Updated graphics recommendations for WU6

I have seen some really good guides in the past, but after WU5, I am missing some general guidelines so I can balance FPS impact vs value of better graphics.

I know I can do the extensive testing of every setting, but before I dive into that, I am looking for other players recommendations and learnings.

I have a Ryzen 3600 XT and a AMD 5700 XT on a NVMe.
Running the game at 1920*1280.
I would like to achieve 30-60 FPS…

1 Like

Clouds for me make a big impact, getting15-20 fps more when goimg from ultra to high. (With 9900k and 3070•uwqhd]

1 Like

I’ve found that, apart from the other strange effects it causes at the moment, supersampling causes a big hit in FPS, without (for me) any noticeable increase in quality. So keep that rendering down to 100% max.

This is definitely a threat to follow.

In my case (i7 7700 + 3070) I still enjoy the higher FPS count after SU5. Clouds Ultra to High makes a big difference when landing in London today.

I have mostly a mix of High and Ultra in Widescreen 3440p. Getting 30-50 FPS depending on area and height.

i9-10900KF / RTX 3070 / 32 GB RAM @ 3200 SSD / 2560x1440p

Personally I’m fully maxed. I read over on Avsim that oddly enough having the pre-cache of terrain data set to ultra gives you better performance than on low. I personally haven’t tested this but given I’m fully maxed out I’m curious what happens if I drop this and only this to low. I will try this out in a bit and report my findings.

This video came out a week or two after MSFS launched. It shows a side by side of just about all of the graphics settings. It’s long, so grab paper and pencil, plus a snack. The things that make a big difference can be set to high/ultra, while the things that make very little visual impact (and sometimes the biggest frame impact) can be set lower. I’ll paste in my settings list below, with my core specs at the bottom. I just did a test with the drone over Manhattan, and with these settings, I was getting 30+ FPS. It does drop when loading new scenery, and with the drone’s speed maxed out, I was still getting ~25 FPS at max transit speed. If I stop and let everything load, I’m back in the 30’s.

V-Sync > Off
Anti-Aliasing > TAA
Terrain LOD > 100
Off Screen Terrain Pre-caching > High
Terrain Vector Data > Medium
Buildings > Medium
Trees > High
Grass/Bushes > High
Objects LOD > 50
Clouds > Ultra
Texture Resolution > Low
Anisotropic Filtering > 4x
Texuture Supersampling > 4x4
Texture Synthesis > Low
Water Waves > High
Shadow Maps 2048
Terrain Shadowns 128
Contact Shadowns > Ultra
Windshield Effects > High (no glass screen reflections)
Ambient Occlusion > Ultra
Reflections > Ultra
Bloom > On
Depth of Field > Off
Lens Correction > Off
Lens Flare > Off
Glass Refresh Rate > High

Core basics
TUF Gaming FX505DT
AMD Ryzen 5 3550H
NVIDIA GeForce GTX 1650
16.0 GB RAM

4 Likes

OP: We have similar graphics cards (I have a 2070 non-super), and I have a slightly better CPU (Ryzen 5600X), and I can can run everything on Ultra at 2560X1440.

I would suggest starting on the ultra preset and tweak down from there (if needed). However, I have a fixed FPS of 30 and I’m not worried about maximizing FPS. 30 FPS is smooth and easy to obtain on Ultra.

2 Likes

I agree that this will result in higher FPS … but I can only fly with V-sync = ON because otherwise each frame has some disturbing vertical discontinuity, at some changing random place, which my brain gets so attracted to that it totally kills the immersion for me.

2 Likes

Not true for me, cost me about 12% of my framerate or so, at least in some cases (many objects to load). Still worth it though, the reduced if not eliminated pop-in and stutter is sooo nice.

Interesting, where did you get this percentage? Would like to see myself.

Nice way to describe screen tearing :upside_down_face:

I’ve removed increased LOD to 3 and just used standard Ultra preset and its running and looking great.

Sometimes the less you mess around with the better

1 Like

I used this before, but I think WU5 changed so much, that I wonder if those tests are still valid

1 Like

I have a few standard “benchmark runs” I do with CapFrameX. I use the exact same settings (including date and time); no traffic or MP for repeatability, etc. Fly the same way and engage autopilot ASAP to avoid accidental changes.
So I simply ran these with the low setting and with the ultra setting. From two of the runs I got this:


So at almost every point of the run, I had about 7-10 fps higher with the low setting.
However, there’s no camera motion in this benchmark, and if there had been, there would have been far more stuttering with the low setting. I find it almost unplayable.

1 Like

Awesome, thank you. I will definitely give this a go later!

I think for the most part they are. The core code has been changed to increase overall performance, but individual settings can still have an effect on performance. So I keep the settings that have next to no visual impact dialed back to save resources. Prior to the major code update that gave most people these current frames, Manhattan and London would bring my rig to its knees. Now I get fluid frames most of the time. There are still times that I get a momentary stutter down to about 5 FPS, but those don’t last long, and it’s usually when the sim is having to load scenery at low altitude, and usually on approach. I’m hoping the increased caching ability will help that.

Here’s another vid that OP might find useful: I DOUBLED my FPS in Microsoft Flight Simulator 2020 | Optimization Guide 2021 - YouTube

2 Likes

Same for me. What I learned a few months ago was that since i have a fixed refresh rate monitor, I should target 30 or FPS and turn on vertical refresh.
It turned out that my experience in the game was much better at 30 FPS than f.ex. 50 FPS using these settings.

V-sync must be on for me as well, despite having a G-Sync monitor. Just looks bad with variable refresh.

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.