Getting Your Speedrun Timing Right on the Xbox Series X

I spent about three weeks trying to nail frame-accurate splits on a speedrun for Elden Ring on my Series X before I figured out why my times were consistently 400 milliseconds slower than runners on PC. The console isn't necessarily slower at running the game, but the way it handles frame pacing and display sync throws off any timing tool you hook into it. Once I stopped fighting the console's output and started working with what it actually does, everything changed. What people are looking for when they search Gameplay Speedrun Xbox Series X is a combination of recording tools, frame counter overlays, and input-lag testing methods that let you break a game down into segment times. It's not one app you download. It's a setup. The core pieces are OBS or your console's built-in capture system, a frame counter from an external source, and a split timer you can reference without switching apps during playback. The Series X outputs at a fixed refresh rate depending on what the game supports, usually 60Hz or 120Hz, and that matters because your split timing precision depends entirely on whether you're counting in frames or seconds. One frame at 60Hz is 16.67 milliseconds. At 120Hz it's 8.33. If you're tracking splits by the second, you're leaving over half a frame of accuracy on the table, and speedrunners will notice.

Gameplay Speedrun Xbox Series X Setup Method

Start with the frame counter. The Xbox has a built-in developer frame rate counter you can enable through Settings > General > TV and display options > Frame rate, but it only shows current FPS, not a running count you can use for splits. You need something that timestamps each frame. A capture card running at 120Hz connected to your PC works best. Elgato 4K60 Pro+ won't cut it if the game runs at 120fps, since it caps at 60fps passthrough. The Sony HD700 or even a cheap AVerMedia live gaming card set to 120Hz input will let you see every frame. From there, route your Series X HDMI output through the capture card into your PC, run OBS or Streamlabs, and set your recording resolution to match the game's native output. Don't upscale. If the game runs at 1440p, record at 1440p. The Series X will downsample to 4K for output, but the frame data is already baked at the game's native resolution, and capturing at the wrong scale introduces frame interpolation artifacts that throw off frame-perfect split timing. This is the first thing most people mess up. For the actual split timer, use SplitTime or StopTimr on your phone or a second monitor. Don't try to run it on the same PC capturing your footage unless you can isolate the timer from the capture stream. I had a run where I was splitting on a browser tab, the browser hibernated in the background, and my splits were drifting by nearly two seconds over a 45-minute playthrough. I thought my inputs were slow. They weren't. The timer just wasn't running.

If you're timing a game without official RNG manipulation rules, your splits should be based on observable in-game events: crossing a threshold, triggering an NPC dialogue, defeating a boss health bar depletion, or entering a new zone. Games that use save-state warps or sequence breaks require a different approach. You're not just timing yourself anymore. You're verifying that the game state is correct, which means frame-perfect input recording matters more than split precision. The Series X doesn't expose a save-state system like emulators do, so you need to capture at the full frame rate and then scrub through the raw footage to confirm exactly which frame the input landed. Input lag on the Series X varies by game and by your display settings. Enable Game Mode on your TV or monitor. Without it, you're adding anywhere from 15 to 60 milliseconds of post-processing delay before the frame even reaches your eyes. That's enough to miss a frame-perfect input window in games like Celeste or Super Meat Boy, and it shifts your split times consistently in the same direction. I tracked this on a platformer run where my first attempt through a section was 12 seconds and my second was 10.8. Same inputs, same path. The difference was that I hadn't switched my monitor to Game Mode between attempts. I lost a second to processing delay and never knew it. The other thing nobody talks about is storage throttling. The Series X uses a custom NVMe SSD, but the console reserves a chunk of that drive for system overhead and game updates. If your game is installed on the internal drive and you've filled it past about 85 percent, read speeds drop and loading screens stretch out. I noticed this on a run of Metroid Dread where my transition times between rooms were creeping up by half a second per checkpoint over a three-hour session. The game wasn't getting harder. The SSD was thrashing. I freed up about 20 gigabytes and my room-to-room times snapped back to where they were at the start. If you're timing runs that rely on quick transitions, keeping your drive below 80 percent isn't optional. It's a requirement.

Get the Full Details

Skate - Xbox Series X Gameplay + FPS Test - YouTube
Skate - Xbox Series X Gameplay + FPS Test - YouTube

Some games support quick resume from the Series X's suspend state. This is useful for saving time between practice segments, but it's not available in any officially timed run. The console's quick resume feature saves the entire game state to the SSD, which means the game is still allocating resources in the background even when it appears suspended. A few titles, particularly older ones that weren't updated for the Series X, have memory leaks that trigger when you quick resume multiple times in a row. I hit this on a Cyberpunk 2077 run after about eight quick resume cycles in an hour. The game started stuttering on load screens, which inflated my zone transition times. Fresh boot fixed it immediately. If you're doing long practice sessions with quick resume, plan to restart the console every couple hours or you'll be measuring slowdown instead of your own skill. When it comes to sharing or verifying your splits, the standard format is a raw video file with the timer visible in frame. Don't crop the timer out. Don't compress to a social-media format that drops frames. Upload in H.264 or ProRes if you can. The speedrun community checks timing by frame-counting, and a compressed upload can drop intermediate frames, making it look like an input landed on the wrong frame. I've had runs rejected because the YouTube upload had dropped frames around a critical sequence break. The original recording was clean. The upload wasn't. Always share the source file. The bottleneck for most Series X speedrunners isn't the console itself. It's the gap between what the game renders and what your timing tool can actually measure. The hardware can run games at high frame rates, but unless your capture chain, your display settings, and your timer are all aligned to the same frame clock, you're just guessing at your splits. Fix the setup first. The times will follow.