it might be the case that some setting changes don't show any difference until you restart the game, specially if you are doing a lot of tinkering.
Advanced Questions on Graphic Settings
- Thread starter Exigeous
- Start date
it might be the case that some setting changes don't show any difference until you restart the game, specially if you are doing a lot of tinkering.
I'm probably the wrong person to ask, since I prefer to remove effects like DOF, bloom, lens flare, light cones, etc. for a more minimalistic graphics presentation. As I've said before, my Rift S gives me all the bloom and lens flare I could ever ask for (and then some), along with godrays and DOF
As for terrain settings, I usually max these out unless I encounter a heavy loss in FPS. With these settings, more is visually and obviously better. When I do start losing FPS, it's usually because of too many ultra-quality shadows. Here's a little tip - turn off your SRV lights in the daytime for a boost in performance, because dynamic shadows cast by SRV (even in sunlight) cause a performance hit. At night in VR where I am dependent on ship and SRV lights down at a planet surface POI, I sometimes will turn off shadows completely so I don't have a vomit-inducing slideshow in VR.
That's all I got
That is an excellent observation znôrt! Good catch!
Just a small addition. LOD blending describes both the blending between assets with different polygon count and textures of different resolution as well. While both options primarily save performance, textures with lower resolution also prevent aliasing.
For some settings to apply, you need to reload the star system. Maybe this goes for this on, too.
Edit: https://community.pcgamingwiki.com/gallery/image/1041-comparison-material-quality-ultra/
Last edited:
Deleted member 182079
D
Probably not the most useful post, but I think Material quality affects stuff like mining fragments, or mats from geo/bio sites, and their related assets. I even think that textures of asteroids are affected by this.
As for the Volumetric Lighting, I think that @Craith is correct - I think it's usually called god rays, i.e. the light of the star shining through the fog in the asteroid belts and casting shadows on each other - the light beams in High Tech stations would also be affected by this I reckon. I didn't see this effect in the rings until I upgraded from my laptop to my current desktop PC - I can now run everything on Ultra with 1.5 super sampling, and I was gob-smacked how good the game can look, despite its age.
As for the Volumetric Lighting, I think that @Craith is correct - I think it's usually called god rays, i.e. the light of the star shining through the fog in the asteroid belts and casting shadows on each other - the light beams in High Tech stations would also be affected by this I reckon. I didn't see this effect in the rings until I upgraded from my laptop to my current desktop PC - I can now run everything on Ultra with 1.5 super sampling, and I was gob-smacked how good the game can look, despite its age.
@Exigeous - Can't answer your questions directly, but have you seen these threads?
I suspect @Morbad would have more knowledge than most on what your queried graphics settings do.
Hardware & Technical - Some provisional tweaks for better performance vs. IQ in Elite: Dangerous on higher-end setups
ED's default graphics settings and Ultra preset are pretty solid, for the most part, but there is definitely still room for improvement. In particular, I find that the lighting relies too much on bloom instead of tonemapping; there are too many instances of 'Peter Panning' shadows; texture...
forums.frontier.co.uk
Hardware & Technical - Custom Elite Dangerous display quality settings continued...
Since I've started playing ED again actively after several months off, I decided to revisit the tuning I used as a basis for my previous thread: https://forums.frontier.co.uk/showthread.php/400235-Some-provisional-tweaks-for-better-performance-vs-IQ-in-Elite-Dangerous-on-higher-end-setups To...
forums.frontier.co.uk
I suspect @Morbad would have more knowledge than most on what your queried graphics settings do.
Yeah, there's been a bug for a long time that if you set it to anything but Ultra it adds that to the edges - as for the framerate changes I'm not seeing that at all, and visually I see absolutely no difference. Here are all 4 versions in the same place. I even double checked Frontier Flopping to the main menu to make sure the changes were taking effect (as a few require this) and still I see absolutely no difference (well slight given everything updating from going to the settings but that's all)
Do you see anything I'm missing? I did get reading glasses recently so maybe I'm just blind
Icy Rings - Volumeteric differencs
Source: https://imgur.com/a/uYU7Xsu
Yep, it is, wasn't sure where to put it.
You're clearly not understanding my goal - I'm analyzing the in game settings, nothing external, nothing "modified". So if you want to understand what each setting does visually and in terms of framerate that's what I'm doing.
DING DING DING DING - exactly this.
Huh? What does that have to do with changing the settings in game? ALL PC's would render the same image, only the framerate would differ. I think you're not following something here...
Yep, this is absolutely true for some settings but for all of these I have restarted both to main menu and fully to desktop.
Yep, that's why I've done this many many times.
Yep, should have been clear to begin with as I've tested that many, many times. I did finally find Environment Quality last night - it's the skybox. It's just really hard to see the differences in the bubble since the sky is pretty basic. When I was out in Colonia we realized that's what it is
I haven't been to a Lagrange cloud yet to check those assets but I can't see any visual difference in the others you name. Logical thought though.
Have a look above, I think it's pretty clear that isn't the case (or I'm more blind than I think)
Yes, these cover the very granular settings you can change with the XML files and while these are very helpful for my next video (specifically about these, the xcfg files, customizing the presets, etc) it doesn't really help with my current project.
Thanks for all the input guys but I'm still pretty stuck with a few. I'm going to reach out to the devs but I doubt I'll get any answers from them...
~Exigeous
Looks like what effects this is in fact shadows - here are the differences in only Ultra Shadows to No Shadows
Shadows - Ultra
Source: https://imgur.com/31M5MCY
Shadows - Off
Source: https://imgur.com/IH7tpuM
Pretty clear that's what's effecting the "rays", as I understand it from other threads there are no "god rays" in Elite at all.
Thanks but that's affected by "FX Quality", not volumetric. All the interior station lighting it (well that and bloom). Set to to off and they completely disappear, with the varying levels showing varying degrees of fog, etc.
DING DING DING - that's it, from all I can tell it only effects the camera suite. When turned off you can't focus/blur at all (if that's set and you turn this off the effect goes away) and the difference in medium/high is the quality of what's rendered under the blur. Really silly that they'd put this setting here with no explanation like that. Ugh.
Big thanks for pointing me in the right direction. Now if I can only find volumetric... Not rings, not lagrange clouds, not station lighting, not not not.
DoF does only affect the third person views (HoloMe and vanity camera stuff), but it otherwise straightforward. Performance hit is significant where it's present.
Material quality is a bit vague, but seems to affect the LODs of geometry, specular lighting, and normal maps for certain objects. Not sure if it affects object textures.
Volumetrics is how light is scattered by fog and similar effects. It's not the same as the light cones or fog settings that are control via "FX", but subtly influences the appearance of them. I think about the only place I could notice a performance difference was while doing donuts in the SRV in reverse with the lamps on, or the super thick fog left over from core mining. It's not the same as fog shadows either (which, as previously noted, is controlled by the shadow settings), but might change the appearance of the light rays between the shadowed areas...been a while since I've messed with it.
Terrain LOD blending reduces pop-in by blending LODs together. Performance impact is negligible and the effect is subtle except during planetary approach.
Terrain Work slider also has almost no performance impact on decent GPU hardware and does seem to reduce pop-in at higher settings for me, most noticeably when moving rapidly over planet surfaces at low altitude. Almost impossible to benchmark on the GPUs I've tried without recording frametimes and looking for stutter. Would also use a controlled test area, like the SRV demo. I normally run 85-95% on the slider.
Material quality is a bit vague, but seems to affect the LODs of geometry, specular lighting, and normal maps for certain objects. Not sure if it affects object textures.
Volumetrics is how light is scattered by fog and similar effects. It's not the same as the light cones or fog settings that are control via "FX", but subtly influences the appearance of them. I think about the only place I could notice a performance difference was while doing donuts in the SRV in reverse with the lamps on, or the super thick fog left over from core mining. It's not the same as fog shadows either (which, as previously noted, is controlled by the shadow settings), but might change the appearance of the light rays between the shadowed areas...been a while since I've messed with it.
Terrain LOD blending reduces pop-in by blending LODs together. Performance impact is negligible and the effect is subtle except during planetary approach.
Terrain Work slider also has almost no performance impact on decent GPU hardware and does seem to reduce pop-in at higher settings for me, most noticeably when moving rapidly over planet surfaces at low altitude. Almost impossible to benchmark on the GPUs I've tried without recording frametimes and looking for stutter. Would also use a controlled test area, like the SRV demo. I normally run 85-95% on the slider.
Deleted member 182079
D
I just wish developers would explain better what each setting does, instead of us having to second guess and trial and error things, at least for the stuff that isn't named based on terms that are usually a known variable (like AA options, or AO).
Some descriptions in the game I never heard of are almost verbatim to the names of the option ("X" = "changes X").
Not just a problem in Elite; at least in Elite though I can run everything at max settings now, even though certain situations can result in some temporary frame drops - namely switching on the light while mining very close to the asteroid, the dust kicked up by the SRV, and core explosions. Working out what setting is to blame was previously quite a PITA when I ran the game on inferior hardware.
Some descriptions in the game I never heard of are almost verbatim to the names of the option ("X" = "changes X").
Not just a problem in Elite; at least in Elite though I can run everything at max settings now, even though certain situations can result in some temporary frame drops - namely switching on the light while mining very close to the asteroid, the dust kicked up by the SRV, and core explosions. Working out what setting is to blame was previously quite a PITA when I ran the game on inferior hardware.
For VR the impact of the Terrain Work slider can be significant. Besides that I agree, it is important to find a controlled area, that is not affected by anything other than CPU and GPU Performance. Since we have seen network connection can have a serious impact on frames rendered, it should be ruled out as much as possible.
@Exigeous Just something, that came to my mind. Have you checked the GraphicsConfiguration.xml? It should should show you the values behind each detail setting.
This is not true. Older gfx (eg. dx10) will not render dx11 and so on.
I don't think every computer setup would return the same result with the same settings. Mainly gfx card I'd assume.
Well I just did a total of 9 passes over 3 planets (Ultra, High, Off) and I see absolutely zero framerate difference. I dropped from glide and I jumped to SC (stick over really quickly so I'm basically flying over the terrain but at SC speeds). All exactly the same performance (and near impossible to see the difference that way).
The search continues...
APIs do not necessarily change the look of the rendered images. As for ED, it only supports DX11 anyway for quite some time now.
There differences are so marginal, it would be hard to notice (if at all) even on direct comparison. In the past you could tell cards apart by the way they applied anisotropic filtering, but that was around ten years ago.
Interesting idea, and I'll check a few things but I can already tell you several things it claims are wrong, as it shows volumetric effecting the interior station lights which are only effected by FX Quality. So I think someone created that making assumptions based on how other games/engines do things.
Yep, exactly what I'm seeing now - thanks
Yeah, I still can't find anything to show an A/B/C comparison here. Nowhere.
Definitely no effect on the explosion cloud, tested that tonight extensively.
No crap, obviously that is not in any way what I'm talking about as that's basically like switching games/rendering engines so of course it would look different. What I thought was clear is the difference in Elite from a 2080Ti to a 970 would be the same image with the same settings, only the framerate would differ. Thought that would have been obvious but I guess not.
Well I did find Terrain Quality - it's the rendering of objects off in the distance like hills, mountains, etc. One more down, 4 to go (with Material Quality and Volumetric being the big ones).
I did send a note to Arthur, the lead community manager to see if he'd pass my questions along. Now that hope may be naive....
I did send a note to Arthur, the lead community manager to see if he'd pass my questions along. Now that hope may be naive....