Three options cost the most frame rate in Dying Light: The Beast: ray-traced reflections (about 20%), ray-traced global illumination (about 20%) and the Draw Distance Multiplier slider (up to about 20%). Techland’s own advice for switching ray tracing on is to pick the Ultra preset, and its Update 1.4 notes add that Ultra also turns frame generation on automatically. If you picked Ultra on an RTX 3060 and watched the frame rate collapse, that preset is the first suspect.
Those percentages come from one RTX 4090 test by PC Optimized Settings, so treat them as a ranking of what hurts, not a promise for your card. The same test says fog and post-processing “don’t appreciably impact” performance, which is the opposite of what most settings guides claim. I could not find a single published benchmark of an RTX 3060 in this game, so the 3060 config below is built from measured costs on other cards and labelled as a starting point, not a result I have clocked myself.
Verified 14 October 2026 against Techland’s Update 1.7 notes (17 September 2026); the game launched on 18 September 2025. The cost figures date from launch week to the Update 1.4 ray-tracing release; where a source is older than the latest patch, the section says so.
Quick start: the 60-second setup for an RTX 3060-class card
Start from the High preset, not Ultra, then change these seven things in order. Every value is explained further down.
- Preset: High (keeps ray tracing and automatic frame generation off).
- Upscaler: set it after the preset, because upscaling quality is tied to the preset by default. DLSS Quality at 1080p, Dynamic Resolution Scaling off.
- Frame Generation: off.
- Renderer: D3D12 with Asynchronous Compute on.
- Draw Distance Multiplier: 140 as a starting point.
- Shadows Quality: Medium, Screen Space Shadows on, Ambient Occlusion Medium.
- Advanced menu, ray tracing: confirm every ray-tracing toggle is off.
What each setting actually costs in FPS
Ray tracing and the Draw Distance Multiplier are the only options that cost double-digit percentages; everything else in the menu is worth single digits or less. The table pulls every figure from the PC Optimized RTX 4090 test unless another source is named, and it says what you give up for each.
| Setting | Measured FPS effect | What you lose | Call for an RTX 3060 |
|---|---|---|---|
| RT reflections | about 20% lower | Reflections on diffuse and semi-glossy surfaces only; mirror-like surfaces stay screen-space either way | Off |
| RT global illumination | about 20% lower | Biggest visible gain outdoors, on surfaces out of direct sun | Off |
| RT shadows | a few percent lower | Sunlight only. Looks worse than the normal shadow maps in DSOGaming’s test | Off |
| Draw Distance Multiplier | up to about 20% across the slider | Distant geometry and shaders; mostly visible from vantage points | 140 |
| Shadows Quality | Medium is about 5% faster than High; Low up to 10% | Softer shadow edges; artefacts in artificial-light shadows at Low | Medium |
| LOD Quality | Low is about 3% faster than High | Subtle extra pop-in | High |
| Asynchronous Compute | up to 5% faster when on (RTX 4090) | Nothing | On |
| Bloom, lens flare, motion blur | about 1% each | Taste only | Off if you dislike them |
| Ambient Occlusion, Global Illumination, Screen Space Shadows, Reflection Quality | minimal to negligible | AO High tends to overshade | AO Medium, GI High, SSS on |
| Fog Quality and Postprocess | “don’t appreciably impact” | Nothing measurable | Medium |
Two measurements frame the rest. First, the presets are blunt: at native 4K the Very Low preset is only 10-12% faster than High, and maxing everything out costs another 13%. Second, upscaling dwarfs every individual setting: Quality, Balanced and Performance gained 43%, 55% and 67% at 4K. Working that out per step, each notch down the upscaler is worth only about 8% more (1.55 ÷ 1.43 and 1.67 ÷ 1.55), which is my arithmetic from their numbers.
The Ultra preset trap: what ray tracing costs and what it buys
In both tests that measured it, ray tracing roughly halved the frame rate or worse, so it is not a free upgrade you forgot to enable. Techland says the feature needs a ray-tracing-capable card with at least 8 GB of video memory (16 GB recommended for 4K), and warns it “may impact performance”.
The size of the hit depends on the scene. DSOGaming measured 87 FPS minimum and 91 average at native 4K with an RTX 5090 in what it calls the most demanding area it could find, then “the 30s” with ray tracing on. GameGPU’s Patch 1.4 test logged 30 minimum and 38 average on the same card at 4K Ultra RT, and 20 / 27 on an RTX 4090. The two tests use different scenes, so do not subtract one from the other. They agree on the size of the cliff, and neither tested a card in the RTX 3060 class.
The per-effect split is the useful part. If reflections and global illumination each cost about 20% and the two simply multiply, the pair costs roughly 36% (0.8 × 0.8 = 0.64). That is my back-of-envelope figure; nobody I found measured the pair on its own. It is close to the “about a third” most people quote, but the culprits are reflections and GI, not fog or shadows.
Here is how I would use that, as a reading of the numbers rather than a tested rule:
- RTX 3060, 4060 or weaker: leave every ray-tracing toggle off. PC Optimized says mid-range and low-end cards need frame generation to hold a stable frame rate with ray tracing on, and Update 1.7 only just fixed screen tearing with ray tracing plus frame generation on RTX 40 cards.
- RTX 4070 class or better, 1440p with upscaling: if you want one effect, make it global illumination. PC Optimized found it gives outdoor shaded surfaces the biggest lighting improvement.
- Any card: skip ray-traced shadows. They cost little, but DSOGaming, the one review I found that compared them, said the ray-traced shadows are “not as good as the normal shadow maps”: tree shadows “look less detailed, and they animate at a much slower update rate”.
Fog: the option people blame
Fog Quality is a cheap setting in the only measured test I found. PC Optimized says it and post-processing “don’t appreciably impact” performance, with one honest caveat: fog can matter more in heavily foggy areas, and the author had not met one yet. That caveat is why I give a Medium recommendation instead of “turn it off”.
Settings sites that claim big fog savings are usually modelling, not measuring. CheckFPS’s page lists “Volumetric Fog +14% FPS”, “Volumetric Clouds” and “Crowd Density”, and says its numbers come from “our FPS model”. Neither of the two full menu listings I found (Deltia’s and Sportskeeda’s) contains any of those three options. The game’s menu has Fog Quality, Particles Quality and Postprocess Quality, and nothing named volumetric.
Draw Distance Multiplier: the cost hiding in a slider
Draw Distance Multiplier is the most expensive non-ray-tracing option in the game, cutting frame rate by up to about 20% across its range, and its extra distance rarely shows. PC Optimized’s measurement says the higher values “don’t impact detail unless viewing from a vantage point”, and it calls 150-200 the sweet spot. Guides disagree on the exact number: Deltia’s uses 140 and Sportskeeda’s list a “LOD Range Multiplier” of 140, so start at 140 and move in steps of 20 or so until your frame-time graph stops complaining.
There is a CPU angle. PC Optimized saw “a mild CPU bottleneck” in upscaled 1080p and 1440p, “particularly when parkouring across rooftops”, and recommends lowering draw distance when frame rate drops often. Upscaling shrinks the GPU work, so settings that load geometry rather than pixels take a bigger share of the frame. That reasoning is mine, not a measurement, but it fits why the slider matters more on a 3060 with DLSS than the table’s 4K figure suggests. GameGPU’s launch-week test also called the Castor Woods foliage rendering suboptimal, with trees, bushes and small objects that “pop out” in front of the player, so do not expect any value of this slider to remove pop-in entirely.
RTX 3060 starting config (1080p)
Use this table as a starting point and expect to adjust it. No source I found measured an RTX 3060 here; Deltia’s RTX 30 guide says the game “plays very well” on RTX 30 cards, that a 3060 can play 1080p with native anti-aliasing, and that its own table is “a starting point”; Sportskeeda’s VRAM note says the game does not spike past 6-7 GB, while PC Optimized measured up to just under 10 GB at maximum settings at 1080p and 1440p, which is why I suggest Medium textures on 8 GB cards. Where the two guides disagree I picked the value with the lower measured cost and list the other.
| Setting | Value | Why | Where guides disagree |
|---|---|---|---|
| Resolution / upscaler | 1920 × 1080, DLSS Quality, DRS off | Both guides use DLSS Quality on a 3060 | Deltia’s: Balanced for an RTX 3050 |
| Frame Generation | Off | NVIDIA’s frame generation needs an RTX 40-series card or newer | None |
| Latency Reduction (Reflex) | Test both | Reflex caused stutter on day one (see stutter section) | Deltia’s: Reflex; Sportskeeda: Off |
| Renderer / Async Compute | D3D12 / On | Up to 5% faster on a 4090 | None |
| Texture Quality | High on 12 GB, Medium on 8 GB | Lowering textures and draw distance to the lowest settings saves over 1 GB (PC Optimized) | Deltia’s: Medium |
| LOD Quality | High | Low only gains about 3% | None |
| Draw Distance Multiplier | 140 | Biggest non-RT cost | PC Optimized: 150-200 |
| Shadows / Screen Space Shadows | Medium / On | +5% for little loss | Both guides: High |
| Ambient Occlusion | Medium | High over-darkens | Sportskeeda: High |
| Global Illumination | High | Negligible cost | Deltia’s: Medium |
| Reflections / Fog / Postprocess | Medium | All near-free in testing | Reflections: Low (Deltia’s) vs High (Sportskeeda) |
| Particles Quality | Medium | Sportskeeda drops it to Medium on the 3060 only | Deltia’s: High |
| All ray-tracing options | Off | See above | None |
That table is a configuration, not a guarantee of 60 FPS. To check it, pick one repeatable route that includes forest and rooftop parkour, record 1% lows with any overlay, and run it before and after each change. If you are short of 60, cut in this order, from least visual damage to most:
- DLSS Quality to Balanced (about 8% by the 4K measurements).
- Draw Distance Multiplier from 140 towards 100.
- Shadows from Medium to Low (about 5% more, with artefacts in artificial-light shadows).
- LOD Quality from High to Low (about 3%).
For the general method behind a list like this, see our complete guide to optimizing PC game settings, and for 3060-specific advice across games, our RTX 3060 best settings guide. The same pattern, where one option dominates the cost, shows up in our Resident Evil 9 settings guide.
Stutter: what is documented and what has been fixed
The documented stutter sources are narrower than forum threads suggest, and one is a confirmed, mostly fixed bug. This table lists each with its evidence and status.
| Cause | Evidence | Status | What to do |
|---|---|---|---|
| Nvidia Reflex with frame generation | Overclock3D’s testing: stutter on day one, and enabling Reflex alone caused it | Hotfix 1.2.3 (30 Sep 2025) fixed “occasional stuttering” with Frame Generation; Overclock3D says “some stutters remain” | Toggle Latency Reduction and watch the frame-time graph |
| Video memory growing during long sessions | allthings.how relays player reports: usage climbs after roughly 30-90 minutes, worse on 8 GB cards | No Techland confirmation found | Returning to the main menu clears it, per the same reports; keep textures at Medium on 8 GB |
| Foliage and LOD in forests | GameGPU, 18 Sep 2025: stutter “even on an RTX 4080” in forest scenes without upscaling | Patch 1.3 (30 Oct 2025) addressed LOD and texture streaming issues; no source ties the two | Use the upscaler; lower Draw Distance Multiplier |
| Settings menu hitching | Techland, Update 1.7 | Fixed 17 Sep 2026 | Update the game |
One more conflict is worth knowing. A Deltia’s stutter guide advises switching the upscaler to Native, while every other source treats DLSS as the way to cut frame-time spikes in forests. I would test Native for five minutes and keep whichever gives better 1% lows. If your stutter looks like uneven frame pacing instead of hitches, our Death Stranding 2 frame pacing guide covers the general diagnosis, and our frame generation guide explains when it helps.
Does Restored Land mode run worse?
I found no source that says it does. Restored Land arrived in Update 1.6 on 26 March 2026 as a solo mode with a persistent world, where, per Techland’s release notes, “enemies remain defeated and loot can only be collected once”. Hotfix 1.6.2 on 13 April 2026 fixed a “potential crash on Restored Land start/load”, and that is the only technical fix tied to the mode in the notes I read. No benchmark or Techland statement measures its frame rate against the normal game.
If you suspect it, test it yourself: run the same route in Restored Land and in a normal save with the same settings and compare 1% lows. A persistent save holds more world state than a fresh one, but that is a guess, and I would not change a single setting for it without your own frame-time graph showing the problem.
Which settings for which player
| If you are | Do this | Skip this |
|---|---|---|
| On an RTX 3060 12 GB or similar | The starting config above; DLSS Quality | Ultra preset, ray tracing |
| On an 8 GB card (3060 8 GB, 4060) | Textures Medium, DLSS Quality, restart to the main menu if frame rate sags after an hour | Native resolution |
| Chasing 144 FPS | Everything in the table, then test Reflex on and off; frame generation only if base FPS is already around 45 or higher (PC Optimized) | Draw Distance above 150-200 |
| On an RTX 4070 or faster at 1440p | DLSS Quality; try ray-traced global illumination alone | Ray-traced shadows and reflections |
| Playing for screenshots at 4K | Ultra preset (ray tracing and frame generation on) | Expecting stable 1% lows |
The honest summary: pick the High preset, set the upscaler by hand, leave ray tracing off unless your card is well above the 3060 class, and spend your frame budget on draw distance only after the rest is settled.
New to the game itself rather than its settings? Our Dying Light: The Beast beginner’s guide covers which mode to start in, which Beast Mode skills to buy first and how to survive night chases.
Frequently asked questions
Is ray tracing worth turning on in Dying Light: The Beast?
Only on a high-end card, and only for global illumination. Reflections cost about as much but the effect is limited to diffuse and semi-glossy surfaces, and ray-traced shadows cost little but look worse than the shadow maps in the one review that compared them. On an RTX 3060 none of it is worth the frames.
Should I use the Ultra preset?
No, because Ultra is the preset that enables ray tracing and frame generation together, and the presets barely differ otherwise: Very Low is only 10-12% faster than High at native 4K. Choosing High and then adjusting the five or six options in the table gives you control the presets do not.
Does frame generation fix low FPS on an RTX 3060?
No. NVIDIA’s frame generation needs an RTX 40-series card or newer (listed as a 40-series feature even by CheckFPS), and it multiplies the frames you already have rather than rescuing a low base. I found no test of AMD’s alternative on an RTX 30 card in this game, so the safe path is DLSS Quality plus the cost table. Our DLSS vs FSR vs XeSS comparison covers the upscalers in general.
Why do settings guides contradict each other?
Because most options cost almost nothing, so authors pick by taste. Reflections Low versus High, fog Medium versus High and Reflex on versus off all differ between the two guides I used, and in each case the measured difference is small or untested. Trust the entries that come with a measured percentage and treat the rest as preference.
Sources
- Techland. Update 1.4 patch notes (ray tracing, Ultra preset behaviour, requirements). Steam Community announcements, 27 Nov 2025 (https://store.steampowered.com/news/app/3008130/view/1817483467036903)
- Techland. Restored Land Edition release notes (Update 1.6). Steam Community announcements, 26 Mar 2026 (https://store.steampowered.com/news/app/3008130/view/1827626365770963)
- Techland. Anniversary Update release notes (1.7). Steam Community announcements, 17 Sep 2026 (https://store.steampowered.com/news/app/3008130/view/1844115010491804)
- Techland. Update Patch 1.3 notes. Steam Community announcements, 30 Oct 2025 (https://store.steampowered.com/news/app/3008130/view/1815034432873494)
- Valve/Techland. Dying Light: The Beast, Steam store page (system requirements, release date 18 Sep 2025) (https://store.steampowered.com/app/3008130/Dying_Light_The_Beast/)
- PC Optimized Settings. Dying Light: The Beast PC Settings Optimization Guide + Ray Tracing (RTX 4090 test, first published 19 Sep 2025) (https://pcoptimizedsettings.com/best-settings-for-dying-light-the-beast-pc-optimization-guide/)
- DSOGaming. Dying Light: The Beast Ray Tracing & DLSS 4 Benchmarks (RTX 5090) (https://www.dsogaming.com/pc-performance-analyses/dying-light-the-beast-ray-tracing-dlss-4-benchmarks/)
- GameGPU. Dying Light: The Beast test GPU/CPU (launch-week test, 18 Sep 2025) (https://en.gamegpu.com/test-gpu/action-fps-tps/dying-light-the-beast-test-gpu-cpu)
- GameGPU. Dying Light: The Beast (Patch 1.4) Ray Tracing Benchmark. Steam Community group post (https://steamcommunity.com/groups/gamegpu/discussions/0/690868707633694345/)
- Overclock3D. Dying Light: The Beast Hotfix 1.2.3 fixes DLSS Frame Generation support (https://overclock3d.net/reviews/software/dying-light-the-beast-hotfix-1-2-3-fixes-dlss-frame-generation-support/)
- DSOGaming. Dying Light: The Beast Hotfix 1.2.3 released, full patch notes (30 Sep 2025) (https://www.dsogaming.com/patches/dying-light-the-beast-hotfix-1-2-3-released-full-patch-notes/)
- Deltia’s Gaming. Dying Light The Beast: Best Settings For RTX 30 Series (https://deltiasgaming.com/?p=351047)
- Sportskeeda. Best Dying Light: The Beast graphics settings for mid-range GPUs (18 Sep 2025) (https://www.sportskeeda.com/gaming-tech/best-dying-light-the-beast-graphics-settings-mid-range-gpus)
- allthings.how. Dying Light: The Beast PC VRAM leak: symptoms and short-term fixes (updated 26 Sep 2025) (https://allthings.how/dying-light-the-beast-pc-vram-leak-symptoms-and-short-term-fixes/)
- CheckFPS. How to improve FPS in Dying Light: The Beast (modelled estimates, not benchmarks) (https://checkfps.io/en/optimize-fps/dying-light-the-beast)
- MP1st. Dying Light: The Beast Hotfix 1.6.2 (Update 1.006.002), 13 Apr 2026 (https://mp1st.com/title-updates-and-patches/dying-light-the-beast-hotfix-1-6-2-update-1-006-002)
I've been playing video games for over 20 years, spanning everything from early PC titles to modern open-world games. I started Switchblade Gaming to publish the kind of accurate, well-researched guides I always wanted to find — built on primary sources, tested in-game, and kept up to date after patches. I currently focus on Minecraft and Pokémon GO.