Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallThere is no universal best-settings preset for every game and PC. The reliable way to improve FPS without making the image harder to read is to identify whether the game is limited by the GPU or CPU, test one change at a time in a repeatable scene, and judge frame-time behavior and moving-image clarity alongside the FPS counter.
[Game] is still a placeholder, so exact menu labels, visibility-critical options, patch behavior and support for DLSS or frame generation cannot be named responsibly. Use the method below now, then apply game-specific values only after the title and its current graphics documentation are known.
As an Amazon Associate I earn from qualifying purchases.
What “best settings” can mean when the game is not identified
A setting is useful only if it addresses the limit that is actually holding back performance. A faster graphics card can raise frame rate in a GPU-bound situation, but it may produce little or no improvement when the CPU, game engine, simulation or an imposed frame cap is the limit. The same slider can also have a very different cost and visual effect from one game to another.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →That is why a responsible guide for an unnamed game cannot provide a universal table such as “set everything to High” or “turn every option to Low.” Instead, use a controlled process:
#1 Best Overall
- Powered by Radeon RX 9070 XT
- WINDFORCE Cooling System
- Hawk Fan
- Server-grade Thermal Conductive Gel
- RGB Lighting
- Choose a repeatable scene that represents the demanding parts of normal play.
- Record average FPS and frame-time behavior before changing anything.
- Determine whether the GPU or CPU is limiting the run.
- Change one relevant option, rerun the same scene and inspect the moving image.
- Keep a change only when its performance benefit is worth its loss in detail, stability or visibility.
Exact advice still depends on the eventual game name, version and renderer. Those details determine whether a title exposes render scale, a temporal upscaler, DLSS, frame generation, ray tracing, a visibility-related shadow option or none of them.
Diagnose the bottleneck before lowering settings
Signs of a GPU-bound run
A run is likely GPU-bound when the graphics processor is working near its practical limit, lowering a GPU-heavy option produces a measurable FPS or frame-time improvement, and the CPU has spare capacity. Resolution, render scale, ray-traced effects, volumetric effects, reflections and some shadow or ambient-lighting implementations are common test candidates, but the exact cost is title-specific.
Signs of a CPU-bound run
A run is likely CPU-bound when lowering resolution or expensive visual options barely changes FPS, one or more CPU threads are saturated, or the game is simulating many characters, physics objects or world systems. A faster GPU does not solve that limit. Options that reduce simulation, crowd density, view distance, foliage updates or other engine work may matter if the game exposes them; the correct control varies by title.
Free tools Windows power users keep installed
One-click scans. No signup required.
Use frame time, not just the headline FPS
FPS is an average rate. Frame time is the duration of each rendered frame, so spikes can feel like hitching even when the average number looks healthy. Record the average FPS and a percentile or graph that reveals slow frames. AMD’s Radeon Game Advisor documentation, for example, describes reporting average FPS and the 95th-percentile frame time, then using those results to suggest an in-game change and rerunning the report. Treat that product behavior as an example of an iterative method, not proof that the tool supports every PC or game.
| Observation during the same scene | Most useful interpretation | First test |
|---|---|---|
| Lowering resolution or render scale raises FPS clearly | The GPU is probably a significant limit | Compare a modest render-scale reduction or an upscaler mode, then inspect clarity in motion |
| GPU-heavy changes barely move FPS while a CPU thread is busy | The run may be CPU- or engine-limited | Test CPU-oriented distance, crowd, simulation or effects controls if the game provides them |
| Average FPS is acceptable but frame-time spikes remain | Stutter or streaming may matter more than average rate | Repeat the route, watch the frame-time graph and test the setting associated with the spike |
| Both CPU and GPU have headroom but FPS is fixed | A frame cap, synchronization mode, menu limit or engine ceiling may be active | Check the game’s cap, display mode and synchronization settings before lowering image quality |
Build a repeatable test scene
- Pick a demanding route. Use the same location, camera movement and combat or traversal sequence each time. A quiet menu or empty corridor can hide the problem you are trying to fix.
- Stabilize the conditions. Keep the display resolution, refresh rate, power mode, background applications and game version unchanged between runs. If the game streams assets, allow the same area to finish loading before comparing.
- Run a baseline. Note the current preset, resolution, render scale, upscaler state, frame cap and synchronization mode. Record average FPS and frame-time behavior for long enough to include the difficult part of the route.
- Change one variable. Do not lower a preset, resolution and several effects at once. A single-variable test tells you what actually helped and what visual cost it introduced.
- Compare in motion. Pan the camera, track distant objects and observe moving targets. A still screenshot can conceal shimmer, ghosting, disocclusion trails or loss of small scene cues.
- Keep a simple log. Write down the setting, before-and-after FPS, frame-time behavior and a short clarity note. Revert changes that do not help the diagnosed limit.
Use the same test for every candidate. Changing the route between runs makes a setting appear better or worse simply because the workload changed.
GPU-bound tuning: reduce cost while protecting the image
Test resolution and render scale deliberately
Resolution and internal render scale can be high-impact GPU tests, but they are not guaranteed wins. Lowering the internal image size can raise performance while leaving interface elements at native resolution in some implementations. The trade-off is reduced fine detail, softer edges or reconstruction artifacts. Compare distant targets, thin geometry and motion rather than judging only a static menu.
Rank #2
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Powered by GeForce RTX 5070 Ti
- Integrated with 16GB GDDR7 256bit memory interface
- PCIe 5.0
- WINDFORCE cooling system
If the game offers separate display resolution and render scale, test them independently. A lower internal scale with a native-resolution interface may preserve readability better than lowering the entire desktop output, but that behavior is implementation-specific.
Use an upscaler when the title supports it
DLSS Super Resolution reconstructs a higher-resolution frame from a lower-resolution render. It is not the same operation as simply changing the monitor’s output resolution, and its quality depends on the game’s integration and selected mode. Availability and menu names vary by game and by implementation.
Compare the native render and each available quality mode at the same target resolution. Look for temporal shimmer, ghost trails behind moving objects, foliage breakup and loss of small silhouettes. Keep the mode that improves frame time without obscuring the cues you need to play. Do not assume a lower-quality mode is automatically best; the useful choice is the one that meets your performance target while retaining legibility.
Test expensive lighting and reflections one at a time
Ray-traced lighting, reflections, global illumination, volumetrics and high-quality shadows can be significant GPU workloads. Disable or reduce one effect, repeat the route and then inspect the same surfaces and moving objects. Some effects are visually dramatic but offer little gameplay value; others may help you read a scene. Because option names and costs differ, do not copy a value from another title.
Do not automatically lower textures
Texture quality often depends more on available video memory than on raw shading time. Lowering it can make surfaces and object markings less distinct without delivering a meaningful FPS gain. NVIDIA’s 2019 Control guide found several texture levels had essentially similar performance in its test, while Ultra carried a small measured cost. That is evidence that each option needs to be measured in the game being tuned, not a reason to assume the same result elsewhere.
Recommended Free Tools
CPU-bound tuning: lower workload, not just pixels
When the CPU or game engine is the limit, dropping resolution may leave FPS almost unchanged. Look for controls that reduce the amount of work the simulation must perform:
Rank #3
- Powered by the NVIDIA Blackwell architecture and DLSS 4. System Requirements: Minimum 850W PSU with 16-pin 12V-2x6 (12VHPWR) connector required. Verify before purchasing.
- Military-grade components deliver rock-solid power and longer lifespan for ultimate durability. Compatibility: 348mm (13.7") length, 3.6 slots, 4.3 lbs. Confirm case clearance and slot spacing. GPU bracket included.
- Protective PCB coating helps protect against short circuits caused by moisture, dust, or debris
- 3.6-slot design with massive fin array optimized for airflow from three Axial-tech fans
- Phase-change GPU thermal pad helps ensure optimal thermal performance and longevity, outlasting traditional thermal paste for graphics cards under heavy loads
- World and view distance: reducing how far objects are updated or drawn can lower CPU and GPU work, but may remove useful distant cues.
- Crowd, traffic or population density: fewer active agents can improve busy scenes while changing how populated the world feels.
- Physics, destruction or simulation detail: a lower setting may reduce CPU work in scenes with many interacting objects.
- Foliage and object update distance: these can affect both draw calls and the visibility of environmental markers.
- Background limits: close launchers, browsers, overlays or recording tasks that compete for CPU time before changing image quality.
Use the same one-change test. If a CPU-oriented option improves the difficult scene but harms navigation or target recognition, restore it and try a less visible workload reduction. A CPU limit can also be a hard engine ceiling; in that case, lowering every visual option will not create unlimited FPS.
Preserve visibility while tuning
Separate competitive visibility from cosmetic detail
The setting that makes a screenshot attractive is not necessarily the one that helps you identify a moving target. Keep detail that communicates position, silhouette, contrast, object state or route information. Reduce effects that add noise, blur or expensive lighting without improving those cues, provided the game lets you do so.
Evaluate motion and peripheral vision
Check fast camera turns, strafing and scenes with particle effects. Some temporal reconstruction methods look clean while stationary but produce trails or shimmer in motion. Test dark corners and bright outdoor areas as well; a change that looks acceptable in one lighting condition may hide objects in another.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteKeep a visibility checkpoint
- Can you distinguish a distant character or object from its background?
- Do thin rails, wires, foliage edges or other navigation cues shimmer or disappear?
- Do shadows or ambient effects obscure a target rather than define its shape?
- Does motion remain readable during a rapid pan?
- Are interface elements still crisp and correctly scaled?
Which slider controls each cue depends on the game. The missing title must be identified before anyone can name a visibility-critical setting with confidence.
DLSS, frame generation and latency are different decisions
Super Resolution changes the rendered image workload
DLSS Super Resolution renders internally at a lower resolution and reconstructs a higher-resolution output. It can be a useful GPU-bound test, but the quality and artifact pattern are game-specific.
Frame Generation creates additional displayed frames
Frame Generation creates extra frames between traditionally rendered frames. A higher displayed FPS number does not automatically mean lower input latency or a smoother simulation. The underlying rendered frames, CPU/GPU bottleneck and queue behavior still matter. Availability varies by title, hardware generation and implementation.
Rank #4
- AI Performance: 767 AI TOPS
- OC mode: 2632 MHz (OC mode)/ 2602 MHz (Default mode)
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Axial-tech fan design features a smaller fan hub that facilitates longer blades and a barrier ring that increases downward air pressure
- A 2.5-slot design maximizes compatibility and cooling efficiency for superior performance in small chassis
Ray Reconstruction and DLAA are separate options
NVIDIA’s driver documentation distinguishes Ray Reconstruction, which is intended to improve reconstructed ray-traced detail, from DLAA, which applies anti-aliasing at native resolution rather than reducing the internal render resolution. A game may expose one, both or neither, and may use different labels.
Do not combine competing motion technologies
NVIDIA states that “Native DLSS Frame Generation and Smooth Motion are competing technologies and should not be used together. Doing so results in lower performance and visual artifacts.” Apply that warning only when both controls are present in the game or driver. Read the current game’s and driver’s labels carefully; similarly named features are not interchangeable.
Consider latency separately from FPS
Render queues and CPU/GPU bottlenecks affect responsiveness. NVIDIA Reflex is designed to manage the game’s CPU/GPU work and render queue. If a title exposes Reflex, compare responsiveness during the same test rather than assuming the highest displayed FPS will feel best. A stable frame-time pattern and sensible queue behavior can matter more than a large counter increase.
Using Radeon Game Advisor as an example workflow
AMD’s documentation describes Game Advisor as a monitor-and-rerun process: observe gameplay, review average FPS and 95th-percentile frame time, apply a suggested in-game change, and run the report again. Suggested categories can include lowering a quality preset, advanced rendering options, render scale or display resolution.
The documented compatibility qualifications matter. The tool requires exclusive full-screen operation, describes support for Radeon RX 400-series and newer graphics, and excludes APU graphics. Software availability and support can change, so verify the current documentation before relying on it. If the tool is unavailable, the manual workflow in this guide still works with the game’s own performance display or another compatible monitor.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →When a graphics-card upgrade is justified
Consider a graphics-card upgrade only after your repeatable test shows a clear GPU limit and reasonable settings changes cannot meet the target. A faster GPU is a plausible FPS remedy in that case; it is not a guaranteed fix for a CPU-bound game, an engine ceiling or a frame cap.
Best Value
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Powered by GeForce RTX 5060
- Integrated with 8GB GDDR7 128bit memory interface
- PCIe 5.0
- WINDFORCE cooling system
Before buying hardware, record the game resolution, render scale, upscaler mode, frame cap, GPU utilization, CPU behavior and frame-time pattern. Confirm that the rest of the system, power supply, cooling and display can support the change. No particular card, budget or performance tier can be recommended while the game and your hardware remain unspecified.
Game-specific evidence: what the Control example does and does not prove
NVIDIA’s 2019 guide for Control reported DLSS frame-rate increases of as much as 89% in its higher-fidelity mode and up to 197% in lower-resolution performance modes. Those are NVIDIA-published, game-specific results under that guide’s test conditions. They are not typical gains promised for another title, another patch or another PC.
The same guide stated that, at 1920×1080, performance became CPU-bottlenecked at around 80 FPS in its tested configuration. That observation explains why further GPU reductions can stop helping once a CPU limit is reached; it is not a universal CPU ceiling at 1080p. Use the example as a reason to measure your own bottleneck, not as a target value for [Game].
Troubleshooting branches when a change does not help
“I lowered resolution and gained almost nothing.”
Recheck GPU and CPU utilization, frame-time behavior and any frame cap. The run may be CPU-bound, engine-limited or synchronized to the display. Test a CPU-oriented option or remove the cap temporarily for diagnosis, then restore the cap you actually want.
“The counter rose, but the game looks worse in motion.”
Revert the most damaging change and test a different GPU workload. Compare upscaler modes, render scale and individual lighting effects while moving. A smaller gain with stable edges and readable targets is often the better setting.
“The image is sharp, but frame pacing is poor.”
Use the frame-time graph rather than average FPS alone. Repeat the same route to identify streaming or simulation spikes. Check background CPU activity, synchronization and queue-related options, then test the setting most closely associated with the spike.
“Frame Generation reports a high FPS, but input feels delayed.”
Separate displayed frames from traditionally rendered frames and latency. Check the game’s latency options, the underlying GPU/CPU limit and frame-time consistency. Do not combine native DLSS Frame Generation with Smooth Motion when both are exposed, because NVIDIA warns of lower performance and visual artifacts.
“A recommended tool does not monitor my setup.”
Check its documented requirements before troubleshooting further. For Radeon Game Advisor, exclusive full-screen operation, Radeon RX 400-series-or-newer support and the exclusion of APU graphics are material qualifications. Use a manual baseline and one-change loop when the tool is not compatible.
A practical order of operations
- Confirm the exact game, patch and graphics API, then note every available upscaler, frame-generation and latency option.
- Set a sensible baseline and record average FPS plus frame-time behavior in a repeatable demanding scene.
- Determine whether the GPU, CPU, engine, cap or synchronization mode is the limiting factor.
- If GPU-bound, test render scale or an upscaler, then expensive lighting, reflections, volumetrics and shadows one at a time.
- If CPU-bound, test distance, crowd, simulation and update-work options instead of assuming lower resolution will help.
- After every change, inspect moving targets, distant detail, dark scenes, bright scenes and interface sharpness.
- Evaluate frame generation and latency separately from the displayed FPS number; never combine competing motion technologies.
- Keep only changes that improve the diagnosed limit without removing the visual cues you need.
- Consider a GPU upgrade only when your measurements establish a GPU bottleneck and settings changes are insufficient.
Once [Game] is replaced with the actual title, this process can be turned into a precise menu-by-menu guide. Until then, any fixed preset or promised percentage would be less reliable than measuring the bottleneck on your own PC.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
