VALORANT Latency Guide – BEST SETTINGS FOR FPS AND LATENCY
I’ve done a couple of these latency guides now, including the first one on Rainbow Six Siege that goes into a lot more detail on wider factors that you might want to consider. I’ll link them in the cards above if you want to check them out. For VALORANT though, this is one of the tightest spreads for latency that I’ve seen. In CSGO the worst results I saw were over three times slower than the fastest. In VALORANT, we’re looking at, optimistically, like a 40% difference. The main reason for that is that there aren’t all that many settings available to tweak. You’ve got a couple of high/med/low settings, a couple of on/off options, and the only “advanced” feature I could find was NVIDIA Reflex. So, in short then, really no matter what settings you pick, you are probably having as close to the best experience as you can have. Still, for those that want the utmost in performance, let’s take a look at these test results and find out what works best.
Starting off with the base high, medium and low options, where everything else remains default, there is very, very little spread between these. High and Medium are functionally identical – like down to the 0.1 millisecond range, with medium only being fractionally faster in the minimums. Really no big difference there. Low is 1 millisecond faster, with a noticeable drop in the minimums too, although this isn’t anywhere near as precipitous a drop as I’ve seen in Siege or CS2. It becomes pretty obvious why that’s the case when you look at the FPS data. High and medium are running at 530 and 537 FPS respectively, aka basically the same. There’s really no big advantage here, both from outright performance and for latency.
There are a few more options to play with, starting with turning NVIDIA Reflex off. By default, at least on my test system, it was on by default, and turning it off technically did speed the average up, but it also increased the min and max values, making for a clearly less smooth experience. There’s also no major performance difference – save for the 1% lows seem to have rebounded. So if turning Reflex doesn’t help, what about setting it to “On + Boost”? Well that does improve things, dropping the average to 11.5 milliseconds and bringing the min and max results to the lowest we’ve seen so far. Interestingly the FPS actually went back down slightly. We’re still talking about a single-digit performance difference though, certainly not what I’d call noticeable or substantive.
Ok, what about disabling anti-aliasing and anisotropic filtering – two more post-processing style effects? Surprisingly, that makes a big difference, dropping the average latency a further millisecond, along with the min and max another millisecond each. That’s a pretty significant drop, although looking at the FPS, you can see why. Dropping anti-aliasing and anisotropic filtering lifts the framerate from around 550 to just over 600 FPS average. That’s a bit more significant, although there is something else interesting here. If I swap back to the latency results, but I include the frame time, as in how long on average a frame took to render, you can see that there is functionally no difference between the frame times. Even the worst result, high, is still only 1.89 milliseconds, versus the best we’ve seen so far which is 1.66 milliseconds. And yet we’ve dropped four milliseconds off the latency. This shows nicely how multiple factors affect the latency you’ll experience – it isn’t just more frames means lower latency, although that is generally the case, but certain settings actively delay new frames being displayed. Things like anti-aliasing are a post-processing type effect, so if you disable that, you skip a step and the frame can get to the screen faster. Of course, that ends up being a tradeoff between visual quality and performance. Anti-aliasing can make a massive difference to how the game looks, so personally I’d likely leave that one on, but if the ultimate in competitive advantage is what you’re after, you can turn it off.
Speaking of turning settings off, there are a couple more settings we can try disabling. With literally everything set to off, low, and Reflex on, we get… functionally the same result. It is 0.4 milliseconds faster on average, but I’d call that within margin for error. The maximum goes up by 0.3 milliseconds, and the minimum goes 0.6 down. Interestingly, the FPS performance does improve, going from 603 FPS to 631 FPS. Not bad, and quite surprising to see such little improvement in latency between them. The last thing I want to check is if dropping the resolution will help. I doubt it, as even at 1600p on this XMG Core 16 we are clearly CPU limited, but let’s test it anyway. On the latency front, we actually get a slightly higher result, settling between low with no AA and low everything. The FPS data shows a slightly slower result than at 1600p as well, indicating quite clearly that we’re CPU limited here.
As always, different configurations of CPU, GPU, RAM and display will all give you slightly different results here, although the trend should be the same across the board. The lower the settings, the higher the performance, and the lower the latency. Certain settings have effects outside the outright performance, like disabling anti-aliasing here, so it’s worth keeping that in mind too.
So, to sum up then, VALORANT has a remarkably tight spread due to its limited selection of settings. Enabling NVIDIA Reflex to “On + Boost” seems like a good shout, as does disabling anti-aliasing. Otherwise, it seems like you’re pretty good regardless of what settings you opt for.
