluated in the first place. This can be done per GPU warp.

The second optimization works only for rectilinear transformation (no rotation, skew or perspective). Solving the quadratic equation involves a square root and a division (which alone are >30% of the computation), which can be computed for each row and column of pixels instead of for each pixel (2n instead of n^2).

Both optimizations rely on mathematical invariants of monotonicity, ie. the derivative of the Bezier curve must be non-zero. All Bezier curves can be robustly subdivided into monotonic sections using de Casteljau's algorithm.

My simple benchmarks compare favorably to Slug on the GPU and to "fast" rasterization algorithms on the CPU (which is an order of magnitude faster than "fancy" rasterization algorithms with hinting etc).

Unfortunately there are so many hobby projects and so little time. All I have is messy shaders that draw individual characters and a few benchmarks to see how quickly (and something similar for the CPU). Going from there to a complete text rendering system would be a lot of work. Writing a more detailed article with illustrative code examples is something I'd want to do but haven't gotten around to.

If you want to offer words of encouragement or geek out about rasterization algorithms, I welcome any input.

[0] https://www.youtube.com/watch?v=SO83KQuuZvg Sebastian Lague - Coding Adventures: Rendering text.

Const-me 10 hours ago

“What makes glyphs hard” Another reason is hinting. Traditional text renderers like FreeType are aware of the pixel grid and they adjust the curves slightly, snapping them to that grid. For all methods in the article quite hard to do on GPUs.

“Chinese, Japanese, and Korean have tens of thousands of glyphs, and baking all of them at several sizes is a memory disaster” One possible solution is dynamic atlas built on CPU for visible glyphs only.

“The distance-field panels notch, where interpolating between stored samples no longer matches the true curve” Can’t it be fixed in the shader, using screen-space derivatives of the SDF? I think in theory, SDF value for pixel center combined with screen-space gradient vector of that number delivers enough data to compute partial coverage for the pixels on the edge.

flohofwoe 11 hours ago

Here's a simple Slug rendering example on top of sokol_gfx.h:

via WebGPU backend: https://floooh.github.io/sokol-webgpu/slug-sapp.html

via WebGL2 backend: https://floooh.github.io/sokol-html5/slug-sapp.html

There's quite a bit of helper code plus stb_truetype.h and stb_ds.h under the hood to parse TTF files and crunch the TTF curve data into the runtime format expected by the Slug shader (this stuff should better go into an offline asset pipeline tool):

https://github.com/floooh/sokol-samples/blob/master/libs/slu...

...the actual text rendering code is also taking a couple of shortcuts, e.g. no kerning, no right-to-left, and also no text shaping.

There's also a new and complete text rendering stack by Mikko Mononen called Skribidi (AFAIK not based on Slug though):

https://github.com/memononen/Skribidi

...the list of external dependencies is a bit scary for a small self-contained sample though (Harfbuzz, SheenBidi, libunibreak, etc...), but that basically shows that proper international text rendering is really damn hard, even when trying to simplify the code as much as possible.

whizzter 10 hours ago

... or why it's best to just use the OS libraries to render to textures if you only need an OSD :P

Question though, you implemented the slug system? The article mentions root eligibility but doesn't expand on it, what's the point of it because slug doesn't seem too "magical" for only doing winding counts?

seanw265 11 hours ago

Interesting read. Text rendering techniques have always fascinated me.

I'm a bit confused because at some points it seems like the author is conflating "tessellation" and "Rive". Are they the same thing? As an uneducated reader, my understanding would be that Rive is an implementation of a renderer using the tessellation approach. But surely a generic tessellation approach could support perfect arbitrary transformations even if Rive doesn't?

Maybe there's something I'm missing.

Also, a nitpick: in the "head to head" section, the author highlights Slug's better performance in green for the entries that it wins (or ties). For the sake of fairness, shouldn't we highlight the winners in every category? Surely Rive's "low" memory usage beats Slug's "moderate"?

pavlov 11 hours ago

> "In 2017 Eric Lengyel published an algorithm, called Slug, that stopped dodging. It renders glyphs directly from their outlines in the fragment shader, with no texture atlas and no per-frame tessellation. Lengyel patented it in 2019, and on March 17, 2026 he dedicated that patent to the public domain."

It's nice that he gave the patent to public domain, but this is not how patents are supposed to work. You can't patent something two years after it was already published.

I'm guessing he actually filed for a patent before publishing, and the article should read: "Lengyel was granted a patent for it in 2019"

It's an AI-written article, so maybe it's not reasonable to expect it to be consistent on this level...

flohofwoe 10 hours ago

The details are in Eric's post here:

https://terathon.com/blog/decade-slug.html

The timeline basically checks out, maybe it simply took the patent office over a year to process the patent application?

eviks 9 hours ago

No, application filed 2018-02-01, after the publication. Though yes it took more than a year to grant it

eviks 9 hours ago

He filed in 2018-02-01, after publishing in 2017. What do you think breaks in the patent system with such an approach? https://terathon.com/blog/decade-slug.html

pavlov 3 hours ago

You need to file before publishing to establish the priority date.

A sibling comment notes that he indeed filed a provisional patent in 2017 before going public. That’s the part that the article muddled. 2019 was when the patent was granted, not filed.

elengyel 6 hours ago

* The provisional patent application for the Slug algorithm was filed on March 27, 2017, and this is incorporated into the final patent at the beginning of the Description section. A provisional patent establishes a priority date for everything it discloses (which was the whole algorithm), and it gives the inventor one year to complete the application process.

* The JCGT paper was published a few months later on June 14, 2017, after the priority date.

* The final patent application was submitted on February 1, 2018, before the one-year deadline.

* The patent was granted about 1.5 years later by the USPTO on August 6, 2019.

jayd16 10 hours ago

Is there a runtime comparison of the shaders needed? Seems like you might have to do a lot of texture sampling if you have to walk the ray through the data but maybe there's a trick?

MSDF is pretty much just the target texel in question plus the surrounding samples in a way the GPU can do entirely upfront before the math starts. (M)SDF glyphs also play nicely with mip mapping and I would think this Slug algorithm needs uncompressed data. Maybe that doesn't matter because you just don't scale the data ever?

flohofwoe 10 hours ago

Slug doesn't use "bitmap" font data but more like parameter lookup tables in storage buffers, pre-processed from TTF files. The lookup data can of course also live in textures (so that it works in WebGL), but using those textures as "poor-man's storage buffers", e.g. fetching the data directly without filtering.

The pixel shader is indeed quite complex compared to SDF:

https://github.com/floooh/sokol-samples/blob/8afa83928ce1870...

rezmason 11 hours ago

Slug looks impressive!

The project I'm most known for is basically an MSDF shader with a bloom pass. It serves my needs 100%, though if I expand to support arbitrary text, I may reach for Slug.

The one issue with MSDF I want to raise is, it seems everybody uses the same msdfgen texture creation program from Viktor Chlumský's master's thesis 11 years ago. I wish there were other implementations. Who ever heard of a graphics technique that was only ever programmed once, and then used everywhere without substantial iteration? We need to de-XKCD-2347 MSDFs for everyone's sake, including and especially Chlumský.

torginus 10 hours ago

I haven't heard of Chlumský or MSDF before, but I think a lot of people have been aware of the technique before he published his paper, thanks to Valve's TF2 text rendering paper:

https://web.archive.org/web/20120505013814/https://www.valve...

The pdf details only simple SDF, but in the closing paragraphs, it mentions the weakness of the technique and mentions how it can be solved with multiple SDFs, but the exact technique wasn't showcased.

Which led to people trying to reverse engineering it, and making their implementation, for years (it's Valve after all). Including me. Not sure if what I came up with was exactly MSDF, but certainly there are a lot of implementations out there.

rezmason 10 hours ago

We mostly agree! I should elaborate.

I'd say what you and the others did was research, not implementation. Your results were solutions to the same problem, not variations of the same recipe. What you did is important but separate from what's concerned me.

A recipe of Chlumský's— the msdfgen utility— is widely used, but only has one producer. I'm just saying that that's a liability. Like, imagine if HarfBuzz was the only text shaper, and was maintained by one person.

sirwhinesalot 11 hours ago

There's another GPU text rendering algorithm missing from the comparison: Rook & Possum's Scanline Sweeper:

https://rookandpossum.com/posts/scanline-sweeper/

Sean Barret (creator of the stb public domain libraries) independently invented a CPU-based implementation of the same idea, used in stb_truetype.

logdahl 11 hours ago

I knew the second Eric posted about releasing Slug to the public domain we'd get 100s of "OpenSlug" slopped-up. Will be interesting to see which implementation will win or if Slug will keep being a thing.

flohofwoe 10 hours ago

Tbf, the Github repo doesn't look particularly 'sloppy' to me, more like some mild AI assistance. There's several months of fairly regular looking git history, a couple of commits with an LLM disclaimer for python bindings, and two llm-context files.

https://github.com/AlphaPixel/slughorn

bel8 12 hours ago

MSDF looks better than Slug for me in the first example.

But slug wins in perspective in my eyes.

I use MSDF to render crisp text in my webgl hobby game. Hope to publish it with source code when I get the time.

thanks for sharing the article. I'll take a deeper look at it later.

Keyframe 12 hours ago

I do MSDF as well and since I'm not doing a huge-ass text editor (and even then) it works and it's great. One thing to consider is that with any solution that does things directly from vector outlines means you also then have to distribute proper font outlines and you might not have a license to do that. With MSDF and similar atlas-like solutions, you pre-render a font and distribute that rendered image instead of an actual font.

dcrazy 10 hours ago

Which you also might not have the license to do. I once licensed a Monotype typeface that prohibited its use in anything editable, including fillable PDF forms.

hncbw02z5a 11 hours ago

MSDF corners gave me trouble til I bumped the distance range at small sizes.

reactordev 11 hours ago

Excellent write up. The cited references are on point too. Eric Lengyel is a legend.

LoganDark 7 hours ago

> By Chris Hanson

What a name!

(Skipped the rest of the article cause it's AI.)