Low FPS, Stutter, Input Lag, or Tearing? A One-Change Test Plan

Start by naming the symptom, not guessing the cause. Low FPS looks consistently slow or heavy; stutter interrupts otherwise acceptable motion; input lag separates your action from the visible response; tearing appears as mismatched portions of an image. Choose a repeatable in-game location, record the display mode and relevant settings, then change one item at a time. Restart when the setting requires it and compare the same scene before keeping anything.

These categories can overlap. NVIDIA distinguishes graphics lag, in which insufficient frame rate also delays visible control response, from stutter lag, which appears as momentary freezes. It also recommends paying attention to minimum FPS rather than judging a game only by its usual or highest rate, because demanding moments can still feel poor. That makes the lowest and roughest part of your repeatable run more useful than an easy scene that produces an impressive number. Treat your first classification as a working hypothesis, not a verdict.

Prepare the PC before changing anything

Pick a short test that you can reproduce without relying on a random match. A saved checkpoint, training area, benchmark sequence, or familiar route can work. Keep the camera path and actions similar. The goal is not laboratory precision; it is a fair before-and-after comparison.

• Write down the game, test location, approximate route and actions you will repeat.
• Record whether the game is fullscreen, borderless-windowed or windowed.
• Record the monitor selected in Windows and the refresh rate currently shown for it.
• Note any active frame cap, in-game synchronization option and GPU control-panel override without changing them yet.
• Close optional overlays or background tasks only if you will keep them in the same state for every comparison.

Classify the dominant symptom

Run the baseline once and choose the description that best matches the worst moment. If two descriptions fit, investigate the more obvious one first and list the other as a secondary symptom.

• Low-FPS candidate: movement remains heavy or uneven through much of the route, rather than failing in one isolated instant.
• Stutter candidate: motion is acceptable between distinct pauses, hitches or sudden jumps. Note exactly where each interruption occurs.
• Input-lag candidate: camera motion looks reasonably smooth, but a button press, mouse movement or stick movement seems to appear late.
• Tearing or sync candidate: different horizontal portions of the image appear out of alignment during movement, or responsiveness changes noticeably when an existing sync option is toggled.

Create a repeatable baseline

Repeat the route at least once before touching a setting. If the game has an FPS or frame-time display, record what it shows during the worst section, but also write a plain-language observation such as “continuous heaviness,” “one pause at the doorway” or “turn begins after mouse movement.” A number without a location and symptom is difficult to compare later.

If consecutive baseline runs differ dramatically, postpone configuration changes and try to make the test more repeatable. A setting cannot be judged fairly when the route, scene or display mode keeps changing.

Choose one symptom-specific change

For a low-FPS candidate, select one existing graphics option and move it down by one available step. For a stutter candidate, keep settings unchanged for another run and check whether the hitch recurs at the same place or action before choosing a variable. For suspected input lag or tearing, test one presentation or synchronization option while leaving graphics quality alone.

Windows 11 offers optimizations for compatible DirectX 10 and DirectX 11 games running windowed or borderless-windowed. Microsoft says the feature changes compatible games from legacy blt presentation to flip-model presentation to reduce frame latency. For a per-game test, open Settings, System, Display and Graphics, select the game, change its windowed-game optimization preference, and restart the game. Test this only when the recorded display mode and game fit that scope; do not combine it with a sync or quality change.

On a system with compatible AMD graphics and a certified FreeSync display, AMD Software can show the selected monitor’s FreeSync status, certification tier and current refresh range. It also offers per-game choices of AMD Optimized, On or Off. Use those as separate comparison states rather than assumptions about which must feel best. AMD also notes that some monitor presets or custom display settings can disable FreeSync support, so record the monitor preset and consult the display maker’s recommended configuration if support unexpectedly appears unavailable.

• Change exactly one setting and write its original value before applying the new value.
• Restart the game whenever Windows or the setting itself requests a restart.
• Repeat the same route, camera movement and control action in the same display mode.
• Keep the change only if the target symptom improves without introducing a drawback you consider worse.

Worked example: a hitch mistaken for low FPS

Suppose a borderless-windowed game feels bad while entering a busy courtyard. The first log entry says “low FPS,” but a repeated run shows smooth motion until one sharp pause at the same doorway. Reclassify the primary symptom as stutter and record the doorway as the trigger location. Do not lower several quality options at once.

As a suggested next test, keep the route and quality settings unchanged, alter only the game’s per-game windowed optimization preference, restart the game, and repeat the doorway run. If the pause is unchanged, revert the preference and mark the result neutral. The next experiment should start from the restored baseline. If the pause changes but controls or image presentation feel worse elsewhere, record both outcomes rather than calling the test an uncomplicated success.

Use a copyable test log and a stop rule

Create one entry per run. Short, specific notes are more useful than a long verdict written after several undocumented changes.

• Game and version shown:; test date and time:; test location or checkpoint:; route and repeated actions:.
• Display mode:; selected monitor:; refresh rate shown:; frame cap:; sync state:; monitor preset:.
• Primary symptom:; secondary symptom:; exact moment or location:; FPS or frame-time observation if available:.
• Single setting changed:; old value:; new value:; restart performed: yes or no.
• Result for target symptom: better, unchanged or worse; new drawback:; decision: keep, revert or retest.