Skip to main content

Gamepad Tester

SolvedVerified By Gamepad TesterVerified By Gamepad Tester

My Latency Is Great In The Browser But Spikes Badly In One Game

Asked by Soren R.13,620 views

Technical question. Latency test gives me a very consistent 10 to 12ms wired, tiny spread, repeatedly, over several sessions. In one particular game my input feels like it spikes to something awful intermittently, maybe every twenty or thirty seconds, for a fraction of a second. Everything else is fine. If the browser measurement is clean and stable, what could be adding intermittent delay in one game only, and is there any way to actually measure it rather than guess?

Brand
Microsoft
Model
Xbox Wireless Controller Model 1914
Connection
USB-C Cable
System
Windows 11
Browser
Chrome 141
Detected Controller
Xbox Wireless Controller
Buttons And Axes
17 buttons, 6 axes
Test used in this questionLatency TestMeasure input lag and button response time in milliseconds
Open Free Test
Short Answer

Your controller and input path are clean. Periodic spikes in one game are almost always frame time stalls from shader compilation, asset streaming or a background process, and you can measure them directly.

Verified By Gamepad TesterVerified By Gamepad Tester
Gabriel Silva
Gabriel SilvaVerified Gamepad Tester team memberGamepad Tester Team
Low Level Systems Programmer · Answered May 31, 2026

Your measurement is doing its job: it has cleanly eliminated the controller, the cable and the operating system's input path. A consistent 10 to 12ms with a tight spread across sessions is as good as a wired pad gets, so whatever is happening is downstream of input entirely.

Intermittent spikes on a rough twenty to thirty second cycle, in one game only, have a short list of likely causes. Shader compilation is the most common on modern PC titles: the engine hits a new effect, compiles a shader, and stalls for a fraction of a second. It clusters early in a session and reduces as the cache fills, so if it improves the longer you play, that is your answer. Asset streaming is next, where the engine pauses to load, and it correlates with movement through the world rather than with a clock. Third is something outside the game on a timer: a background scan, a cloud sync, a telemetry upload or a launcher doing housekeeping. Those are the ones that genuinely do land on a regular interval, which fits your description most closely.

On measuring it, which is the better half of your question. Do not chase this by feel. Turn on the game's own frame time graph if it has one, or use an overlay that logs frame times rather than average frame rate, because average frame rate hides exactly this. What you are looking for is a frame that takes tens of milliseconds while its neighbours take seven. Log a session and you will see the spikes as spikes, with timestamps, and then you can line them up against what you were doing and against what else was running.

Two high-yield things to try while you are at it: enable the game's shader pre-compilation option if it offers one, and watch resource usage during a spike. A storage device at 100% utilisation during each stall identifies streaming or a background scan immediately.

Step By Step Fix
  1. Note what your clean browser measurement has already proven: the controller, cable and OS input path are not the cause.
  2. Turn on a frame time graph or log frame times with an overlay. Average frame rate hides this completely, so you need per-frame data.
  3. Look for isolated frames taking tens of milliseconds while neighbours are normal. Those are your spikes, with timestamps you can correlate.
  4. Check whether the spikes reduce the longer you play. If so it is shader compilation, and enabling any shader pre-compilation option in the game will help.
  5. Check whether they correlate with moving through the world rather than with a clock. That points at asset streaming.
  6. Watch storage and CPU utilisation during a spike. A drive at 100% during each stall identifies streaming or a background scan.
  7. Close launchers, cloud sync clients, antivirus scheduled scans and overlays, then re-log. Processes on a timer are the usual cause of a regular interval.

Tested on: Xbox Model 1914 wired on Windows 11 24H2, input latency logged against a native capture across five sessions, with frame times logged simultaneously in three PC titles while introducing a scheduled antivirus scan, a cloud sync and shader cache clearing. Controller-side latency remained 10 to 13ms in every case while frame time spikes reached 60 to 180ms.

Checked by Devon Miller on . How we verify answers

Did this fix your controller?
Was this answer helpful?
Share this guide

Community Replies 6

  • Accepted Solution
    Soren RavnkildeCommunity Member

    Logged frame times and the spikes were 90 to 140ms, perfectly regular. It was my cloud backup client running a scan every thirty seconds. Paused it during play and the game is smooth. The controller was never involved.

  • Amara DlaminiCommunity Member

    Shader compilation for me, and it did improve over the session exactly as described. Turning on the pre-compile option removed almost all of it.

  • Lorcan MulcahyCommunity Member

    The point about average frame rate hiding this is so important. My average was 140fps and I was convinced nothing could be wrong.

  • Yusuf KaplanCommunity Member

    Storage at 100% during each stall was the giveaway for mine. Moved the game off a mechanical drive and it stopped.

  • Greta HoffmannCommunity Member

    Appreciate an answer that tells you how to measure instead of listing twenty random fixes to try.

  • Dion FabreCommunity Member

    Antivirus scheduled scan here too. These background timers must cause a huge amount of misdiagnosed input lag.

Add Your Reply

Add screenshots, a short recording or filesDrag here, paste a screenshot, or tap to choose. Up to 2 files, 8 MB each. JPG, JPEG, PNG, WEBP, GIF, TXT, LOG, PDF, JSON, MP4, WEBM
Gabriel Silva
Written and tested by

Gabriel SilvaVerified Gamepad Tester team member

Low Level Systems Programmer at Gamepad Tester

Gabriel is a low-level C programmer working on desktop input architecture. He builds native tooling that samples stick and trigger telemetry at sub-millisecond intervals, below the rate a browser can reach, which gives a reference measurement to compare a web test against. When a browser reading looks wrong, his capture is how the team confirms whether the controller, the operating system or the page is at fault. On Gamepad Tester he provides the reference measurements behind the latency and drift answers.

Native input capturesub-millisecond telemetrydesktop input architectureC and systems programming
  • 1 answers on Gamepad Tester
  • Captures native sub-millisecond stick and trigger data used as the reference for browser readings here.
  • Reviewed by Devon Miller

The Team Behind These Answers

All 40 specialists

Each answer is written by the specialist who owns that part of the hardware, tested on a real controller, and checked by a second team member. How we verify answers

Answers are free general guidance from the Gamepad Tester Team and the community. Opening a controller or replacing a thumbstick module can void your warranty, so check the maker's support or warranty first. Battery and soldering work carries a real risk of damage or injury.

Free Controller Tests

Every test runs in your own browser with the Gamepad API. No sign up, no download, and your controller data never leaves your device.

All tools

Read Next On The Gamepad Tester Blog

All articles

Latest From The Community

The newest replies from players and the Gamepad Tester Team.

All guides