Quick answer: run the same short route twice after a fresh launch. Keep fullscreen mode, FPS cap, VSync, upscaler, and graphics preset fixed. If pass two is smoother without a settings change, treat shader/cache/streaming warmup as a likely factor before chasing another preset.
The two-pass route test
- Restart the game and close capture overlays, browser video, and heavy downloads.
- Pick a route with one camera turn, one traversal transition, and one combat or effects-heavy moment.
- Turn frame generation, NVIDIA Smooth Motion, or similar interpolation off for diagnosis.
- Pick one fullscreen mode and one FPS cap, then keep them fixed.
- Run the route once and note the first hitch, traversal pause, or camera-turn roughness.
- Run the same route again without changing settings.
- If the second pass is smoother, record it as a warmup/cache signal rather than a solved preset.
How to read the result
| Result | Likely read | Next test |
|---|---|---|
| First pass hitches, second pass is smoother | Shader, cache, or asset streaming may be part of the stutter. | Repeat after a restart and keep the same route. |
| Both passes stutter in the same spot | The base preset, VRAM pressure, CPU spikes, or display pacing may still be weak. | Use the full stuttering fix checklist. |
| Frame generation hides the hitch but input feels odd | Generated frames may be masking a bad base frame-time path. | Disable generation and stabilize the base preset first. |
| Only one upscaler causes traversal hitches | The upscaler interaction may matter more than overall graphics quality. | Use the DLSS / FSR / TSR test. |
Do not overclaim the fix
A smoother second pass does not prove the game is fixed. It means the test route may be warmed. That is still useful because it tells you not to judge every setting change from a cold first traversal.
Useful community proof: "Same route, same cap, same fullscreen mode: first pass hitched at the transition, second pass was smoother." That is better than only posting average FPS.
Copyable no-link reply
I would test this as shader/cache warmup, not just low FPS. Restart, pick one short route, turn frame generation/Smooth Motion off, keep the same fullscreen mode and FPS cap, then run the exact route twice. If pass two is smoother without changing settings, that points more toward warmup/streaming/frame-time behavior than raw GPU shortage.
Related tests
FAQ
Is a second-pass improvement proof of shader compilation?
No. It is only a useful signal. Cache, asset streaming, route warmup, overlays, and frame-time behavior can all make the second pass feel different.
Should I clear shader cache?
Only after recording a clean baseline. Clearing cache can make first-pass stutter worse temporarily, so use it as a deliberate test rather than a first step.
When should I repeat the two-pass test?
Repeat it after a game update, GPU driver change, cache clear, or major graphics preset change. Keep the route and other settings fixed so the comparison remains useful.