Apex Legends Latency and Settings Guide – THE BEST SETTINGS FOR PERFORMANCE & LAG

If you’ve not seen one of these guides from me before, you’ll want to check out the first one I did on Rainbow Six Siege which explains more about what I’m testing, and some of the wider factors you might want to consider. As for this video, we’re gonna get straight into it! Apex doesn’t have all that many settings to play with – it does have NVIDIA Reflex, but there’s no settings for DLSS, FSR, XeSS, or even quality presets. We do have the usual Ultra, Very High, High, Medium, Low and Very Low when it comes to the texture streaming budget, but otherwise it’s only a handful of individual settings. The texture streaming budget does include recommendations for how much VRAM you should have to use that setting. Now I’m testing with this RTX 4060 Laptop system which has 8GB of VRAM – albeit on a tiny 128 bit bus – so in theory this setting tweak won’t do much, but we’ll be testing that to find out for sure. So, let’s get testing!

Now the first result here is on default settings – save for VSYNC being turned off, as it is for every test here. That means an ultra streaming budget, 2x anisotropic filtering, high ambient occlusion, and the rest of the settings you see here. This is really close to using the maximum settings by default, so we can expect this to be on the slower side. Now just showing you a single result isn’t all that interesting, so the first comparative test I’ll include is simply turning Reflex on. Just looking at those two, you can see how much of a difference Reflex can make. We go from 28 milliseconds of end to end latency, to just 20.9 milliseconds. Now that’s still pretty high, so let’s see if we can do anything about that. 

One of the settings I haven’t mentioned yet is the adaptive resolution FPS target. This essentially drops the render resolution on the fly to try and maintain a target FPS, which by default is set to 60 FPS, and the highest option is only 100 FPS, so as you’ll see that isn’t all that useful here. Strangely, disabling it only serves to add a tiny bit of latency, bringing the average up to 22 milliseconds. That’s still an awful lot better than without Reflex on, but it seems it’s worth leaving that one on. Interestingly, looking at the performance numbers, so far the highest we’ve had is actually the default settings run at 183.6 FPS average. As we’ve seen, turning Reflex on generally either doesn’t do much, or like in this case, provides a little less performance as it works to shorten the delay between inputs and frames showing up on screen. What’s strange about the AA only run is that it provided functionally the same performance as the Reflex and Adaptive Resolution FPS Target run – so the added latency (as small as the difference may be) is from a pipeline step, rather than just missing FPS.

Now I think it’s worth having a play with that textures slider. Since we were testing on Ultra thus-far, our only option is downwards. Even with the same AA only setting, swapping to very high textures did provide a slight drop back down to the Reflex only result at 20.9 milliseconds. High textures went even further, dropping a whopping… 0.2 milliseconds. Amazing, I know. Medium was a fraction higher at 21 milliseconds flat, and Low was a few microseconds slower at 21.1 milliseconds. Interestingly, looking at the FPS results, there really isn’t much in it here. Low did push ahead slightly, but it’s so marginal. So, unless your system doesn’t meet the VRAM requirements, it seems just changing the textures setting won’t have much of an effect.

One thing I feel it’s important to mention is that in my testing, Apex had a bit of an usual behaviour. Most results from basically every other game have been pretty consistent. They all generally come out as a singular band, but in Apex, basically every single test had this pattern. For the first 50 moves it would get a considerably lower result, then after 50 it’d settle in at a decently higher average. I’d normally do 100 moves, but because of this behaviour I upped it to 250 instead and you can see that the higher result stays consistent after that. It’s hard to know how this would reflect in your actual user experience, because this wasn’t like a new game load-in every time. The only thing that changed was the settings while still in-game, and I tried leaving a longer gap between changing the settings and starting the test, but that didn’t change the behaviour, so I really don’t know. Still, I thought you should know as it’s an important factor in discussing these results.

Now, seeing as changing the texture settings didn’t do much, let’s try something more brash. Let’s set everything to low, or off. The lowest setting we can get. Finally, that has an effect! We drop the average latency to just 17.3 milliseconds, the first significant change since turning Reflex on! What I find most interesting here is actually in the FPS results, where we can see a significant jump in performance, going from around 175 FPS, to 240 FPS, and if you convert that to frame times, that’s around 5.7 milliseconds and around 4.1 milliseconds, or about 1.5 milliseconds difference. That means most of our latency benefit came straight from more FPS, although there is likely some level of step-skipping in the pipeline that helped a touch too. Still, it’s clear that more FPS gives you a competitive advantage here. 

Lastly, I wanted to try dropping the resolution from this laptop’s stock 2560×1600, down to 1920×1200, to see if that makes any difference. Remarkably, it does, dropping the average to just 15.3 milliseconds! What’s more interesting for me is seeing the FPS results aren’t any different, which is quite a surprise really. VSYNC is off, so the FPS capping at 240 – the screen’s refresh rate – is a bit of a weird behaviour. Still, lower latency is possible with lower resolutions, at least on this machine.

So, regardless of your settings, enabling NVIDIA Reflex seems to be your best bet to getting the competitive advantage in Apex Legends. If you want the utmost advantage, setting everything to low settings might be useful, and you can even try dropping the resolution too if you can stomach the visual quality difference. Of course, these results are still somewhat specific to my system, so if you want to be able to test games like this on your own system, you can check out my Open Source Latency Testing Tool over on OSRTT.com. I hand build each unit right here at home, and ship them worldwide.