Nvidia DLAA (Deep Learning Anti-Aliasing) is more of a niche prospect than its upscaler cousin DLSS (Deep Learning Super Sampling), though it has the same ultimate goal of making your PC games look sharper. Like DLSS, DLAA uses a dash of AI brainpower to stitch frames together in more detail than conventional anti-aliasing techniques like TAA and MSAA – only with DLAA no upscaling is needed. That doesn’t mean a performance gain, but for those with enough powerful Nvidia RTX GPUs, potentially superior image quality at all resolutions. Maybe even better than DLSS.
However, is DLAA worth using? I tried two of the most recent implementations, in No Man’s Sky and Deep Rock Galacticand found that any image quality benefits sometimes come at a hefty performance cost — even on one of the best graphics cards† On the other hand, since DLAA primarily requires a beefy GPU, the difference in frames per second may not even be noticeable unless your monitor has an esports-grade refresh rate.
Here’s everything you need to know about Nvidia DLAA, including exactly how Nvidia’s AI does the heavy lifting and what games can use it. You’ll find my quality comparisons and performance benchmarks a little further down.
What is DLAA and how does it work?
Explaining DLAA is a lot like explaining DLSS as they both rely on the same AI system. The work actually begins again at Nvidia’s headquarters, where a supercomputer feeds extremely high-resolution images into a neural network — essentially an AI model. These images teach the AI how to accurately predict the content of upcoming frames, which in turn generates anti-aliasing algorithms that enable rendered images with more detail and more intelligently placed pixels. The AI model, including the latest insights, is then sent to your RTX graphics card via a driver update.
Both DLAA and DLSS allow the GPU to apply these AI-powered algorithms when running a supported game, theoretically producing a sharper image than other dumber AA systems. The big difference is that DLSS involves upscaling, so games are first rendered at a lower resolution than your monitor’s, then reassembled to resemble native resolution, while also including that AI anti-aliasing. With DLAA, games are initially just displayed in native resolution. It’s much more of a ‘simple’ anti-aliasing option in this regard, even if getting there requires a supercomputer and some machine learning.
On paper this should allow DLAA to look slightly sharper than DLSS, all other things being equal it will always use a higher display resolution as a starting point. However, it also means that DLAA gives up the performance advantage of DLSS; The latter’s upscaling process isn’t terribly GPU-heavy per se, which along with the less demanding render resolution means games will just run faster with it. It’s also worth noting that Nvidia’s upscaling is very, very good: in Quality and Balanced DLSS modes, it usually looks just as good as native, if not even a little better.
It’s all very scientific, which I know isn’t hugely interesting to every PC owner. But the bottom line is that DLAA aims to improve the visual quality of games that can already run smoothly without upscaling, giving RTX graphics card owners the choice to dig into that extra FPS headroom in exchange for nicer anti-aging. alias.
Which games support DLAA?
Not much, indeed. DLSS support may be approaching the 200-game mark, but as of June 2022, the newer and more specialized DLAA will only work in a relative handful:
- Chorus
- Deep Rock Galactic
- Farming Simulator
- Jurassic World Evolution 2
- No Man’s Sky
- The Elder Scrolls Online
unofficially, death loop can also use DLAA. In the eyes of the game, it would run DLSS while rendering at native resolution, but since they use the same algorithms, that’s all DLAA really is. Setting it up is also more involved than clicking a single switch: you need to enable DLSS, set it to use Adaptive Quality, set the target frame rate to 30 fps, and finally enable the Quality setting. Strange, you know.
You may also be able to achieve a DLAA-like effect in a wider variety of games by combining DLSS with DLDSR (Deep Learning Dynamic Super Resolution). Another piece of Nvidia image technology for RTX cards, DLDSR is a downsampler that can improve image quality by rendering games above native resolution, and because it works at the driver level, it doesn’t require a game-by-game implementation. By combining DLSS upscaling with DLDSR downsampling, you should be able to get the AI anti-aliasing of the former with an internal render resolution balanced to match your monitor’s native resolution. DLAA, actually.
It’s a total job and requires at least some math to get the right resolution, but it works if you select the right combination of settings. If none of that appeals, then stick to the games mentioned above. Also, for the curious: no, you can’t have DLSS and DLAA enabled at the same time.
What does DLAA look like?
In terms of pure image quality, DLAA lives up to its promise of higher image quality – just plain. In the galleries below, you can see how DLAA compares to the highest quality DLSS setting, as well as native/TAA, in No Man’s Sky and Deep Rock Galactic (remember, you can click the photos to embed them).
It’s not a completely perfect form of anti-aliasing, as there are occasional frays, if not particularly pronounced: on that circular ship engine in No Man’s Sky, for example, which seems to smooth out even the basic TAA better.
But TAA’s whole deal blurs out details to hide fraying, which has the obvious downside of…blurred details. DLAA therefore generally looks a lot sharper, and especially in motion. Even if you can’t see it in these screenshots, tricky environmental elements like mesh grilles and distant stairs don’t shine as much with DLAA as they do with TAA. Or for that matter, DLSS. Both Nvidia technologies are, of course, very similar as they use the same machine-learned AA, although DLAA can look a bit more polished in the smaller details.
A zoom in on our space person’s space backpack, for example, emphasizes a sharper, more in-focus look of the textures and edges. And it’s not just close-up details where you can tell the difference. In DRG, a distant monitor displays noticeably clearer text using DLAA:
Again, TAA proves better at reducing the jagged effect, but only at the expense of making the whole scene look a little less sharp. In any case, for me it’s worth giving that up in exchange for DLAA’s overall better detail rendering. On this shredded slope, TAA even seems to be chopping away bits of metal, while it’s all rendered and presented correctly with DLAA.
I don’t want to exaggerate the quirkiness of DLAA either. It looks fine in motion, which should come as no surprise to anyone who likes to use DLSS, as the anti-aliasing is basically the same. It’s just the higher startup resolution that DLAA can use for a modest, but real quality improvement.
How does DLAA perform?
Since DLAA has no upscaling, that would never happen improve performance like DLSS could. But maybe, with all its AI smarts, it could impose less performance strain than old-fashioned anti-aliasing?
No luck. In both games tested, and regardless of native resolution, DLAA was consistently the heaviest at frames per second. And this was with a Nvidia GeForce RTX 3070one of the more powerful models in the compatible RTX range.
Admittedly, it was a difference of less than 10 fps with TAA at 4K, and even with both games’ Ultra graphics presets enabled, the RTX 3070 never wanted high frame rates at 1440p and 1080p. In any case, these results show the exact conditions that Nvidia had in mind for DLAA: when the performance is already so high that it doesn’t matter much whether you cut a few frames away to improve the image quality. Indeed, Deep Rock Galactic losing 36 fps at 1440p (compared to TAA) sounds awful, but you wouldn’t even notice the difference on a 144Hz monitor. And you have to very, very squint to see it on a 240Hz screen.
That said, results will vary depending on your own hardware setup, and if you can get by with an older, lower spec RTX GPU – the RTX 2060, for example – then I wouldn’t blame you for looking more closely at that performance gap between DLAA and DLSS. The 4K results in particular make a compelling case for letting you serve AI anti-aliasing with a side of upscaling. The little extra detail with DLAA is nice, but I’m not sure it’s worth more than a 50% speed boost, you know?
As such, DLSS is certainly still the king of Nvidia’s graphics wizardry: the kind you’d specifically buy a GeForce graphics card for (apologies for FSR 2.0† DLAA is neat, and it might well be worth switching over if you had the headroom for performance – it’s just not such a must-have feature, especially when so few games have built-in support for it.
Speaking of which, I wonder if game developers (and Nvidia) are missing a trick here. Other than The Elder Scrolls Online, all currently supported games are fairly recent releases. Understandable considering DLAA is a new technology, but wouldn’t it make even more sense to add it to older games? These would be easier to run than today’s releases, leaving GPUs with more performance space, and the extra detail reproduction could be like a fresh coat of paint on aging graphics. Just a thought.