Why does my flight simulator feel choppy at high FPS?

A flight simulator can feel choppy at high FPS because the frames are not arriving at even intervals. The FPS counter usually reports an average, while main-thread stalls, scenery streaming, shader compilation, add-ons, VRAM pressure, frame generation, or a display synchronisation mismatch can all produce visible judder and micro-stutters.

This is a general flight-simulator issue affecting Microsoft Flight Simulator 2024 and 2020, X-Plane, Prepar3D, FSX and other simulators. The exact bottleneck varies, but the useful measurement is always frame time: how long each frame takes to reach the display.

What does high FPS but choppy motion mean?

High FPS with choppy motion means frame delivery is inconsistent or does not match the display's refresh cycle. A steady 60 FPS produces a frame every 16.7 ms, while 30 FPS produces one every 33.3 ms. Both can look smooth when the interval remains stable.

An average of 60 FPS can still include a short frame at 10 ms followed by a long one at 70 ms. That long frame appears as a hitch, yet it has little effect on a counter averaged over several seconds. A frame-time graph or 1% low figure exposes this behaviour more clearly than average FPS; our guide to displaying FPS and frame-time information in flight simulators explains the available measurement methods.

Total CPU usage can also be misleading. A simulator may depend heavily on one main thread, leaving the other processor cores partly idle. The system might report 35% CPU usage even though the thread responsible for preparing each frame is saturated.

How do I diagnose uneven frame times?

The quickest diagnosis is to test CPU, GPU, streaming, add-ons and display synchronisation separately under repeatable conditions.

1. Create a fixed test. Use the same aircraft, airport, weather preset, traffic level and camera movement. A busy default airport is more useful than an empty cruise because it shows the frame rate the system must sustain during demanding phases.
2. Measure native performance. Open the simulator's performance display or a platform overlay and watch CPU and GPU frame times. Temporarily disable frame generation and VR motion reprojection so synthetic frames do not disguise the native rendering rate.
3. Establish a stock baseline. Use a default aircraft and scenery with third-party plugins, traffic injectors, weather utilities and external telemetry disabled. In Microsoft Flight Simulator, temporarily remove or rename the Community folder and disable other non-default packages where practical; do not delete add-ons merely to test them.
4. Lower render resolution. A large improvement in GPU frame time indicates a graphics limit. Little or no improvement, after accounting for any active frame cap, points towards the CPU, simulation workload or asset streaming.
5. Compare simple and complex locations. Test a default rural airport against the airport where the problem occurs. A hitch confined to one area usually indicates scenery complexity, overlapping packages or asset loading rather than a general GPU shortage.
6. Test online data separately. Where the simulator allows it, compare online scenery and live services with a local or simplified setup. This separates connection and content-streaming stalls from rendering stalls.
7. Set a sustainable cap. Base the limit on the lowest performance seen during the demanding test, not the highest FPS reached at altitude. Leave enough headroom for scenery loading, avionics and traffic updates.

Patterns matter. A pause at the same place on every approach points towards scenery. A spike every few minutes suggests an autosave, weather refresh or utility. Stutter that occurs only the first time an effect or object appears may be shader compilation; it should diminish once the relevant shaders have been prepared, whereas a persistent repeatable pause has another cause.

Which settings fix choppiness without ruining image quality?

Change the settings attached to the longest frame time instead of lowering every option indiscriminately.

• CPU or main-thread limited: reduce terrain and object detail distance, AI or live traffic, parked aircraft, ground vehicles and airport activity. Test a default aircraft because detailed avionics, displays and system simulations can consume significant CPU time.
• GPU limited: lower render scale or resolution, volumetric clouds, shadows, reflections and ambient occlusion. Upscaling can reduce the GPU rendering load, but it is different from frame generation: upscaling renders fewer pixels, while frame generation inserts synthetic frames.
• VRAM limited: lower texture resolution and render resolution, then remove overlapping high-detail scenery. Exceeding available graphics memory can trigger abrupt asset transfers and pauses even when the average FPS remains respectable.
• Streaming or storage limited: compare online content with simplified scenery and check that simulator caches are on healthy local storage. Do not repeatedly clear caches as a routine performance tweak; rebuild or disable one only when a controlled comparison points to corruption or storage stalls.
• Background task limited: stop duplicate monitoring overlays, recording tools, autosave utilities and telemetry exporters for the test. A utility that wakes at regular intervals can produce an equally regular frame-time spike.

Add-ons can affect frame pacing through scripts, avionics, scenery, traffic and background network requests, not just through higher polygon counts. Use our method for isolating aircraft, scenery and utility add-on performance rather than removing an entire library at random.

Why does frame generation still look choppy?

Frame generation raises the displayed FPS without repairing uneven native frames underneath it. If the simulator pauses while its main thread loads scenery or updates an aircraft system, there is no timely pair of completed frames from which to create smooth intermediate motion.

Disable frame generation while diagnosing the problem and inspect the native frame rate. Re-enable it only after the native frame-time graph is reasonably flat. It can then improve visual fluidity, but it will not remove control latency, streaming hitches or a persistent CPU stall.

Generated frames can also produce artefacts around cockpit frames, propellers, glass displays and rapidly moving labels. An FPS counter may include those synthetic frames, so a reported 80 FPS does not necessarily mean the simulator itself is calculating 80 complete frames per second.

Should I cap FPS or use V-Sync or VRR?

Use a frame cap when it produces a rate the simulator can hold consistently, then match that cap to the display's synchronisation method.

• Fixed 60 Hz display: a stable 60 FPS is ideal, but a locked 30 FPS with V-Sync often looks smoother than movement fluctuating between 40 and 60 FPS. Thirty divides evenly into 60, allowing each frame to be shown for two refresh cycles.
• Fixed 120 Hz display: 60, 40 and 30 FPS all divide evenly into 120 Hz. Choose the highest one the simulator can maintain at demanding airports, not only in cruise.
• Variable-refresh display: enable VRR on the display and platform, then keep performance inside the supported refresh range. A cap a few FPS below the upper boundary can prevent repeated contact with the VRR ceiling. Behaviour below the lower boundary depends on whether the display supports low-frame-rate compensation.
• No synchronisation: disabling synchronisation can reduce latency, but tearing during turns may make otherwise consistent rendering appear choppy.
• VR headset: target either the headset's native cadence or one stable reprojection mode. Continually crossing the reprojection threshold is usually more distracting than remaining at a lower but consistent cadence.

Confirm that the operating system or console is actually outputting the intended refresh rate. Use one frame limiter at a time: stacking the simulator limiter, graphics-driver limiter and an overlay limiter can create pacing conflicts. If several limiters are available, test them separately and keep the one that produces the flattest frame-time graph.

What if only camera pans or head movement look choppy?

Choppy camera movement with stable frame times usually indicates display cadence, tearing or an input update problem rather than insufficient rendering performance. Test a keyboard-controlled pan against mouse look and head tracking; if only one input method stutters, inspect that device's software, smoothing and update settings.

Also distinguish world motion from cockpit animation. Glass displays, gauges, propellers or traffic models may update less often than the rendered world. Lower-rate cockpit animation can look jerky without representing a whole-simulator frame-time problem.

Why does the simulator become choppy only after a long flight?

Choppiness that develops gradually points towards accumulating memory pressure, traffic, weather data, logging or an add-on that is not releasing resources correctly. Compare RAM and VRAM use near departure with the point where performance deteriorates, then repeat the route with external utilities and third-party content disabled.

A restart restoring performance is a useful clue, but not a permanent fix. Our diagnostic process for FPS that degrades during long flights covers memory growth, thermal behaviour, traffic accumulation and add-on testing in more detail.

Are the causes different in each flight simulator?

The frame-time principle is universal, but each simulator has characteristic sources of stutter.

• Microsoft Flight Simulator 2024 and 2020: streamed scenery, dense airports, traffic, glass-cockpit systems and third-party packages can all stall the main thread independently of average FPS. Microsoft Flight Simulator 2024 relies heavily on streamed content, so connection and asset-delivery behaviour deserve a separate test. Use our MSFS-specific high-FPS stutter checks when the developer display identifies MainThread, GPU or memory pressure.
• X-Plane: complex aircraft, scenery loading and plugins are common causes of isolated spikes. Disable plugins by category and compare the same location in a default aircraft before reducing global graphics quality.
• Prepar3D, FSX and FS2004: these older engines often depend heavily on one CPU thread. Running with an unlimited frame rate can consume processor time needed for scenery loading, so a sensible internal cap may feel smoother despite reporting fewer FPS.
• Console versions: developer graphs and graphics-driver controls may be unavailable. For MSFS 2020 on Xbox, or MSFS 2024 on Xbox Series X|S and PlayStation 5 or PS5 Pro, use the same repeatable stock-aircraft test, remove optional third-party content and verify the console and display refresh or VRR settings where supported.

The correct target is the highest frame rate the simulator can sustain with flat frame times during the most demanding part of the flight. A locked 30 FPS at 33.3 ms can look smoother than an average of 55 FPS interrupted by regular stalls; peak FPS matters far less than consistent delivery.