**Getting Started with Modding and Performance Settings**

For those new to modding and performance settings, we've created this guide as a starting point. As modders have more time to experiment with the game, they can refine their understanding of the game's mechanics and optimize performance. This is especially useful for those who want finer tuning between Ultra and High or other settings that offer better performance without sacrificing visual quality.

By using these performance settings as a foundation, you can then fine-tune them manually to achieve your desired level of quality. There are also variations of this approach that increase graphics quality even further, allowing you to experiment with different settings and see what works best for you. With modding and performance tweaking, the possibilities are endless, and we're excited to explore the many options available in Starfield.

**Technical Considerations**

From a purely technical standpoint, there's a lot to consider when it comes to optimizing performance and visual quality in Starfield. One of the most impactful settings is Shadow Quality, which can make or break the visuals in certain areas of the game. When Low and VRS are turned off, shadows look particularly poor, but turning on VRS (Variable Rate Shading) resolves this issue immediately.

In contrast, Contact Shadows and Volumetric Lighting don't show significant improvements at higher quality settings, although they're still important for creating a immersive experience in certain areas of the game. The quality differences are noticeable, especially when it comes to Shadow Quality and Reflections, which both look their best when maxed out.

**Performance Impact**

While some performance settings have a significant impact on GPU usage, others are relatively minor. For example, Volumetric Lighting has a moderate performance impact, while Grass Quality is more impactful in terms of GPU usage.

However, with the right tuning and optimization, it's possible to find a balance between visual quality and performance that meets your needs. By experimenting with different settings and adjusting them as needed, you can create a gaming experience that suits your style and system capabilities.

**Crowd Density and Grass Quality**

Two other settings that are worth exploring are Crowd Density and Grass Quality. Crowd Density is particularly interesting because it affects the behavior of NPCs in certain areas of the game. When set to High, crowd density is more realistic, but this requires reloading the save or re-entering a zone to see the effects.

Grass Quality also has an impact on the visual fidelity of the game world, with higher settings resulting in thicker grass and a greater view distance. This can be particularly noticeable when exploring large areas of the game, making it worth experimenting with different Grass Quality settings to find one that suits your style.

**Conclusion**

This guide provides a starting point for those new to modding and performance settings in Starfield. By experimenting with these settings and fine-tuning them to suit your needs, you can create a gaming experience that's both visually stunning and relatively stress-free on your system. Remember to check out mods on Nexus Mods as they develop, and don't hesitate to reach out if you have any questions or need further guidance.

**Supporting This Content**

If you want to support this type of content and help us continue creating high-quality guides like this one, be sure to check out our store at Gamers Access.net. We've recently launched a new line of modding tools and accessories that make PC building easier than ever, including our brand-new gn15 mod mats. These mats provide a large work surface with diagrams, screw tracking, and other features to help you build the perfect gaming rig.

Thanks for watching, and we'll see you all next time!

"WEBVTTKind: captionsLanguage: enthis massive Graphics guide for Starfield looks at the mod packs that are out now including some dlss and an Optimizer mod we'll also be looking at Graphics comparisons A to B and we're doing a ton of benchmarks across different graphic settings presets and explaining a deeper level of optimization you can do Beyond just changing the settings word in the Starfield menu and this is our final installation of our Starfield coverage for right now at least we'll come back to this topic Starfield will become a permanent test in our suite we've decided and as more mods come out and patches to the game come out we'll revisit of course whenever those happen but for now this is the last immediately planned piece and we're excited for it because uh for those of you who've been around a while in 2015 when GTA 5 and The Witcher 3 came out our some of our original sort of uh claim to internet success was GTA 5 optimization guides and benchmarking and The Witcher 3 Graphics benchmarking uh and I loved working on those and we're back now with Starfield so let's get started before that this video is brought to you by the thermal T swappable blade fans available in 120 and 140 mm sizes the new thermal T fans include three sets of swappable blades so that even as you change builds or cases you can ensure the LEDs are always presented on their best side the swappable blades allow Builders to get the fan frame out of the way of the lights by reversing the blade direction to reconfigure the fan as push or pull while keeping the struts relatively hidden and keeping the fan frame oriented One Way swapping blades is done by applying pressure evenly to opposite sides then pressing until the click each fan also has Pinto pad connections for cess daisy chaining and you can learn more at the link in the description below okay so the way this works is we're going to look at each of the settings individually for the most part we will be comparing them not only visually but with a number of benchmarks across a couple different platforms those include CPU bound constraints and GPU bound constraints and additionally we'll be testing the graphic settings and tuning them from two directions one of them is from a baseline of 1080p low meaning we are adjusting the setting under test up into higher uh demand and the other one is from a baseline 1080p High meaning we're adjusting the setting down so you can sort of view it like 1080p low Baseline adjusting settings up is you are at the the lower end of performance capabilities with your hardware and you're almost shopping for settings what can I add to my cart to get the most visual impact with the least performance impact or meet some kind of trade in between and then likewise from 10 AP High Baseline you're more looking for I'm pretty close to the performance I want what can I take away to just get over whatever that threshold is that you're targeting for your performance so that's how we're going to be looking at this that involves a lot of testing we tested in a couple different areas as well uh and we're looking at some mods optimizing starfields Graphics can go a lot deeper than just configuring the menu to those settings that means you could if you wanted to get into fine-tuning parameters like maximum decals the draw distance per setting like grass mesh coling the radius of things such as ambient occlusion the shadow map count the particle count and quality resolution scaling and more these individual settings have already been tuned by one of the mods if you don't feel like you understand how to change them to get the Maximum Impact and we'll be looking at that today additionally we're benchmarking those from the default configuration and we're testing the game for an optimization mod we'll also be testing two dlss mods and we're adjusting fov for just a quick test there as well one quick thing right at the top of this too as we were finalizing this video Bethesda posted a note that it's working on official dlss integration an fov slider HDR support and Gamma adjustments these are not integrated yet but Bethesda did post a small patch none of these features though were included the patch claimed Improvement to Performance generally speaking without being specific at all we just ran two quick validation tests and we found that the RX 7600 had no change to Performance when tested at 1080p high or 1080p low in our test conditions the RTX 4060 saw a 6% performance uplift in one test condition but upon running Nvidia profile inspector we found that this is because Nvidia pushed a patch that added rebar profiles the 6% uplift could be a mix of the rebar profile with whatever Bethesda changed or it's just from the rebar profile alone either way as of right now there are no major changes from the patch it's mostly to fix a few Quest breaking bugs we'll go back to the rest of the content now especially since the patches in the future won't affect most of the graphics comparison any way in terms of visual changes or optimization relative values and they will affect of course absolute values but we'll test those as they come out we're going to give you a quick summary chart upfront of our findings to try and provide the most helpful information as fast as possible couple things some of these matters will be subjective so our summary chart looks at the performance impact and the visual impact visual impact is a matter of opinion uh and we have a couple people on the team who contributed opinions to that and we've kind of averaged them the chart is going to be for your quick reference the content today is relatively in-depth and it aims to provide a detailed understanding of the settings and the benchmarks and hopefully this simplification is helpful so here's the table the scale ranks from none to critical performance impact is defined objectively and from our GPU bound test scenario however CPU bound scenarios are explored later in this video the performance impact rankings are objectively defined by the ranges at the bottom of the table some have caveats like grass and crowd density and indirect lighting is also a particularly interesting one for visual impact most of the time it's not too important but there is one specific condition that really ruins the visuals for us and if you're playing with indirect lighting at low without variable rate shading like if you have an older GPU that might be what's making the game look bad for you we're providing the summary chart because we don't want to offer flat recommended settings and that's because they would change based on the hardware and we can't recommend flat settings that will maximize everyone's Hardware so rather than giving a a sort of safe configuration uh that you'll probably be able to run what we're hoping to do is provide the information so that you can then go do the tuning yourself to get the performance that you need based on your Hardware so uh there's be a little bit of work on the users end interpreting that but we've made it as simple as we can while giving you the the the most to work with as an example though a higher-end machine might only focus on the items we've marked as critical performance for extra frame rate while a low-end machine might focus on say moderate to critical or try to balance the preservation of visual impact high and critical items while reducing the quality of other things either way the testing we're doing today is relative scaling from impact of settings which means that for the most part this will all remain basically a hopefully a permanent reference unless Bethesda makes a a very specific tweak to how some of these settings behave uh also a lot of our testing was done with an RX 7600 we have 49s in here as well but that comprises the the bulk of the numbers we're looking at today and additionally uh unless we're talking absolute numbers driver updates aren't going to change the relative scaling of individual settings for the most part so this should be fairly useful long term as well even if you're watching this in a year or so just know that the absolute frame rate numbers might change okay okay enough of the summary time to get into showing our work for what made the summary first of all for this we are looking for any change greater than a couple percent to investigate into greater detail so if we see a Swain of say 6% moving from low to ultra we'll look at the invidual ones in between as well uh we'll start with our new Atlantis testing then we deviate from that for a couple specific areas like a field for grass quality for example uh further all capture was done on an RTX 490 the video capture was an 4K borderless which is uh sort of the the closest Native thing to full screen without modifying the game that's in vanilla right now and our Baseline capture was done with all settings manually maxed out uh with the exception of a couple those are vsync and dynamic resolution both of which were off we also disabled all forms of upscaling for the capture and we disabled variable rate shading for the capture now capture is not benchmarking so the capture we did to show the comparisons was made specifically for visual comparison and has no bearing whatsoever on the performance benchmarking that we're doing later moving to GPU bound benchmarks with the RX 7600 we'll start with Baseline low testing and adjust each setting up individually we're starting in our new Atlantis setting for this first we'll just dump all the test results on the screen and then we'll split them up accordingly and look at some different environments here's everything look at this graph kind of like a shopping cart each item you add costs performance and we're looking at the cost on a per item basis so they're not cumulative here but obviously as you add more of these high demand things the performance will get worse some settings like grass quality don't have any impact in this test area but they do later we'll split them out separately for testing in a field others like motion blur just don't affect performance in any meaningful way in any location that we tested just using this as a quick glance though we saw the biggest impact from Shadow quality we're changing literally only that setting to ultra cost us a staggering 18% of our total possible per perance this setup this is the first setting to adjust if your GPU bound highlighting volumetric lighting as the next one or maybe Reflections as the one after that and then indirect lighting indirect lighting here has a potentially large impact on visuals when using low specifically with no variable rate shading we'll come back to that but it looks really bad jumping quickly to the GPU bind from the other direction with a 1080p High Bas line we get this chart we've intentionally sorted this chart with the Baseline at the top despite it not being the best setting this is the opposite direction from before we start at high settings not Ultra and then we reduce them as we go that restricts the maximum top to bottom swing of course but it attempts to look at this more holistically a lot of people probably try to start with the higher overall settings then individually reduce them which is what this shows reducing volumetric lighting had the largest Improvement going from high to low ignoring ultra settings and with medium volumetric we got the most performance without that much sacrifice at least for this setting Reflections and Shadow quality were also High on the list as was indirect lighting note that several items like everything in the 55 to 56 FPS range are with an error of basine in other words changing them had no impact on performance in this test scenario we can see some changes in other areas but not always let's move to individual charts Shadow quality has particularly high cost at Ultra when tested with our GPU bind configuration medium and high are barely differentiated here with them respectively costing 4% and 7 1.2% of the performance against Baseline that is 4% is still a relatively large change considering it's just one setting but Ultra Remains the obvious choice to avoid for most setups if you're on a higher-end set of hardware and you're bordering on handling the ultra preset this would be the one setting to drop to try and boost it to more satisfactory performance while still maintaining Ultra overall looking at Shadow quality from a high Baseline rather than low we saw 5 1 12% Improvement switching to low Shadow quality which materializes as about 3 FPS average this test moving to medium next gave us a 2 and 1/2% uplift Ultra was obviously much more expensive here let's take a look at what shadow quality actually does though to see if it's worth it the most obvious results of the Shadow quality setting can be seen in the edges of Shadows which generally become softer as the setting is lowered from Ultra to low fine details like the shadows of chair legs are blurred out at low as are the shadows of individual leaves at the bottom right of the frame the differences between medium and high are more complex with shadows on the wall at the right of the frame appearing softer at medium but sharper under the table and chair at the center indoors where Shadows are generally closer to the source casting them the scale from low to ultra can be more subtle moving from Ultra to high practically erases the shadow of the coffee table cast by the fire but otherwise there's little difference medium is similar but moving to low then removes much of the Shadow under the portrait cast by the ceiling light Shadow draw distance also scales with quality even just going from Ultra to high in this scene eliminates many Shadows within the tree canopy in the background and each step down and quality brings the edge of that draw distance closer until at low even the Shadows within the trees closest to the camera are almost gone so this really seems to affect the draw distance indirect lighting didn't have as large of a performance impact as some other settings like Shadow quality but it does have a gigantic impact on visuals depending on which setting you use low with variable rate shading off produces some of the worst looking blurring in the game so older gpus that don't support variable rate shading will look particularly bad at low we'd strongly recommend running this minimally at medium for performance though we're seeing about a 5.8% performance loss by going from Baseline low to ultra indirect lighting when GPU bound it's a fairly large change switching to Baseline high and still GPU bound moving from high indirect lighting to low gave us 4.2% more performance but it has a big visual impact at low so let's look at that it's so incredibly ugly at low that we think the setting itself might be bugged at least at the time of writing if you have to play on low settings to get a decent frame rate we'd strongly suggest that you at least consider turning this one up to medium just to prevent the huge reduction in overall image quality at low at least from what we think here indirect lighting shows no meaningful difference between Ultra and medium though in any area that we check then at low it suddenly looks Jagged shimmery and lickery anywhere there's a straight line areas of the screen appear blurry and low resolution it's especially noticeable in the background highrise here and the metal awning over this balcony for what it's worth the effect is strongest in areas that actually are indirectly lit Shadows illuminated by reflected light in other words this scene in a park shows the same effect distant objects like the Mercury Tower sign and Center screen become a watery mess while some other objects are only affect on the edges like the blue pillars in the center screen this really does look terrible in game if you've been running this setting at low or you're running at 1080p for example we highly recommend pushing this up to at least medium it may change your perception of the graphics quality overall and it should be one of the first settings to increase on the list of players using lowquality settings alternatively enabling variable rate shading also solves this issue which makes it seem even more likely that the low Behavior is a bug we did notice that in some scenes the visual impact isn't clearly visible for example we tried to find it in the club in this bottle full of bub and it's not as obvious there's still some difference specifically at the lowest setting though notice the flickery noise and the reflections on the text in circling the stage again moving to medium makes the scene appear effectively identical to ultra or at least close enough that it's a massive upgrade over low time to get into grass quality our previous summer charts showed nearly zero impact from grass quality but that's because there's not much grass in the new Atlantis scene looking at grass quality tested specifically in a field we see the differences emerge more meaningfully in the numbers tested from all low settings at 1080P and only adjusting grass without upscaling the Baseline result was 86 FPS average with good lows this field doesn't put load on the GPU in the same way that the city does but the grass quality clearly shows an impact overall increasing quality to medium results in a loss of 2. 4% performance while increasing to high cost us a total from low of 6.9% ultra cost us 11% from Baseline low that's a massive swing considering this is just one quality setting and once again Ultra proves to be sort of that last jump that really gets overbearing for a lot of gpus if you're in a GPU bound scenario and you particularly notice issues in grassy areas this would be the setting to adjust just know that it won't help performance issues in areas without much literal grass like cities where there's nowhere near the grass density if there's grass at all crowd density is our next setting to test this setting has a high and observable visual impact but a low impact to Performance we'd strongly recommend keeping this as high as you can to keep the game feeling more alive in a GPU bound setting from a low Baseline we saw a cost of around 2% to increase crowd density to high it wasn't worth testing anything in between since obviously it'd be less than 2% let's look at the visuals crowd density is a weird one our method ology for this piece was to change a setting exit the game open the game load our save and start recording in this process we noticed that Starfield is weirdly consistent in a good way about loading saves that extends to the positions and the paths of the NPCs like this NPC that always walks on top of the couch at least the path finding is consistently buggy but that explains why we didn't see any real difference in the number of NPCs between low and high c density we need to load into a new area and generate a fresh batch of NPCs straight from the Starfield NPC Bakery but that also breaks the side by-side comparison we're trying to do by loading into the Mast District we can get an extremely rough picture of crowd density we captured a couple passes at each setting then synced them up paused and counted the visible NPCs in that single frame with that small sample size we saw 22 NPC C's on average at high 17 at medium and 13 at low we wouldn't base any performance calculations on those numbers but it's enough to confirm that the setting does do something now for Reflections this is another one with a relatively high performance cost for a singular setting depending on how bound you are on the GPU starting with the low Baseline numbers on the 7600 we saw a maximum cost of 7.2% moving to high Reflections or about 2.2 with medium from the other direction tested at a high Baseline we saw an improvement of 5 . 7% by dropping Reflections to low or about 3% dropping it to medium even with everything else loading the GPU heavily Reflections quality remains one of the first settings you could go for to get a performance uplift maybe after Shadow quality and here's how the reflections actually look at the highest level mirrored surfaces show a fairly sharp but non-r traced reflection of the surrounding space while lowering the setting to medium leads to more streaky abstract Reflections even at Max objects like furniture and NPCs aren't reflected at all here moving to the lowest setting wipes away large details and also appears brighter in this instance although we ensur that each individual set of Clips was done at precisely the same time of day so as to minimize any impact there note that the real pillar and floor visible at the right edge of the screen are lit the same at all settings only their Reflections change there's some slight changes in non-mirrored surfaces as well but there's not a clear better or worse in this regard this outdoor Lake scene illustrates the same point the surface of the lake and the exterior of the curved building are brighter with Reflections at high but medium and low are effectively identical even when considering the reflections of scenery on the water surface like some of the other settings in the game it appears as if the biggest change is at the highest quality option at least visually volumetric lighting benchmarks are next and then the comparisons when GP bound and from Baseline low settings we see a performance loss maximally of 12% this setting has a large impact on performance high and Ultra costs about the same medium costs about 3.3% from the Baseline performance tested from a high Baseline instead we saw a 9% performance uplift moving from high to low or about a 6.6% uplift going to medium and a slight loss going to ultra even when more heavily constrained in general this setting still has a large performance cost time to talk about how it looks though the Smoky streets of neon seemed like the ideal place to check for volumetric lighting but they weren't it's not the right kind of smoke we'll use that scene to check some other settings though instead let's look at a facility on Mars Ultra and high looks similar here bringing up the brightness and the sharpness of the fog shrouded lights at the center screen but dropping to medium and low progressively softens the effect with low eliminating some of the individual bright points altogether we also noticed flickering artifacts around one of the lights at all settings gtao testing is next for this ambient occlusion testing we found that a GPU constraint from 1080p low Baseline resulted in a 4% performance loss by changing to ultra for the setting the range for this one is overall low considering the relatively low maximum cost from high Baseline we found no change going to low the performance was within error the lodge is an ideal place to check on ambient occlusion because has lots of small objects cluttering the scene this setting performed completely as expected at Ultra the crevices behind things like the cushions for example are darkened as are areas around items on the coffee table and pieces of furniture the effect is lessened noticeably with each step down to high medium and low with low showing brightly lit areas that should definitely be in a shadow like inside the arms of the couch the effects aren't always so pronounced though although the setting is still working as intended this brightly lit office on Titan mostly shows ambient occlusion around the base of the desk the complicated geometry of the office chair also benefits here with higher settings significantly darkening the areas underneath and inside the chair on the contact Shadows from 1080p low Baseline we observed a 3.6% performance cost hiking to ultra and settling in between would yield most of the visuals from Baseline High settings reducing contact Shadows to its lowest setting gave us 1.4% more performance there's not much worth considering here when other options give us far more range the most obvious instance of contact Shadows that we could find were around NPCs and bright sunlight in this regard ultra high and medium look pretty much identical while shifting to low drops the resolution of the shadow where vasco's legs for example contact the ground and it eliminates some detail in the shadows of the arms and the legs elsewhere the differences are even harder to notice in this outdoor scene the joints and the honeycomb roof structure become darker when the setting is raised above low but changes beyond that are minor most of the NPCs are standing in Shadow so even ones that pass close to the camera like the security guard are unaffected by the setting and the ones in the distance are too far away to see clearly anyway for particle quality benchmarking we found no differences from our Baseline low settings by increasing this setting we even checked in some areas that seemed to more explicitly show particles but found no differences the setting has described as setting quote the quality of particle lighting so thanks Bethesda that really helps we tried examining three specific things closer Sparks from a mining laser Smoke backl by Neon Signs and Sparkle effects on the artifact in a script sequence starting with the mining laser the Sparks generated by firing it have no noticeable effect that we could see on any of the surrounding Shadows at any setting the laser itself does illuminate its surroundings but there's no flickering or variability to suggest that the Sparks themselves are lit there's also not a difference in the quantity of particles generated at any of the settings that we could notice since we didn't see particles lighting the scene we hoped to see the scene lighting particles by moving to the smoke and neon the differences are almost invisible but but we noticed that the cables in the support to the right of the Neon security sign are obscured at the lowest setting accompanied by some shimmering medium and high still look identical in this area though finally in the animated sequence that starts after bringing the artifact to the lodge there are visible differences in lighting and shadows between recordings but flickering and intensity of the light appears to be randomized we can't make a conclusion based on this scene time to briefly look at some benchmarks from the CPU limited angle for this we chose an R5 3600 with an RTX 490 to create a CPU bound test condition overall we found most settings were unable to affect the performance the best shot was Shadow quality again even though it's a GPU heavy setting it's still affecting performance when mostly CPU bound bottlenecks are really so simple as being one component or the other so this makes sense crowd density also affected performance but not much we saw 3.6% performance loss from Baseline going to high for crowd density otherwise everything was within 1% chance or 1.2 to 1.3 for Reflections and contact Shadows we tested these from Baseline High also but stopped because we saw no change time to talk about mods a couple weeks into launch and Nexus mods unaffiliated with us but a great website is already filled with tons of options the modding scene for Starfield is still developing and the modding aspect of Bethesda games is probably the most exciting part so even though all of this stuff is going to get much more sophisticated we downloaded and tested a couple of the early mods each independent from the other the mods we tested include Starfield performance optimizations by Eck Starfield upscaler for dlss by pure dark Starfield frame Generation by Luke FZ 564 and we did a manual fov adjustment in the inii files we'll do the simplest Benchmark first field of view or fov has been a common complaint since the game launched it can be changed without modifying the game other than creating a custom inii file similar to Skyrim but currently at least there's no in-game slider increasing fov widens the peripheral vision of the player's camera we ran some quick tests from the default fov up to a 150 just to see how it did between all of these performance was the same 150 looked awful obviously but it was just there for a test we tried up to 360 uh it didn't work well and most people might adjust to 90 to 110 and in those cases we found no performance loss up next we tested two different dlss mods we tested with up to dlss 3.5 at least according to the dlls and we used frame generation with the DLS SG mod our Baseline performance for this test was about 50 FPS average on the test platform with an RTX 460 scaling obviously improves performance since it's reducing the base render resolution so it's no surpris that 79% render scale brought us up by 18% but the whole point of dlss and FSR is that you can render the scale lower than 100% and you're supposed to be able to get at least comparable quality to previously or far better than it used to be when we didn't have any type of upscaling or other super resolution super sampling Technologies Dro into 70% scale but with dlss improved us further now it's a 62 FPS average and an increase over Baseline of 24% it's a lot removing that mod and switching instead to the dssg mod with frame generation we saw an immediate increase of about 30 to 40 FPS average it's not necessarily because any one mod implements it better it's because of frame generation the 70% SC 100 FPS average is around our CPU limit on this test platform well for the one specifically we're using for dlss that is and judging against the 79% increase we saw an improvement of 57% over the test without frame generation at 58 FPS average or 85% over Baseline at 100% scale if you're using an Nvidia card these mods may be a place to look up next is the optimization mod or at least the highest used one at the time we checked this one changes the actual implications of the named Gra graic settings so ultra high medium and low now carry a different set of parameters that means the words themselves aren't like for like but that's the point of this quick test the mod author on the page noted that quote it slightly tweaks some of the graphical settings like Shadows Reflections grass NPC spawning non-baked light counts and vrs setting the blue bar is the Bas line the red bar is the modded result and this is just for average FPS for the results we found that Ultra improved the most it moved from 44 fps to 51 and increased increase of 16% the medium results posted a 12% Improvement low barely moved it's likely that this approach will get more refined as modders have time to really figure out the game if you need more performance than you can get from these settings natively or you want finer tuning between Ultra and high or other settings this would be a good starting place you could use this as a foundation and then tune it manually further for what you want there are also variations of this II modding approach that increase graphics quality of ultra if you want it to go the other direction so this has been fun and we have a lot more we could do with the game of course really enjoyed working on it uh despite whatever issues Starfield may have just from a purely technical standpoint and a Content standpoint this is fun work because we get the visual comparisons and we get a lot of benchmarking and that's the fun part for us is working through that so what we walk away with here uh a couple of these settings are particularly impactful so Shadow Quality volumetric Lighting can be indirect lighting is an interesting one where in most cases it doesn't appear to do that much but then with low and vrs off it looks awful it like it ruins the visuals of a lot of places anywhere with a straight line in the game looks terrible you won't notice that if you have vrs on for the most part from what we've seen uh and probably most people will especially on newer cards if you're on a card that doesn't support vrs or for some reason you don't want to use variable rate shading and it looks bad make sure you're not on low indirect lighting move it up to at least medium that will resolve your issue immediately it it it looks so bad we're still curious whether it might be broken at that setting and that wouldn't surprise us with h with Bethesda launches so contact shadows and volumetric Lighting on the other hand aren't broken but they don't show serious Improvement in quality above high for volumetrics and medium for contact Shadows with volumetrics it's a matter of opinion whether the higher settings actually do look better of course gtao Shadow quality and Reflections all show clear differences at each quality level and they all look their best when maxed out so these are settings to focus on when balance ing visuals and performance especially Shadow quality and particle quality from what we've seen with other uh reviewers and with users online also we just struggled to find many differences with we saw a couple places we gave you one example earlier but it doesn't really seem to impact performance at least where we've tested it now maybe there's some massive boss encounter or whatever at some point where it's just particles everywhere all over the place and in that case maybe it's different but uh anyway the last couple we thought cra density you should definitely max out if possible it doesn't have the biggest impact on performance in most instances and it really influences how alive the game feels crowd density is particularly interesting because you need to reload the NPCs regenerate them so the save from what we've seen appears to save the crowd density from the last settings which means if you set it to high you Zone somewhere you save an exit you come back with it at low it's still going to show the Den that you saw last time at least from what we've seen so you need to actually load a different area of the game or run around a bit to reload those NPCs to really see that setting take effect and then grass quality also directly affects the quantities of of grass and the view distance of it so if you're doing a lot of exploring we'd recommend trying to drag that one higher but it does have a relatively high performance impact on the GPU and your GPU limited engine a grassy field that's enough summarizing though if you want to get the full detail obviously just go back through the video this is built to be a long-term reference for what do these things do uh we are not experts in individually tuning these settings like Shadow Maps or Draw distances or things like that it's all very specific for the inis and these types of Bethesda games uh but there are a lot of experts out there and we highly recommend you check out some mods on Nexus mods as they develop give the modders some time it take a couple months for them to really work through things and figure out how the game behaves but uh we've given you some Basics and some stuff to look at initially and hopefully get the game running as well as it can for what it is right now today and otherwise we'll be back with some review content and Hardware pretty soon so check back for that subscribe for more as always if you want to support this type of work go over to store. Gamers access.net and grab one of our brand new gn15 mod mats we just launched them and they give you a large PC Building work surface with a lot of diagrams screw tracking and other features to make PC Building easier thanks for watching we'll see you all next timethis massive Graphics guide for Starfield looks at the mod packs that are out now including some dlss and an Optimizer mod we'll also be looking at Graphics comparisons A to B and we're doing a ton of benchmarks across different graphic settings presets and explaining a deeper level of optimization you can do Beyond just changing the settings word in the Starfield menu and this is our final installation of our Starfield coverage for right now at least we'll come back to this topic Starfield will become a permanent test in our suite we've decided and as more mods come out and patches to the game come out we'll revisit of course whenever those happen but for now this is the last immediately planned piece and we're excited for it because uh for those of you who've been around a while in 2015 when GTA 5 and The Witcher 3 came out our some of our original sort of uh claim to internet success was GTA 5 optimization guides and benchmarking and The Witcher 3 Graphics benchmarking uh and I loved working on those and we're back now with Starfield so let's get started before that this video is brought to you by the thermal T swappable blade fans available in 120 and 140 mm sizes the new thermal T fans include three sets of swappable blades so that even as you change builds or cases you can ensure the LEDs are always presented on their best side the swappable blades allow Builders to get the fan frame out of the way of the lights by reversing the blade direction to reconfigure the fan as push or pull while keeping the struts relatively hidden and keeping the fan frame oriented One Way swapping blades is done by applying pressure evenly to opposite sides then pressing until the click each fan also has Pinto pad connections for cess daisy chaining and you can learn more at the link in the description below okay so the way this works is we're going to look at each of the settings individually for the most part we will be comparing them not only visually but with a number of benchmarks across a couple different platforms those include CPU bound constraints and GPU bound constraints and additionally we'll be testing the graphic settings and tuning them from two directions one of them is from a baseline of 1080p low meaning we are adjusting the setting under test up into higher uh demand and the other one is from a baseline 1080p High meaning we're adjusting the setting down so you can sort of view it like 1080p low Baseline adjusting settings up is you are at the the lower end of performance capabilities with your hardware and you're almost shopping for settings what can I add to my cart to get the most visual impact with the least performance impact or meet some kind of trade in between and then likewise from 10 AP High Baseline you're more looking for I'm pretty close to the performance I want what can I take away to just get over whatever that threshold is that you're targeting for your performance so that's how we're going to be looking at this that involves a lot of testing we tested in a couple different areas as well uh and we're looking at some mods optimizing starfields Graphics can go a lot deeper than just configuring the menu to those settings that means you could if you wanted to get into fine-tuning parameters like maximum decals the draw distance per setting like grass mesh coling the radius of things such as ambient occlusion the shadow map count the particle count and quality resolution scaling and more these individual settings have already been tuned by one of the mods if you don't feel like you understand how to change them to get the Maximum Impact and we'll be looking at that today additionally we're benchmarking those from the default configuration and we're testing the game for an optimization mod we'll also be testing two dlss mods and we're adjusting fov for just a quick test there as well one quick thing right at the top of this too as we were finalizing this video Bethesda posted a note that it's working on official dlss integration an fov slider HDR support and Gamma adjustments these are not integrated yet but Bethesda did post a small patch none of these features though were included the patch claimed Improvement to Performance generally speaking without being specific at all we just ran two quick validation tests and we found that the RX 7600 had no change to Performance when tested at 1080p high or 1080p low in our test conditions the RTX 4060 saw a 6% performance uplift in one test condition but upon running Nvidia profile inspector we found that this is because Nvidia pushed a patch that added rebar profiles the 6% uplift could be a mix of the rebar profile with whatever Bethesda changed or it's just from the rebar profile alone either way as of right now there are no major changes from the patch it's mostly to fix a few Quest breaking bugs we'll go back to the rest of the content now especially since the patches in the future won't affect most of the graphics comparison any way in terms of visual changes or optimization relative values and they will affect of course absolute values but we'll test those as they come out we're going to give you a quick summary chart upfront of our findings to try and provide the most helpful information as fast as possible couple things some of these matters will be subjective so our summary chart looks at the performance impact and the visual impact visual impact is a matter of opinion uh and we have a couple people on the team who contributed opinions to that and we've kind of averaged them the chart is going to be for your quick reference the content today is relatively in-depth and it aims to provide a detailed understanding of the settings and the benchmarks and hopefully this simplification is helpful so here's the table the scale ranks from none to critical performance impact is defined objectively and from our GPU bound test scenario however CPU bound scenarios are explored later in this video the performance impact rankings are objectively defined by the ranges at the bottom of the table some have caveats like grass and crowd density and indirect lighting is also a particularly interesting one for visual impact most of the time it's not too important but there is one specific condition that really ruins the visuals for us and if you're playing with indirect lighting at low without variable rate shading like if you have an older GPU that might be what's making the game look bad for you we're providing the summary chart because we don't want to offer flat recommended settings and that's because they would change based on the hardware and we can't recommend flat settings that will maximize everyone's Hardware so rather than giving a a sort of safe configuration uh that you'll probably be able to run what we're hoping to do is provide the information so that you can then go do the tuning yourself to get the performance that you need based on your Hardware so uh there's be a little bit of work on the users end interpreting that but we've made it as simple as we can while giving you the the the most to work with as an example though a higher-end machine might only focus on the items we've marked as critical performance for extra frame rate while a low-end machine might focus on say moderate to critical or try to balance the preservation of visual impact high and critical items while reducing the quality of other things either way the testing we're doing today is relative scaling from impact of settings which means that for the most part this will all remain basically a hopefully a permanent reference unless Bethesda makes a a very specific tweak to how some of these settings behave uh also a lot of our testing was done with an RX 7600 we have 49s in here as well but that comprises the the bulk of the numbers we're looking at today and additionally uh unless we're talking absolute numbers driver updates aren't going to change the relative scaling of individual settings for the most part so this should be fairly useful long term as well even if you're watching this in a year or so just know that the absolute frame rate numbers might change okay okay enough of the summary time to get into showing our work for what made the summary first of all for this we are looking for any change greater than a couple percent to investigate into greater detail so if we see a Swain of say 6% moving from low to ultra we'll look at the invidual ones in between as well uh we'll start with our new Atlantis testing then we deviate from that for a couple specific areas like a field for grass quality for example uh further all capture was done on an RTX 490 the video capture was an 4K borderless which is uh sort of the the closest Native thing to full screen without modifying the game that's in vanilla right now and our Baseline capture was done with all settings manually maxed out uh with the exception of a couple those are vsync and dynamic resolution both of which were off we also disabled all forms of upscaling for the capture and we disabled variable rate shading for the capture now capture is not benchmarking so the capture we did to show the comparisons was made specifically for visual comparison and has no bearing whatsoever on the performance benchmarking that we're doing later moving to GPU bound benchmarks with the RX 7600 we'll start with Baseline low testing and adjust each setting up individually we're starting in our new Atlantis setting for this first we'll just dump all the test results on the screen and then we'll split them up accordingly and look at some different environments here's everything look at this graph kind of like a shopping cart each item you add costs performance and we're looking at the cost on a per item basis so they're not cumulative here but obviously as you add more of these high demand things the performance will get worse some settings like grass quality don't have any impact in this test area but they do later we'll split them out separately for testing in a field others like motion blur just don't affect performance in any meaningful way in any location that we tested just using this as a quick glance though we saw the biggest impact from Shadow quality we're changing literally only that setting to ultra cost us a staggering 18% of our total possible per perance this setup this is the first setting to adjust if your GPU bound highlighting volumetric lighting as the next one or maybe Reflections as the one after that and then indirect lighting indirect lighting here has a potentially large impact on visuals when using low specifically with no variable rate shading we'll come back to that but it looks really bad jumping quickly to the GPU bind from the other direction with a 1080p High Bas line we get this chart we've intentionally sorted this chart with the Baseline at the top despite it not being the best setting this is the opposite direction from before we start at high settings not Ultra and then we reduce them as we go that restricts the maximum top to bottom swing of course but it attempts to look at this more holistically a lot of people probably try to start with the higher overall settings then individually reduce them which is what this shows reducing volumetric lighting had the largest Improvement going from high to low ignoring ultra settings and with medium volumetric we got the most performance without that much sacrifice at least for this setting Reflections and Shadow quality were also High on the list as was indirect lighting note that several items like everything in the 55 to 56 FPS range are with an error of basine in other words changing them had no impact on performance in this test scenario we can see some changes in other areas but not always let's move to individual charts Shadow quality has particularly high cost at Ultra when tested with our GPU bind configuration medium and high are barely differentiated here with them respectively costing 4% and 7 1.2% of the performance against Baseline that is 4% is still a relatively large change considering it's just one setting but Ultra Remains the obvious choice to avoid for most setups if you're on a higher-end set of hardware and you're bordering on handling the ultra preset this would be the one setting to drop to try and boost it to more satisfactory performance while still maintaining Ultra overall looking at Shadow quality from a high Baseline rather than low we saw 5 1 12% Improvement switching to low Shadow quality which materializes as about 3 FPS average this test moving to medium next gave us a 2 and 1/2% uplift Ultra was obviously much more expensive here let's take a look at what shadow quality actually does though to see if it's worth it the most obvious results of the Shadow quality setting can be seen in the edges of Shadows which generally become softer as the setting is lowered from Ultra to low fine details like the shadows of chair legs are blurred out at low as are the shadows of individual leaves at the bottom right of the frame the differences between medium and high are more complex with shadows on the wall at the right of the frame appearing softer at medium but sharper under the table and chair at the center indoors where Shadows are generally closer to the source casting them the scale from low to ultra can be more subtle moving from Ultra to high practically erases the shadow of the coffee table cast by the fire but otherwise there's little difference medium is similar but moving to low then removes much of the Shadow under the portrait cast by the ceiling light Shadow draw distance also scales with quality even just going from Ultra to high in this scene eliminates many Shadows within the tree canopy in the background and each step down and quality brings the edge of that draw distance closer until at low even the Shadows within the trees closest to the camera are almost gone so this really seems to affect the draw distance indirect lighting didn't have as large of a performance impact as some other settings like Shadow quality but it does have a gigantic impact on visuals depending on which setting you use low with variable rate shading off produces some of the worst looking blurring in the game so older gpus that don't support variable rate shading will look particularly bad at low we'd strongly recommend running this minimally at medium for performance though we're seeing about a 5.8% performance loss by going from Baseline low to ultra indirect lighting when GPU bound it's a fairly large change switching to Baseline high and still GPU bound moving from high indirect lighting to low gave us 4.2% more performance but it has a big visual impact at low so let's look at that it's so incredibly ugly at low that we think the setting itself might be bugged at least at the time of writing if you have to play on low settings to get a decent frame rate we'd strongly suggest that you at least consider turning this one up to medium just to prevent the huge reduction in overall image quality at low at least from what we think here indirect lighting shows no meaningful difference between Ultra and medium though in any area that we check then at low it suddenly looks Jagged shimmery and lickery anywhere there's a straight line areas of the screen appear blurry and low resolution it's especially noticeable in the background highrise here and the metal awning over this balcony for what it's worth the effect is strongest in areas that actually are indirectly lit Shadows illuminated by reflected light in other words this scene in a park shows the same effect distant objects like the Mercury Tower sign and Center screen become a watery mess while some other objects are only affect on the edges like the blue pillars in the center screen this really does look terrible in game if you've been running this setting at low or you're running at 1080p for example we highly recommend pushing this up to at least medium it may change your perception of the graphics quality overall and it should be one of the first settings to increase on the list of players using lowquality settings alternatively enabling variable rate shading also solves this issue which makes it seem even more likely that the low Behavior is a bug we did notice that in some scenes the visual impact isn't clearly visible for example we tried to find it in the club in this bottle full of bub and it's not as obvious there's still some difference specifically at the lowest setting though notice the flickery noise and the reflections on the text in circling the stage again moving to medium makes the scene appear effectively identical to ultra or at least close enough that it's a massive upgrade over low time to get into grass quality our previous summer charts showed nearly zero impact from grass quality but that's because there's not much grass in the new Atlantis scene looking at grass quality tested specifically in a field we see the differences emerge more meaningfully in the numbers tested from all low settings at 1080P and only adjusting grass without upscaling the Baseline result was 86 FPS average with good lows this field doesn't put load on the GPU in the same way that the city does but the grass quality clearly shows an impact overall increasing quality to medium results in a loss of 2. 4% performance while increasing to high cost us a total from low of 6.9% ultra cost us 11% from Baseline low that's a massive swing considering this is just one quality setting and once again Ultra proves to be sort of that last jump that really gets overbearing for a lot of gpus if you're in a GPU bound scenario and you particularly notice issues in grassy areas this would be the setting to adjust just know that it won't help performance issues in areas without much literal grass like cities where there's nowhere near the grass density if there's grass at all crowd density is our next setting to test this setting has a high and observable visual impact but a low impact to Performance we'd strongly recommend keeping this as high as you can to keep the game feeling more alive in a GPU bound setting from a low Baseline we saw a cost of around 2% to increase crowd density to high it wasn't worth testing anything in between since obviously it'd be less than 2% let's look at the visuals crowd density is a weird one our method ology for this piece was to change a setting exit the game open the game load our save and start recording in this process we noticed that Starfield is weirdly consistent in a good way about loading saves that extends to the positions and the paths of the NPCs like this NPC that always walks on top of the couch at least the path finding is consistently buggy but that explains why we didn't see any real difference in the number of NPCs between low and high c density we need to load into a new area and generate a fresh batch of NPCs straight from the Starfield NPC Bakery but that also breaks the side by-side comparison we're trying to do by loading into the Mast District we can get an extremely rough picture of crowd density we captured a couple passes at each setting then synced them up paused and counted the visible NPCs in that single frame with that small sample size we saw 22 NPC C's on average at high 17 at medium and 13 at low we wouldn't base any performance calculations on those numbers but it's enough to confirm that the setting does do something now for Reflections this is another one with a relatively high performance cost for a singular setting depending on how bound you are on the GPU starting with the low Baseline numbers on the 7600 we saw a maximum cost of 7.2% moving to high Reflections or about 2.2 with medium from the other direction tested at a high Baseline we saw an improvement of 5 . 7% by dropping Reflections to low or about 3% dropping it to medium even with everything else loading the GPU heavily Reflections quality remains one of the first settings you could go for to get a performance uplift maybe after Shadow quality and here's how the reflections actually look at the highest level mirrored surfaces show a fairly sharp but non-r traced reflection of the surrounding space while lowering the setting to medium leads to more streaky abstract Reflections even at Max objects like furniture and NPCs aren't reflected at all here moving to the lowest setting wipes away large details and also appears brighter in this instance although we ensur that each individual set of Clips was done at precisely the same time of day so as to minimize any impact there note that the real pillar and floor visible at the right edge of the screen are lit the same at all settings only their Reflections change there's some slight changes in non-mirrored surfaces as well but there's not a clear better or worse in this regard this outdoor Lake scene illustrates the same point the surface of the lake and the exterior of the curved building are brighter with Reflections at high but medium and low are effectively identical even when considering the reflections of scenery on the water surface like some of the other settings in the game it appears as if the biggest change is at the highest quality option at least visually volumetric lighting benchmarks are next and then the comparisons when GP bound and from Baseline low settings we see a performance loss maximally of 12% this setting has a large impact on performance high and Ultra costs about the same medium costs about 3.3% from the Baseline performance tested from a high Baseline instead we saw a 9% performance uplift moving from high to low or about a 6.6% uplift going to medium and a slight loss going to ultra even when more heavily constrained in general this setting still has a large performance cost time to talk about how it looks though the Smoky streets of neon seemed like the ideal place to check for volumetric lighting but they weren't it's not the right kind of smoke we'll use that scene to check some other settings though instead let's look at a facility on Mars Ultra and high looks similar here bringing up the brightness and the sharpness of the fog shrouded lights at the center screen but dropping to medium and low progressively softens the effect with low eliminating some of the individual bright points altogether we also noticed flickering artifacts around one of the lights at all settings gtao testing is next for this ambient occlusion testing we found that a GPU constraint from 1080p low Baseline resulted in a 4% performance loss by changing to ultra for the setting the range for this one is overall low considering the relatively low maximum cost from high Baseline we found no change going to low the performance was within error the lodge is an ideal place to check on ambient occlusion because has lots of small objects cluttering the scene this setting performed completely as expected at Ultra the crevices behind things like the cushions for example are darkened as are areas around items on the coffee table and pieces of furniture the effect is lessened noticeably with each step down to high medium and low with low showing brightly lit areas that should definitely be in a shadow like inside the arms of the couch the effects aren't always so pronounced though although the setting is still working as intended this brightly lit office on Titan mostly shows ambient occlusion around the base of the desk the complicated geometry of the office chair also benefits here with higher settings significantly darkening the areas underneath and inside the chair on the contact Shadows from 1080p low Baseline we observed a 3.6% performance cost hiking to ultra and settling in between would yield most of the visuals from Baseline High settings reducing contact Shadows to its lowest setting gave us 1.4% more performance there's not much worth considering here when other options give us far more range the most obvious instance of contact Shadows that we could find were around NPCs and bright sunlight in this regard ultra high and medium look pretty much identical while shifting to low drops the resolution of the shadow where vasco's legs for example contact the ground and it eliminates some detail in the shadows of the arms and the legs elsewhere the differences are even harder to notice in this outdoor scene the joints and the honeycomb roof structure become darker when the setting is raised above low but changes beyond that are minor most of the NPCs are standing in Shadow so even ones that pass close to the camera like the security guard are unaffected by the setting and the ones in the distance are too far away to see clearly anyway for particle quality benchmarking we found no differences from our Baseline low settings by increasing this setting we even checked in some areas that seemed to more explicitly show particles but found no differences the setting has described as setting quote the quality of particle lighting so thanks Bethesda that really helps we tried examining three specific things closer Sparks from a mining laser Smoke backl by Neon Signs and Sparkle effects on the artifact in a script sequence starting with the mining laser the Sparks generated by firing it have no noticeable effect that we could see on any of the surrounding Shadows at any setting the laser itself does illuminate its surroundings but there's no flickering or variability to suggest that the Sparks themselves are lit there's also not a difference in the quantity of particles generated at any of the settings that we could notice since we didn't see particles lighting the scene we hoped to see the scene lighting particles by moving to the smoke and neon the differences are almost invisible but but we noticed that the cables in the support to the right of the Neon security sign are obscured at the lowest setting accompanied by some shimmering medium and high still look identical in this area though finally in the animated sequence that starts after bringing the artifact to the lodge there are visible differences in lighting and shadows between recordings but flickering and intensity of the light appears to be randomized we can't make a conclusion based on this scene time to briefly look at some benchmarks from the CPU limited angle for this we chose an R5 3600 with an RTX 490 to create a CPU bound test condition overall we found most settings were unable to affect the performance the best shot was Shadow quality again even though it's a GPU heavy setting it's still affecting performance when mostly CPU bound bottlenecks are really so simple as being one component or the other so this makes sense crowd density also affected performance but not much we saw 3.6% performance loss from Baseline going to high for crowd density otherwise everything was within 1% chance or 1.2 to 1.3 for Reflections and contact Shadows we tested these from Baseline High also but stopped because we saw no change time to talk about mods a couple weeks into launch and Nexus mods unaffiliated with us but a great website is already filled with tons of options the modding scene for Starfield is still developing and the modding aspect of Bethesda games is probably the most exciting part so even though all of this stuff is going to get much more sophisticated we downloaded and tested a couple of the early mods each independent from the other the mods we tested include Starfield performance optimizations by Eck Starfield upscaler for dlss by pure dark Starfield frame Generation by Luke FZ 564 and we did a manual fov adjustment in the inii files we'll do the simplest Benchmark first field of view or fov has been a common complaint since the game launched it can be changed without modifying the game other than creating a custom inii file similar to Skyrim but currently at least there's no in-game slider increasing fov widens the peripheral vision of the player's camera we ran some quick tests from the default fov up to a 150 just to see how it did between all of these performance was the same 150 looked awful obviously but it was just there for a test we tried up to 360 uh it didn't work well and most people might adjust to 90 to 110 and in those cases we found no performance loss up next we tested two different dlss mods we tested with up to dlss 3.5 at least according to the dlls and we used frame generation with the DLS SG mod our Baseline performance for this test was about 50 FPS average on the test platform with an RTX 460 scaling obviously improves performance since it's reducing the base render resolution so it's no surpris that 79% render scale brought us up by 18% but the whole point of dlss and FSR is that you can render the scale lower than 100% and you're supposed to be able to get at least comparable quality to previously or far better than it used to be when we didn't have any type of upscaling or other super resolution super sampling Technologies Dro into 70% scale but with dlss improved us further now it's a 62 FPS average and an increase over Baseline of 24% it's a lot removing that mod and switching instead to the dssg mod with frame generation we saw an immediate increase of about 30 to 40 FPS average it's not necessarily because any one mod implements it better it's because of frame generation the 70% SC 100 FPS average is around our CPU limit on this test platform well for the one specifically we're using for dlss that is and judging against the 79% increase we saw an improvement of 57% over the test without frame generation at 58 FPS average or 85% over Baseline at 100% scale if you're using an Nvidia card these mods may be a place to look up next is the optimization mod or at least the highest used one at the time we checked this one changes the actual implications of the named Gra graic settings so ultra high medium and low now carry a different set of parameters that means the words themselves aren't like for like but that's the point of this quick test the mod author on the page noted that quote it slightly tweaks some of the graphical settings like Shadows Reflections grass NPC spawning non-baked light counts and vrs setting the blue bar is the Bas line the red bar is the modded result and this is just for average FPS for the results we found that Ultra improved the most it moved from 44 fps to 51 and increased increase of 16% the medium results posted a 12% Improvement low barely moved it's likely that this approach will get more refined as modders have time to really figure out the game if you need more performance than you can get from these settings natively or you want finer tuning between Ultra and high or other settings this would be a good starting place you could use this as a foundation and then tune it manually further for what you want there are also variations of this II modding approach that increase graphics quality of ultra if you want it to go the other direction so this has been fun and we have a lot more we could do with the game of course really enjoyed working on it uh despite whatever issues Starfield may have just from a purely technical standpoint and a Content standpoint this is fun work because we get the visual comparisons and we get a lot of benchmarking and that's the fun part for us is working through that so what we walk away with here uh a couple of these settings are particularly impactful so Shadow Quality volumetric Lighting can be indirect lighting is an interesting one where in most cases it doesn't appear to do that much but then with low and vrs off it looks awful it like it ruins the visuals of a lot of places anywhere with a straight line in the game looks terrible you won't notice that if you have vrs on for the most part from what we've seen uh and probably most people will especially on newer cards if you're on a card that doesn't support vrs or for some reason you don't want to use variable rate shading and it looks bad make sure you're not on low indirect lighting move it up to at least medium that will resolve your issue immediately it it it looks so bad we're still curious whether it might be broken at that setting and that wouldn't surprise us with h with Bethesda launches so contact shadows and volumetric Lighting on the other hand aren't broken but they don't show serious Improvement in quality above high for volumetrics and medium for contact Shadows with volumetrics it's a matter of opinion whether the higher settings actually do look better of course gtao Shadow quality and Reflections all show clear differences at each quality level and they all look their best when maxed out so these are settings to focus on when balance ing visuals and performance especially Shadow quality and particle quality from what we've seen with other uh reviewers and with users online also we just struggled to find many differences with we saw a couple places we gave you one example earlier but it doesn't really seem to impact performance at least where we've tested it now maybe there's some massive boss encounter or whatever at some point where it's just particles everywhere all over the place and in that case maybe it's different but uh anyway the last couple we thought cra density you should definitely max out if possible it doesn't have the biggest impact on performance in most instances and it really influences how alive the game feels crowd density is particularly interesting because you need to reload the NPCs regenerate them so the save from what we've seen appears to save the crowd density from the last settings which means if you set it to high you Zone somewhere you save an exit you come back with it at low it's still going to show the Den that you saw last time at least from what we've seen so you need to actually load a different area of the game or run around a bit to reload those NPCs to really see that setting take effect and then grass quality also directly affects the quantities of of grass and the view distance of it so if you're doing a lot of exploring we'd recommend trying to drag that one higher but it does have a relatively high performance impact on the GPU and your GPU limited engine a grassy field that's enough summarizing though if you want to get the full detail obviously just go back through the video this is built to be a long-term reference for what do these things do uh we are not experts in individually tuning these settings like Shadow Maps or Draw distances or things like that it's all very specific for the inis and these types of Bethesda games uh but there are a lot of experts out there and we highly recommend you check out some mods on Nexus mods as they develop give the modders some time it take a couple months for them to really work through things and figure out how the game behaves but uh we've given you some Basics and some stuff to look at initially and hopefully get the game running as well as it can for what it is right now today and otherwise we'll be back with some review content and Hardware pretty soon so check back for that subscribe for more as always if you want to support this type of work go over to store. Gamers access.net and grab one of our brand new gn15 mod mats we just launched them and they give you a large PC Building work surface with a lot of diagrams screw tracking and other features to make PC Building easier thanks for watching we'll see you all next time\n"