Free, open-source mouse calibration and aim testing for Windows.
Download the Windows ZIP · Release notes and checksums · Roadmap
Extract the entire ZIP, open the TrueTrack folder and run TrueTrack.exe. No build required. Windows x64 and .NET Framework 4.7.2 or later are required. The application is unsigned; Windows may show a reputation warning.
Measure your flicks and precision, compare sensitivity profiles, and check whether changes help without simply taking longer. Designed for standard mice across brands.
Available now: free flick/precision/tracking/target-switching practice, saved routines and comparable practice history, movement replay, optional three-check profile confirmation, optional 3D training room, fixed first-test baseline, explained recommendations, balanced target distances, optional measured circle verification, expanded flick scoring rings, precision-first assessment, four before/after gauges, calibration, A/B testing, repeat confirmation, optional game setup with FOV and preferred cm/360, blind distance estimation, and optional global profiles.
Our long-term goal is a free, open-source aim-training platform shaped by its community. The current release is experimental—not yet a full Aimlabs replacement. Testers and contributors are welcome.
The app supports English (first-launch default) and Norwegian; saved language choices are retained. Calibration and local comparisons do not require installing an input driver. Global profiles use optional drivers with separate setup and compatibility considerations.
The welcome screen previews your crosshair and offers optional customization before you begin. New preferences default to a larger cyan cross with a dark outline and open centre; saved choices are retained.
Circle calibration begins with large centred instructions that fade in and wait for your click. They explain 30 clockwise and 30 counterclockwise turns, natural movement and the direction-change pause. Recording starts after the fade-out and countdown. Progress bars are 32–34 pixels tall and show completed counts and percentage. Flick bullseyes are red; precision targets remain green.
New calibration baselines use a subtle moving dot background without room landmarks. Camera and aim reset between targets. Free practice retains the 3D room. Existing sessions preserve their original background and target geometry for fair comparisons.
Choose Customize crosshair under Calibrate for six shapes, six colors, size, thickness and centre gap. The selection applies to new baselines and free practice; ongoing comparisons retain their saved crosshair. This is an in-app feature, not a global game overlay.
Flick targets now have a wider scoring area and a smaller 100-point bullseye, with 99–1 points outside the bullseye and zero outside the target. Start a new baseline to use them. Progress bars are thicker and calibration instructions fade in and out before measurement starts.
Completed precision blocks also show a separate filtered timing summary for possible distractions: successful hits above twice their median are excluded from that summary only. At least eight hits are required. Raw data, misses, timeouts and all profile acceptance checks remain unchanged.
These app previews illustrate the interface, not measured performance gains. They contain no personal measurement results.
DPI, physical sensitivity (cm/360) and field of view are optional setup steps. Defaults are starting points, not a promise of optimal settings.
Click once per target. Misses count, and the scoring rings distinguish a near miss from a distant one. The moving dot field provides camera-motion feedback without a room.
Each part starts with instructions. In precision trials, correct a miss and try the same target again; the clock keeps running. Then compare the proposal with your baseline and repeat it to check whether an apparent benefit holds up.
Not implemented or available yet. We want to investigate whether repeated, consistent curved deviations in mouse movement can be distinguished from intentional motion and reduced with an optional correction. The intended focus is a recurring personal tendency, not every small variation.
Current movement-path analysis and sensitivity/axis adjustments do not establish that TrueTrack can straighten an arbitrary movement in real time. The proposed feature would need repeated comparisons showing fewer corrective movements without slower hits or worse accuracy, including checks that intentional curves remain usable. Global integration is undecided; Raw Accel support alone does not imply support for this future correction.
This is a research direction, not a promised benefit or release date. See the roadmap.
See the roadmap for the proposed path from calibration to broader aim training: reliable onboarding, tracking and target-switching exercises, practice routines, and local progress history.
Report a bug or suggest an exercise. For bug reports, include the app version, Windows version, mouse model, steps to reproduce and expected versus actual behavior. Review logs and screenshots before sharing; you do not need to upload personal measurement files.
Calibration is designed to work with any standard mouse that Windows exposes as mouse input, including wired and wireless mice. No Corsair, Logitech or other manufacturer-specific hardware is required. Every mouse model has not been tested.
In-app comparisons work locally. Applying a profile across Windows and games requires a compatible input driver, installed through the optional built-in setup. Device support, game input paths and anti-cheat restrictions can affect global operation. The tool does not write calibration to mouse firmware.
- Select English. Start with the defaults or optionally enter your normal DPI, desired cm/360 and horizontal FOV. The defaults (1600 DPI, 40 cm/360, 103 degrees) are starting points, not personalized recommendations. Keep setup unchanged during comparisons. Game setup offers an optional vertical-to-horizontal FOV converter, an in-game sensitivity calculator, and five blind horizontal movements summarized by their median. This estimates preferred physical travel from raw counts and nominal DPI; verify actual cm/360 with a ruler in your game.
- Record the baseline. In flick trials, aim quickly and click once; misses count. In precision trials, keep correcting until you hit; every click and the elapsed time count.
- Inspect the proposed sensitivity and axis changes. Read accuracy, hit rate and time together.
- Run A/B to compare the measured profile with the proposal.
- Choose Confirm the same proposal to repeat that comparison without changing the proposal.
- Only apply a global profile when you choose to. After changing a profile, measure a new calibration phase.
Optional circle calibration records 30 natural rotations in each direction. Move naturally rather than trying to draw perfect circles. It estimates consistent shape tendencies; a rounder simulated circle alone does not establish better aim.
Read the research and game-configuration notes for the evidence behind repeated comparisons and the difference between matching cm/360 and matching FOV.
- Promising: descriptive improvements appeared, but the acceptance checks have not established an improvement.
- Closer but slower: lower aiming error came with increased time.
- Mixed or worse: the proposal has no clear advantage under the checks.
- Supported in repeated checks: the same settings passed consecutive checks. This is not a guarantee of better performance in every game.
Repeated tests retain their earlier results. Different profiles are not pooled as if they were one repeated experiment. Practice and fatigue can affect comparisons across phases.
Requires Windows x64, .NET Framework 4.7.2 or later and Windows PowerShell. Obtain Raw Accel 1.7.1 from https://github.com/RawAccelOfficial/rawaccel/releases and extract it locally. Supply a dependency directory containing its matching wrapper.dll and Newtonsoft.Json.dll, plus a neutral rawaccel-default.json configuration. Keep upstream license notices in that directory.
.\build.ps1 -DependencyDirectory C:\path\to\dependencies
Outputs are in artifacts/TrueTrack. Run TrueTrack.exe. Tests can be run without installing or activating a driver:
$report = Join-Path $PWD 'artifacts\tests.txt'
Start-Process .\artifacts\TrueTrack\VectorFit.Tests.exe -ArgumentList '--selftest', $report -Wait
Get-Content $report
Optional tests of historical local measurements are skipped when those private files are absent. No user measurements are distributed.
Raw Accel integration requires its separately installed driver. Open Use in Windows and games → Set up Raw Accel to install the bundled official driver. If the package is missing, setup downloads the fixed official release. Its SHA-256 is checked before extraction. Installation requires your choice and Windows administrator approval; restart and check status afterwards. No profile is activated by installation. Applying a global profile is an explicit action with backup and readback. Reset to neutral disables adjustments; Restore profile retrieves the last backup. This does not program mouse firmware.
Interception is an optional experimental backend with separate licensing. Optional: Interception setup explains the license and anti-cheat risks, then lets you choose an official download and installation or uninstallation. Its binaries are not included in the repository. Check upstream terms for your intended use, particularly commercial use. Stopping the filter does not uninstall the driver. Check the rules and compatibility of each game's anti-cheat before using any input driver; compatibility is not guaranteed.
Measurements and profiles are stored locally. They may contain personal movement and device information. Do not commit them. This repository excludes private profiles, measurements and reports. The unmodified official Raw Accel installer archive is included with its license; Interception binaries are not included.
Original project code is offered under the MIT license; see LICENSE. Third-party dependencies retain their own licenses; see THIRD_PARTY_NOTICES.md. This project is independent and not affiliated with Corsair, Aimlabs, Logitech, Raw Accel or Interception.
The working name and implementation have not received trademark or patent clearance. MIT licensing does not grant rights owned by unrelated third parties.
From v0.18.0, measurements and profiles are stored in %LOCALAPPDATA%\TrueTrack\profiles. Existing files next to the app are copied there on first launch; originals are retained. Keep the complete old folder until you have checked your history. Conflicting imports are retained in import-conflicts.
Use History / updates → Check for updates. Startup checks are optional. Accepting an update closes the app, verifies the release checksum and replaces program files while retaining your data. Experimental releases are included. No driver installation or profile activation happens during an app update. Older versions need one manual download first. The app and updater remain unsigned.