Frame-rate tuning becomes complicated when resolution, graphics presets, display modes, power limits, temperature, and overlays all change at once. The simpler workflow is to decide what experience you want, confirm what the display supports, choose one achievable target, and test only the settings that prevent the system from holding it consistently. A stable lower target can feel better than a higher number that repeatedly drops because uneven frame delivery is visible as stutter.
Define the target before changing settings
Start with the game, device, and display you actually use. Record the display refresh rates, current resolution, power mode, graphics preset, frame cap, vertical-sync or variable-refresh setting, and whether the device is plugged in. Pick one target such as 30, 45, 60, 90, or 120 frames per second only when the display and game support it.
Android's power-efficiency guidance explains the central tradeoff: higher refresh rates can improve smoothness, but they also use more power, and a display rate above the game's useful target may provide no benefit. Although the controls differ on a console, computer, or phone, the planning lesson is portable: do not begin with the highest number. Begin with the lowest target that meets the play requirement and that the hardware can sustain.
Measure consistency, not a single peak
Choose one repeatable scene: the same training area, benchmark, race, or short route. Warm the device for a few minutes, then observe average frame rate and visible or measured drops during the same sequence. Keep background downloads, capture software, and power settings unchanged. A peak seen in a quiet menu is not a useful result.
Android's frame-pacing documentation describes how alternating short and long presentation intervals can be perceived as stutter even when a simple average appears acceptable. That is why the check should include consistency. If the platform exposes frame-time graphs, use them; otherwise note whether motion remains even during the demanding part of the route.
Change one tier at a time
Use a short order of operations. First set the target and cap. Second, lower the settings most likely to affect performance, such as resolution scale, ray tracing, shadows, reflections, crowd density, or view distance. Third, repeat the exact scene. Do not change five controls between runs because you will not know which one mattered.
Record each run in one line: date, game version, device, display mode, target, the one setting changed, observed result, and temperature or throttling note. Stop when the target is consistently met with acceptable image quality. Keep the previous good configuration so a driver, game, or operating-system update can be compared without reconstructing the whole experiment.
Separate performance from display behavior
The game producing frames and the display presenting them are related but distinct. Confirm that the operating system and display are using the intended refresh mode. If the platform supports variable refresh, check its supported range and whether the game is running in the required display mode. If it does not, use a supported fixed target and cap rather than assuming a label automatically solves pacing.
Avoid switching targets every few seconds. Android's documentation notes that frame-rate changes can have costs and that supported modes and system conditions affect what is actually selected. For a player, that means the final check must observe the screen and measured behavior, not merely trust the requested setting.
Keep two practical profiles
Most people need only a quality profile and a performance profile. The quality profile can favor resolution or visual effects at a lower stable target. The performance profile can reduce costly settings to hold a higher target for fast play. On a portable device, add a battery profile only if it solves a real need.
Name the profiles clearly and save screenshots of the settings. After a major update, rerun the same scene before accepting new defaults. If the target is no longer stable, restore the last good profile, check power and thermal conditions, and adjust one setting at a time. The result is a small, repeatable workflow: choose, measure, change one variable, verify, and preserve the evidence.
Sources
- Optimize power efficiency — Android Developers
- Frame Pacing library — Android Developers