PC Gaming
Phantom Blade Zero FPS Drops Aren’t Always a Settings Problem
Path tracing can cost you frames, but lowering every setting isn’t always the answer. Here’s what to change first, what to check next, and why average FPS can be misleading.
You know the parry sound cue cold. Two hours in a single fight and your hands have it memorized. Then the arena opens up — lamp light, rain, three enemies closing in — and the same parry, the one you’ve landed a hundred times, starts whiffing.
That’s what frame drops can feel like in Phantom Blade Zero. Not just a lower framerate, but inconsistent frame delivery. When a hitch lands during a parry or dodge, the fight can fall apart even if the FPS counter looks respectable.
You’ve probably already tried the obvious things. Dragged the preset down to Low, got a blotchy, mushy image, and still noticed a hitch when crossing into a new courtyard. Turned on frame generation, watched the counter jump past triple digits, and found that the game didn’t necessarily feel as responsive as the number suggested. Then came forty minutes in the graphics menu, changing settings without knowing which ones actually mattered.
The more useful approach is to work through the changes in a controlled order, using published benchmarks as a starting point rather than a promise. The numbers below come from specific test systems, and the results on your own hardware may differ.
Before any of that, though.
Path tracing is the first thing to test turning off
The game offers three rendering options: rasterization, ray tracing, and path tracing. If path tracing is enabled and your FPS is dropping, turn it off before changing a dozen other settings. It can produce impressive lighting, but the performance cost can be substantial.
In Wccftech’s 4K test on an RTX 5090, native Ultra averaged 67 FPS, ray tracing averaged 52 FPS, and path tracing averaged 33 FPS. In that test, ray tracing reduced performance by roughly 22% compared with rasterization, while path tracing delivered about half the native Ultra frame rate. These figures describe one GPU at one resolution, not a universal performance penalty.
Start with both RT and PT off, then test the game again. If you have performance headroom and prefer the improved lighting, try ray tracing before considering path tracing. There’s no reason to sacrifice stable combat performance for an effect you may barely notice during a hectic fight.
One caveat: the launch patch included RT/PT performance improvements and adjustments to the rasterized image. Benchmarks from different builds may not be directly comparable, so check the game version and test conditions before using older numbers as a baseline.
What’s worth knowing is that Ultra can be demanding even at 1080p. In Wccftech’s test with upscaling and frame generation disabled, the game averaged 122 FPS on an RTX 5090, 86 on an RTX 5080, 75 on an RTX 5070 Ti, 55 on an RTX 5070, and 41 on an RTX 4070.
1080p. Not 4K.
Those results show why an average of 40–70 FPS isn’t automatically evidence of a broken game on a midrange GPU running native Ultra. But if your performance is far below comparable results, check your driver, temperatures, power limits, and background load before assuming the graphics preset explains everything. The aim is not to hit a benchmark number at any cost; it’s to find a configuration that delivers consistent frame times during combat.
Don’t drag the preset to Low. Tune individual settings instead.
The instinct when frames drop is to go straight to Low and work your way back up. But that can sacrifice a lot of visual detail for a relatively small performance gain.
In Wccftech’s comparison, Low and Medium were separated by roughly 3–5 FPS. Moving from Medium to High improved lighting, textures, and environmental detail, but cost performance. High averaged about 85 FPS versus 99 on Medium at 4K on an RTX 5090. The useful preset depends on your hardware and resolution, not on the assumption that the lowest option is always the best fix.
A more targeted approach can recover performance without sacrificing every visual setting. One optimized configuration tested at 1440p with TSR at native resolution and RT/PT off went from 74 to 103 average FPS — about 39%. The 1% low rose from 63 to 87, and the 0.1% low from 62 to 83. That configuration retained Ultra reflections, textures, character models, environmental geometry, and ambient occlusion.
Start by testing these actual graphics options:
- Shadow Quality. Lower it one step and compare performance in the same scene. The gain will depend on your GPU and settings.
- Effects Quality and Lighting Quality. Test these if performance drops during visually busy combat or scenes with complex lighting.
- Environment Detail Quality. Try Medium if you want to reduce environmental detail without lowering every setting.
- Post-Processing Quality. Test it separately, since the visual trade-off may matter more to you than the performance gain.
Then test the upscaler. DLSS, FSR, or XeSS on Quality is a sensible starting point when supported by your hardware. Try Balanced if you still need more performance. Performance mode can help further, but the softer image may make combat details harder to read.
These percentages come from one test system and won’t transfer directly to every PC. The broader lesson is to compare individual settings under the same conditions instead of assuming that a blanket preset reduction is always the most effective solution.
When the slider roulette gets old
Testing settings properly means changing one option, repeating the same scene, and comparing the results under similar conditions. Then doing it again. And again.
That takes time, and it’s easy to mistake a different fight or a different camera angle for a real performance improvement.
Get Willanlima’s stories in your inbox
Join Medium for free to get updates from this writer.
You don’t necessarily need a tool for this. If you enjoy fine-tuning every option, manual testing gives you more control over individual visual compromises and the scenes you care about most.
But if you’d rather spend less time in the graphics menu, free FPS Graphics Pro tool offers smart graphics settings designed to help improve FPS while maintaining visual quality. It provides another option for players who want a more convenient starting point for graphics optimization, rather than manually testing every setting from scratch.
It isn’t a universal fix for every hitch. Driver issues, shader compilation, and other frame-time problems may require separate troubleshooting. That’s the next part worth checking.
Want a simpler starting point for graphics optimization? → Get the free FPS Graphics Pro Tool Here↗️
The stutter that ignores your settings
Not every hitch comes from the same cause, and lowering graphics settings won’t fix all of them.
Shader compilation comes first. The game performs shader precompilation during its initial launch process, and the reported duration varies: one test recorded about eight minutes, while a settings guide reported closer to half an hour. Avoid using this period as your only performance benchmark. Once compilation finishes, shader-related stutters may decrease, but that doesn’t guarantee a completely hitch-free experience.
Traversal stutter is another possibility. Moving around a corner, entering a new area, or loading a different section can coincide with brief frame-time spikes, even with RT and PT disabled. The frequency and severity vary between test systems, so a hitch during movement shouldn’t automatically be blamed on a single graphics setting.
Drivers are another factor worth checking. Around launch, the developers flagged NVIDIA driver 621.01 for unstable performance and recommended driver 617.42. NVIDIA later released hotfix driver 621.12, but its existence alone doesn’t establish that every issue is resolved. Check the current developer guidance and driver release notes before changing versions.
If the stutter began immediately after a driver update, investigating the driver is a sensible early step. Compare performance before and after the change, and avoid assuming that either the GPU or the game is at fault until you’ve ruled out the more obvious causes.
Your average FPS doesn’t tell the whole story
A stable 50 FPS can feel better than a counter bouncing between 40 and 60. Consistent frame delivery matters, especially in a game where parries and dodges depend on timing.
An FPS counter alone won’t show every hitch. Depending on the tool, the displayed number may be smoothed or averaged, so a reasonable-looking reading can hide short frame-time spikes. The 1% low and 0.1% low provide additional context, while a frame-time graph can help reveal brief stalls. In the optimized configuration mentioned earlier, both low-percentile results improved alongside the average, suggesting that the gain wasn’t limited to the headline FPS number.
Frame generation complicates the picture. It can increase displayed FPS by inserting generated frames between traditionally rendered frames, but those frames aren’t equivalent to new game-simulation frames. A higher counter therefore doesn’t automatically mean better responsiveness or more consistent combat timing.
Test with frame generation off first to establish a baseline. Then enable it and compare the visual smoothness and responsiveness for yourself. The benefit depends on the underlying frame rate, the hardware, the frame-generation implementation, and the latency settings in use.
Steam Deck is a separate case because of its limited performance headroom. Some reported tests found that even Low settings with FSR Performance and dynamic resolution could fall below 30 FPS, with lower results in demanding scenes. External frame-generation tools may improve perceived smoothness, but they can introduce additional latency or visual artifacts. Results vary, so evaluate the trade-off carefully rather than assuming that a higher displayed FPS will solve every performance problem.
The order I’d use to troubleshoot FPS drops
- Check the game version and driver. Install the latest game patch, review known driver issues, and change driver versions only when there is a clear reason to do so.
- Turn off path tracing and ray tracing. Keep both disabled while establishing a stable baseline.
- Start at High or Medium. Compare the visual trade-off and performance before considering Low.
- Set the upscaler to Quality. Try Balanced if you still need more performance.
- Tune individual settings. Test Shadow Quality, Effects Quality, Environment Detail Quality, and Post-Processing Quality one at a time.
- Repeat a demanding scene. Use the same crowded area, boss encounter, or weather conditions, and monitor FPS and frame times.
- Check hardware limits. Look at GPU and CPU utilization, temperatures, and power behavior if performance remains lower than expected.
- Test frame generation and a frame cap. Do this only after establishing a baseline, then judge both smoothness and responsiveness.
The important part isn’t how quickly you can change settings. It’s whether you’re comparing the same workload under similar conditions. A quiet corridor may look smooth while a crowded fight still exposes the problem.
Who this is actually for
If you’re running 1440p on High with an upscaler in Quality and the game already feels smooth, you may not need to change anything. Patches and driver updates can also change performance, so there’s little value in chasing a problem you no longer have.
If you’re struggling with inconsistent frame delivery, start with the settings that impose the greatest performance cost, then test the remaining options under the same conditions. If you’d rather spend less time experimenting, FPS Graphics Pro offers a smart-settings option to explore, but it shouldn’t replace troubleshooting when the cause is a driver issue, shader compilation, or another underlying problem.
The official system requirements are also a useful reference point. Their performance targets are tied to specified hardware and graphics configurations, including the use of upscaling and the relevant ray-tracing settings. Treat those targets as a guide rather than a guarantee that every scene will maintain the same frame rate.
The parry window is the thing. Everything else is bookkeeping.
Image Credits & Fair Use Disclaimer: Screenshots and game assets are property of their respective rights holders, used for commentary under fair use.
