In short
- Oblivion Remastered Steam Deck settings that work: Low preset, 1280×800, Lumen hardware ray tracing off, frame generation off, 30 FPS cap.
- Outdoors the Deck sits near 30 FPS with dips to 20; one report says indoors it can pass 40 with the cap off.
- Removing Lumen in Engine.ini is reported at +5 to 10 FPS, from one unbenchmarked guide.
Checked against: Valve’s compatibility data, Steam’s update feed, Epic’s Lumen guide.
Oblivion Remastered is Deck Verified, yet on a Steam Deck it still lands near 30 FPS in the open world, with dips to about 20. Valve’s compatibility data for the game resolves to the Verified category, and the Deck reports I found agree on about 30 outdoors: The Backdash measured around 30 FPS on average with drops to 20, SteamDeckHQ wrote that “this is not a game you will be playing above 30 FPS”, and Notebookcheck reports drops to 20 FPS in the open world and says more than 40 is possible indoors with the cap off.
So the 30-versus-40 question has a short answer: 40 FPS is, at best, an indoor target. Outdoors, the settings below buy a steadier 30, not a 40. The biggest single lever is Lumen, Unreal Engine 5’s lighting system. One guide reports that removing it gains 5 to 10 FPS, and further down I turn that into milliseconds so you can see what it does and does not buy. Two limits first. Every Deck report I found dates from April or May 2025, and I found no per-setting benchmark on a Deck. The one Deck-specific figure, for Lumen, comes from a guide that states no test method, and the only per-setting percentages come from an RTX 4090 at 4K. Those rank the settings by cost; they do not predict Deck FPS. I did not play-test this on a Deck either, so nothing here is first-hand measurement.
Verified 7 November 2026 against the game’s Steam announcement feed, where the newest game update is 1.2 of 9 July 2025. I found none since, and none of the Deck reports above was measured on that build. Values may change with a future patch.
Quick start: the 2-minute Steam Deck setup
- Keep the Low preset the game boots with. PCGamesN found it pre-configured to Low with XeSS upscaling, and SteamDeckHQ saw the same Low default.
- Set 1280×800, V-Sync off, motion blur off. 1280×800 is the Deck’s native resolution, and it is what The Backdash used.
- Lumen Hardware RT: Off. Lumen Software RT Quality: Low. Both Deck settings guides turn hardware ray tracing off.
- Frame Generation: Off. The Backdash switches it off to avoid latency and ghosting, and PCGamesN warns about the same two problems (the FAQ explains).
- Cap the frame rate at 30 in the Deck’s per-game profile and leave the power limit at the 15 W ceiling. Valve lists the APU range as 4 to 15 W, and Overkill.wtf also uses a per-game 30 FPS cap.
- Optional: the one Engine.ini line in the checklist below, if outdoor dips still bother you.
- Test the same stretch of road or the same city street before and after every change. One change at a time, otherwise you cannot tell which one helped.
What Deck Verified does and does not promise
Verified means Valve’s testers passed the game as shipped, not that it holds 40 or 60. The checklist behind the rating in the report data includes an item called “default configuration is performant”, which is a statement about the default Low preset and nothing about your tweaked one. PCGamesN argues the rating is undeserved because the game is “a struggle to run at a stable 30fps without aggressive upscaling”, and its route above 30 is frame generation, which it says brings “input lag and extreme image ghosting”. It suggests streaming the game through GeForce Now instead. That is a fair option if you have the connection; this guide is for playing it on the Deck itself.
30 or 40 FPS: what the frame-time budget says
Going from 30 to 40 FPS means taking 8.3 ms out of every frame: 33.3 ms becomes 25.0 ms. That is the whole difference, and it is large, because it is a quarter of the frame. A 40 FPS cap feels best when the Deck also refreshes the screen at 40 Hz. SteamOS 3.2 added a 40 Hz mode where the cap can be 40, 20 or 10, and the article quotes Valve calling 40 Hz “a sweet spot for responsiveness, consistency, and smoothness”. Our Fallout 4 Steam Deck settings guide covers that pairing for Bethesda’s other Deck title; the OLED versus LCD comparison covers the two panel types.
Here is where the reports put the two targets. Outdoors the Deck lands near 30 with dips to 20, so a 40 cap would only ever be reached in short stretches. SteamDeckHQ found caves and sewers gave “little to no problem hitting 30 FPS”, and Notebookcheck says indoors with the cap removed it can pass 40. My plan: cap at 30 everywhere, and only build a 40 Hz profile for dungeons if your own frame-rate readout stays above 40 in them, since only one report shows it. A cap you cannot hold is worse than a lower cap you can.
Lumen: the one setting that needs a Deck-specific decision
Lumen is the lighting system that makes Oblivion Remastered look different from the 2006 game, and it is also the most expensive thing on the screen. Epic’s Lumen guide says Lumen targets 30 and 60 FPS on consoles with 8 ms and 4 ms frame budgets at 1080p, for global illumination, reflections, translucency and volumetric fog combined. At 30 FPS that is 8 ms out of every 33.3 ms frame, about a quarter. Epic calls software ray tracing “the fastest tracing method in Lumen” and hardware ray tracing the more expensive one, which is why every Deck guide sets Hardware RT to Off. On a 4090 at 4K, PC Optimized Settings measured hardware Lumen up to 25% slower than software Lumen.
The Deck-specific number comes from Overkill.wtf. Its authors found an engine parameter that switches LumenRT off and wrote that “the performance gains lie between 5 and 10 FPS, depending on the scene”, to see “less dips below our target frame rate of 30FPS”. The same page warns that the game “was most likely built with LumenRT in mind”, so removing it gives many scenes “a more grounded, maybe slightly darker, look”. It is a guide’s figure with comparison shots and no stated test method, so I treat it as a plausible range rather than a benchmark.
Here is what that range means in milliseconds, using plain frame-time arithmetic (1000 divided by FPS). It is our calculation, not a measurement:
| You start at | If you gain 5 FPS | If you gain 10 FPS |
|---|---|---|
| 20 FPS (50.0 ms) | 25 FPS (40.0 ms), saves 10.0 ms | 30 FPS (33.3 ms), saves 16.7 ms |
| 25 FPS (40.0 ms) | 30 FPS (33.3 ms), saves 6.7 ms | 35 FPS (28.6 ms), saves 11.4 ms |
| 30 FPS (33.3 ms) | 35 FPS (28.6 ms), saves 4.8 ms | 40 FPS (25.0 ms), saves 8.3 ms |
Two readings follow. A 20 FPS dip only gets back to 30 at the top of the range, and a 30 FPS scene only reaches 40 at the top of it too. So Lumen off is a stability fix for the bad spots, not a ticket to 40 outdoors. As a consistency check, the 8.3 ms saved by +10 FPS from 30 is close to Epic’s 8 ms console budget for all of Lumen. That supports the upper end as possible, but a console budget at 1080p is not a Deck measurement, and the Deck renders far fewer pixels, so I read the match as a sanity check and not as confirmation.
What each setting costs, and what to set on the Deck
The costs below are relative percentages from PC Optimized Settings, measured on an RTX 4090 at 4K. They tell you which sliders are expensive, not how many FPS each is worth on a Deck. The last column is our pick, built from the two Deck settings guides.
| Setting | Published cost (RTX 4090, 4K) | Deck pick |
|---|---|---|
| Lumen Hardware RT | Up to 25% slower than software Lumen | Off |
| Lumen (removed in Engine.ini) | +5 to 10 FPS on a Deck, per one guide | Optional, see checklist |
| Global illumination level | Up to 3 to 5% between levels with software Lumen | Low |
| Shadows | Ultra up to 20% slower than the lower levels | Low (Medium if you can spare it) |
| Foliage | 7 to 8% gain from reducing foliage quality | Low |
| Effects | Up to 10% | Low |
| Reflections | About 5% with software Lumen | Low |
| Post-processing | 2 to 4% | Low or Medium |
| View distance | Minimal | Medium |
| Frame generation | 60 to 80% gain at 4K on the 4090 | Off |
The takeaway: Lumen and Shadows are where the big costs sit, and the rest of the menu is 10% or less, with frame generation aside, which stays off. The whole preset ladder says the same thing. PC Optimized reports Ultra at 56.5 FPS, High at 75 (+33%), Medium at 85 (+50%) and Low at 97 (+72%) at 4K with quality upscaling, and the Deck already boots at Low, so there is little preset room left to find. That is why the remaining decision is Lumen, not another slider.
The Deck guides do not fully agree with each other, and you should know where. The Backdash sets everything to Low and Screen Space Reflections off, and Overkill.wtf uses Medium for view distance, foliage, shadows, global illumination, textures and reflections, with Screen Space Reflections on. Overkill.wtf aims at a 30 FPS cap and The Backdash runs the cap at 60 and reports an average near 30 with dips to 20, so both land in the same place by different routes. I chose Low with Medium view distance because the view-distance cost is minimal in the data above, and I left Screen Space Reflections for you to compare, since the two guides split on it.
XeSS or FSR 3: the upscaler on a 640×400 image
The game boots with XeSS, and the guides disagree on the mode. SteamDeckHQ saw XeSS Performance; PCGamesN saw XeSS Balanced; Overkill.wtf picks XeSS Balanced; The Backdash picks FSR 3 at Ultra Performance with sharpness 2. The cost of that is resolution. Notebookcheck states that Low with XeSS Performance at 1280×800 renders at an effective 640×400, which is a quarter of the pixels (256,000 against 1,024,000), and Ultra Performance renders lower still. Rendering fewer pixels is the main way an upscaler saves frame time, and the price is softer detail.
On image quality, SteamDeckHQ found FSR 3 Performance blurrier than XeSS Performance and kept XeSS. Bethesda’s Update 1.2 notes include “Fixed flickering shadows when using XeSS Upscaling”, so any older advice to avoid XeSS because of shadow flicker needs a date check. My pick is XeSS Performance, the default SteamDeckHQ saw and the mode it preferred. Step up to Balanced only if your dips are rare, and use FSR 3 Ultra Performance as a last resort for a scene that will not hold 30. For the general trade-off across all three upscalers, see our DLSS vs FSR vs XeSS comparison.
The Engine.ini checklist, and why there are no before-and-after frame times
No published source gives before-and-after frame times for an Engine.ini change on a Deck. What exists is Overkill.wtf’s 5 to 10 FPS figure for removing Lumen and PC Optimized’s 5 to 6% for turning off virtual shadow maps on an RTX 4090. So the checklist below pairs each step with what its source claims, and the ledger afterwards is for your own numbers.
| Step | What to do | Reported effect |
|---|---|---|
| 1 | Launch the game once, then quit. The config folder does not exist before the first run. | Overkill.wtf: run it at least once first |
| 2 | Open Desktop Mode and go to the game’s Proton folder, then Engine.ini. Back the file up. | None, this is safety |
| 3 | Add the two lines below, one under each heading. | +5 to 10 FPS, scene dependent (one guide) |
| 4 | Do not touch the in-game Lumen options afterwards. | Overkill.wtf: otherwise you may need to edit the file again |
| 5 | Optional: set the virtual shadow maps variable to 0. | 5 to 6% on an RTX 4090, no Deck report found |
| 6 | Walk the same route and write the frame rate down. | Your own before and after |
The config lives inside the game’s Proton prefix. Overkill.wtf gives the folder as the following, written here with spaces around the slashes so it wraps on a phone: home / deck / .local / share / Steam / steamapps / compatdata / 2623190 / pfx / drive_c / users / steamuser / Documents / My Games / Oblivion Remastered / Saved / Config / Windows. The lines it adds are:
[/Script/Engine.RendererSettings]
r.Lumen.DiffuseIndirect.Allow=0
[ConsoleVariables]
r.Lumen.DiffuseIndirect.Allow=0Expect a more grounded, maybe slightly darker look, which the guide calls not too drastic; it says the most visible differences are in shaded areas and sunlit city streets. If you would rather not edit files, the same authors published a script that applies an optimised Engine.ini plus the Lumen Begone mod and claims a stable 30 FPS, again with no benchmark data. It runs from Desktop Mode after the first launch, and may need re-running after a game update overwrites the settings. I could not open the Nexus Mods pages for the popular tweak packs (they return an access error), so I do not repeat their claims.
For your own before-and-after, write down the same three scenes each time: a road outside a city, a busy street in the Imperial City, and a cave. Convert each reading to milliseconds with 1000 divided by FPS, and judge a change by the frame time saved in the worst scene, because the dips are what you feel.
| Scene | Before (FPS / ms) | After (FPS / ms) |
|---|---|---|
| Open road | ||
| City street | ||
| Cave |
What changed since the April 2025 guides
Everything the Deck reports measured predates Update 1.2. Its performance section lists fixes for frame rate drops in Deepscorn Hollow, Black Rock Caverns, between Skingrad and Skingrad Castle, and south of Bravil Castle courtyard, plus “General improvements to frame time in many locations” and optimised light and shadow updates in several lairs. That means the 30 FPS reports were taken without those fixes, and no source I found re-tested a Deck afterwards. The evidence from the other side is mixed: Notebookcheck reported in September 2025 that some PC players saw worse performance after the update, one player citing a fall from 50-75 FPS before the patch to 30 FPS in city areas. The article does not say which hardware that player used, so I would not read it as a Deck regression. If you retest on today’s build and get different numbers, trust yours.
There is also a 2026 fan mod, the Oblivion Remastered Performance Fix. Its author’s test PC is a Ryzen 7 5800X with an RX 9070 XT, where one 4K Imperial City flight with frame generation went from 72 to 102 FPS, and a CPU-limited 720p Cheydinhal run without it from 45 to 57 FPS, which its table lists as +25% (each figure there is a mean or median of two to four runs). The Deck renders even fewer pixels and has four Zen 2 cores, so the CPU-limited row is the closest published analogue to it. That is my inference, not a result. The page lists Steam Deck as untested, one Linux report on CachyOS, and every build as beta, with 0.5 to 1 GB more video memory used. I would wait for a Deck report before installing it.
Which build for which player
| If you are | Do this | Why |
|---|---|---|
| Battery-first | 30 cap, a power limit below 15 W, Low, XeSS Performance | SteamDeckHQ got 1.5 to 2 hours on the LCD with no power limit |
| Visuals-first | Medium shadows and foliage, XeSS Balanced, keep Lumen | Overkill.wtf’s Medium list; accept dips toward 20 |
| Dips-hater | Lumen off in Engine.ini, Low settings | 5 to 10 FPS reported, a darker look |
| Dungeon crawler | 40 FPS and 40 Hz profile, only if you hold above 40 | One report says indoors can pass 40 uncapped |
Decide in this order. Start from the Low preset the game boots with and XeSS Performance, the mode SteamDeckHQ saw as the default. If outdoor dips still go below 25 FPS, apply the Lumen line, the largest reported single change. If the darker lighting bothers you, delete the two lines and live with the dips. Battery matters on top of all of this: our Steam Deck battery life guide and the Diablo 4 Steam Deck settings guide, which builds two power-limited profiles and one uncapped one, show how to trade frame rate for battery hours. For the general method behind ranking settings by cost, see our PC game settings optimization guide.
Frequently asked questions
Is Oblivion Remastered playable on Steam Deck?
Yes, at about 30 FPS with dips to 20 in the open world, if you accept Low settings and heavy upscaling. Verified describes the default configuration, not a promise of smoothness, so judge it by the frame rate you see. If you want a clean 40 or 60, the reports I found do not show the Deck delivering it outdoors.
Should I use frame generation to reach 40 or 60 FPS?
No. PCGamesN says going above 30 needs frame generation, which brings input lag and ghosting, and The Backdash turns it off for the same reason. Frame generation inserts extra frames between rendered ones; it does not shorten the slow frames themselves. Spend the effort on the settings above and keep the cap at 30.
Is removing Lumen worth the darker look?
Only if outdoor dips remain after you drop the cheap settings. The reported gain of 5 to 10 FPS is real money at 30 FPS, since +10 FPS saves 8.3 ms, a quarter of the frame, and +5 saves 4.8 ms, but the figure comes from one guide, and the lighting change is permanent until you delete the lines. The edit is reversible, so test it on one route and decide by what you see. The guide says the most visible differences are in shaded areas and sunlit city streets.
Should I wait for the fan Performance Fix instead?
Wait for a Deck report. Its author lists the Steam Deck as untested and every build as beta, and a mod that rewrites code in memory is a bigger bet than two lines in a text file. The 25% gain in its table’s CPU-limited row is interesting for a Deck, but until someone runs it on one, it is a hypothesis.
Sources
- Valve. Steam Deck compatibility report data for The Elder Scrolls IV: Oblivion Remastered (app 2623190), read 7 November 2026. Steam (https://store.steampowered.com/saleaction/ajaxgetdeckappcompatibilityreport?nAppID=2623190)
- Bethesda Game Studios. Oblivion Remastered Update 1.2 (9 July 2025) and Update 1.1 (4 June 2025) patch notes. Steam Community announcements (https://store.steampowered.com/news/app/2623190)
- SteamDeckHQ. The Elder Scrolls 4: Oblivion Remastered Is Mostly Solid on the Steam Deck – First Look (Noah Kupetsky, 22 April 2025) (https://steamdeckhq.com/news/the-elder-scrolls-4-oblivion-remastered-is-mostly-solid-on-the-steam-deck-first-look/)
- The Backdash. Can you run The Elder Scrolls 4: Oblivion Remastered on the Steam Deck optimally? (26 April 2025) (https://thebackdash.com/tech/elder-scrolls-4-oblivion-remastered-steam-deck/)
- Overkill.wtf. Elder Scrolls Oblivion Remastered best settings for Steam Deck (Timo Schmidt, 23 April 2025) (https://archive.overkill.wtf/elder-scrolls-oblivion-remastered-best-settings-steam-deck/)
- Overkill.wtf. Automatically optimize Oblivion Remastered on Steam Deck (2 May 2025) (https://overkill.wtf/steam-deck-oblivion-remaster-automatic-optimization)
- Notebookcheck. Oblivion Remastered zeigt Performance-Probleme am Steam Deck (22 April 2025, in German) (https://www.notebookcheck.com/Oblivion-Remastered-zeigt-Performance-Probleme-am-Steam-Deck-Video-vergleicht-Grafik-mit-Original-aus-2006.1003273.0.html)
- Notebookcheck. Oblivion Remastered performance worsens after patch, leading to stutters and mixed Steam reviews (September 2025) (https://www.notebookcheck.net/Oblivion-Remastered-performance-worsens-after-patch-leading-to-stutters-and-mixed-Steam-reviews.1108215.0.html)
- PCGamesN. Oblivion Remastered doesn’t deserve its Steam Deck Verified status (Niall Walsh, updated 24 April 2025) (https://www.pcgamesn.com/steam-deck/oblivion-remastered-performance-issues)
- PC Optimized Settings. Best Oblivion Remastered Settings and Optimization Guide (RTX 4090 test, 23 April 2025) (https://pcoptimizedsettings.com/elder-scrolls-oblivion-remastered-pc-optimized-settings-performance-guide/)
- Epic Games. Lumen Performance Guide for Unreal Engine. Unreal Engine documentation (https://dev.epicgames.com/documentation/en-us/unreal-engine/lumen-performance-guide-for-unreal-engine)
- LTSharpe. Oblivion Remastered Performance Fix (README, benchmarks and install notes). GitHub (https://github.com/ltsharpe/oblivionremasteredperformancefix)
- Boiling Steam. SteamOS 3.2 Introduces Flexible Refresh Rates for the Steam Deck (29 May 2022) (https://boilingsteam.com/steamos-3-2-introduces-flexible-refresh-rates-for-the-steam-deck)
- Valve. Steam Deck technical specifications. steamdeck.com (https://www.steamdeck.com/en/tech)
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.
