Smooth Experience with These Frame Rates

Sim Update 5 Beta | MSFS 2024 Virtual Reality
1

This is my first time using SU5 beta testing, and it literally makes more sense to me.
I just flew to LS on a 737 in career mode, and my fps was over 60 in the sky, and at the time of landing it was 55-60, on the ground the fps dropped to 40-45, but it was still very smooth and cool.
And considering that I’m using a Pimax Crystal Super with 1.0 rendering resolution, DFR - 25, ultra graphics settings, including 2048 earth shadows, SSAO - ultra, it’s impressive. (9800x3d|5090|32ddr5)
The only thing is, I turned off the terrain of the earth and I don’t know if it has a big effect on the time and number of frames per second, but I want to believe that it wasn’t an accident. Since yesterday, the same flight with an asterisk in Europe was unsuccessful, the number of frames per second barely exceeded 40, and the graphics processor could not work at full capacity, and perhaps it was the weather, because it was heavily cloudy. I will continue my observations.

General SU5 Beta Feedback Thread
General SU5 Beta Feedback Thread
2

That seems an incredible claim. I am stunned how you get 60fps in the sky. Am i getting this right, you have:

Pimax super

Playing at full native res in that super.

Everything ultra in game (but whats your LOD/TLOD)?

I have the pimax light, so have yet to try DFR (Dynamic foveated rendering). Is this giving you the insane perfromance?

One last question: You say you get 60fps, thats really good, and works well with 120hz + locked half frame rate. But how are you enjoying your flight, if your hitting sub-60fps. It must be horrible. So all your ultra settings are moot, because you cant enjoy them. What am i missing here?

Cheers.

3

I will try to answer your questions with this video. I flew at low altitude and on an airliner. League of Trials is a great option to test your system for performance and compare it with my experience. As for the half frame rate, it doesn’t make sense to me, I could set the frame rate to 45, because Pplay is set to 90fps, but for me there is absolutely no problem when the frame rate jumps from 40 to 72. I don’t see any lags because there are no freezes and staters, and 0.1% fps doesn’t fall below the average frame rate.

2 Likes
4

Dude, that is awesome. Great footage, thanks for sharing. I cant get my head around how in VR pimax in particular, a frame rate other than either 60/72 or 90 locked hard, can produce these visually stunning scenes. Ive literally never heard of this, nor experienced this, in over 5 years of VR use. Im gonna unlock my frame-rate, and see what happens. Baffled tbh. But really pleased for you, that looks really good fun.

5

I agree that video looks surprisingly smooth, but a 2D video (especially on a mobile device, if that’s were you’re watching) is a lot more forgiving than actual VR. If you look closely you can see the stutters. In actual VR it would look a lot worse.

6

Not sure if he’s responded but he has DFR at 25, that’s a massive performance boost so I wouldn’t count that as native res.

2 Likes
7

Yes, thanks for the feedback. Never had the chance to use DFR, but hoping to have it witht he new dream air at some point. What i totally do not understand, is how you would achieve smooth frames with anything below 60/72/90 fps. Seriously, in any game, over 5 years of different headsets, ive never had smooth play with a fps that doesn’t exactly meet my HMD refresh rate. Baffled. The OP stated he never sees a hitch or stutter.

8

29 ml pixels is too much even for a 5090… So I use a DFR of 25%, just like in SU3/SU4, but only now do I get excellent smoothness and a fps boost. In TAA 100% mode, the fps are about the same; I just don’t like the anti-aliasing; the edges of objects are jagged, and the lights flicker at night. Some users claim that the video only looks smooth on a phone in 2D, but this is a misconception. I don’t experience any problems with visual perception, provided I use high enough graphics settings. I could get a fixed 72 fps by sacrificing graphics and cropping the frame edges with opxtk, but I really don’t need that. Important! In racing simulators, I need a stable 90 fps without any jumps, otherwise the gaming experience becomes painful. But after the latest Nvidia driver updates, I re-enabled HAGS in the Windows settings, and it seems that fps jumps are no longer critical for me. Due to a poor experience with SU4, I didn’t play FS24 for two months. I wrote a lot on the forum, because the 737 could only show 40 fps even in the sky. Now I’m seeing 60-72 fps in the sky, which is a complete game-changer.

9

Interesting, thank you. One last question, do you use Quad views enabled in game with DFR? Oh edit: Do you notice the DFR, as in, when looking into the sweet zone, do you notice the fuzz or haze at all? Or is it perfectly optically unnoticeable?

10

I understand this is off track, but do what you said and unlock your frames. I am going to make a separate post in VR. I just discovered this yesterday. 14700k, 5080, PCL. I had PCL set to 80 res, 120hz, locked at half. msfs24 max frame at 60, dynamic settings on and set to 60 target. Smooth flight, but always getting awful terrain popping and always looked like tlod was set really low, even with 55-60 fps at 10, 20, 30k ft. I turned off locked half frame in piplay, turned off dynamic settings, set target/max frame rate to 120, tlod 200, olod 100, turned up/on almost all settings and I got MUCH, MUCH better visuals with no popping or fading in. Going to do more testing tonight, but turning all targets up to 120, not limiting any frames and turning dynamic settings off made a world of difference. I would have thought keeping frames low would give way more gpu overhead to crank settings, but that appears to be the opposite of what is happening. My test airports are third party kdca and kbos. Will try others as I test further. Sorry for going off topic.

11

My TV is only 60hz though

12

What fps are you actually getting before and after? Just turning off dynamic setting was probably enough to fix the TLOD issue. If you’re getting 55-60 fps with a 60 target it’s probably lowering TLOD trying to maintain 60 regardless of all the other stuff you changed.

13

Yes, paradoxically I also get better results turning the dynamic settings off. I am bound to a 60 Hz BENQ monitor that I use for other purposes. I am also experimenting but right now letting it go above 60. PS Also turned off 60 FPS frame limits.

14

What do you mean by better results though, visuals, fps or both? If you have the target too high for your machine with dynamic settings on TLOD will scale back and it will look bad, that would be the expected behavior.

If you are turning dynamic settings off and getting better visuals that isn’t surprising.

If you’re turning it off and getting better visuals and better performance that would be unexpected.

15

From my experience of using Dynamic Settings, verified by seeing when it activates with the FPS view, it only behaves as you think it would for a given target frame rate if you are also setting max frame rate within MSFS. If you do not set a max frame rate, Dynamic Settings seems to activate 5 - 10 FPS before your Dynamic Settings target frame rate, not at it.

I suspect that this is because with max frame rate set, MSFS knows exactly how much headroom it has at any point in time so it can apply Dynamic Settings only when that headroom is about to expire right on the target frame rate. Without such a reference, it has to anticipate and activate when you are slightly above your target frame rate, in order to achieve that target frame rate.

For example, my 2D settings are 40 for both max frame rate and dynamic settings target frame rate, and Dynamic Setting will only ever activate it FPS goes below 40 to try and maintain 40 FPS, which is what I want.

2 Likes
16

So first, my apologies for the weird post. I thought I was in the SU5 Beta discussion, which is why I said i’d post in VR and sorry for going off track.

Ok, so the really interesting thing to me was that I got worse performance (not necessarily worse frames) with most settings off or low, with dynamic settings on and a max frame rate set.

I have a Crystal Light, so everything is in reference to 120hz. BATC also used in all tests injecting FSLTL models and settings default at 5. Flying the Vision Jet.

For reference…14700k @ 5.5 all cores, 48gb DDR5 @ 6800, 5080 MSI Gaming Trio OC’d with Afterburner @ +2500 and +350.

My original settings were Piplay 120 locked at half and res at 80. Piplay res always at 80. That’s my sweet spot.
In sim max frames 60, dynamic setting on and target 40. Almost all settings off or low.
Flying from KDCA to KRIC and KRIC to KDCA. KDCA third party with DC third party scenery.

What sparked me to make the changes I made was the blurry ground even at altitude. 10, 20, 30k. I was getting a solid 55-60 with a dynamic setting target of 40, so why did it appear the tlod was set so low? My dynamic settings target was 40, so I expect at altitude where things are stable for the tlod to be moved up significantly until I am around 40. In my new settings I can set tlod at 400 and still be slightly above 40, so what is dynamic settings doing?

This brings me to another issue. MS gets all bent out of shape and forbids even discussing 3rd party software to manage tlod, olod, etc., but then gives no alternative. Dynamic settings is not an alternative imo. It apparently doesn’t work well, at least not for me, nor do I have any control over what it is doing, or know what it is doing. For example, if I crank up all my settings to see the impact, what is dynamic settings doing to those settings? Dynamic settings has access to everything. Did it just turn down most of my settings? No one knows and that’s why people seek out alternate solutions.

Anyway, I decided to just remove all restrictions in regards to fps and see what happened. It couldn’t be any worse than what looked like 50 tlod from 20k. lol. I turned off the lock at half frames in Piplay. In sim I turned off dynamic settings, and I set max frames to 120. I left all my other settings the same at mostly low or off. I set tlod to 200 and olod to 100.

Starting at KDCA I had fps in the mid 40’s. Great visuals and everything smooth. Also BATC injecting traffic. I turned up or on a bunch of settings. Clouds were always at high and texture resolution on medium. Great flight to KRIC with great visuals. Nice and smooth. I flew from KRIC to KDCA last night with tlod at 250, olod 150. Everything was smooth until I got close to KDCA. I expected this as I was pushing it with tlod at 250, at night, and with significant clouds set to high. I backed off tlod to 200 and flew the rest of the way. fps high 20’s to low 30’s. Some stuttering, but not bad and nothing that effected my landing. 150 tlod would have probably been perfect.

All I really need is to make it so that tlod is 150 under 3000’ or so. Ideally I would have tlod at 300+ above 10k and olod at lowest setting. There are obviously programs that will allow me to do this.

In regards to dynamic settings target vs max frames cap, I expect dynamic settings to use the target I set and to maximize at least tlod/olod around that target. The max frames should simply be a cap and not effect dynamic settings tailoring my tlod/olod to the target I set. But again, MS keeps all this secret, so people seek out 3rd party solutions, which MS forbids discussion about. MS really looking silly and very petty with this imo.

1 Like
17

You should set max frames to 40. OpenXR has it’s own algorithms for determining frame rates, and with your refresh at 80hz, OpenXR will likely fallback to 40fps. I have my max frames at 40.
Also, I tune my system for Ultra Texture Resolution. This seems to give me the most stability, and I recommend this if you have the VRAM.

18

Thanks for that info. I’m gonna keep running wide open for the moment and test as it’s the highest my settings have been while still getting smooth performance.

If I move texture resolution to high or ultra my VRAM gets crushed. I compared high to medium and I cannot discern the difference visually. I restarted msfs after the setting change. What are the big changes you see visually with the ultra setting vs high or medium?

Also you confused refresh rate and resolution. I originally had my refresh rate in Piplay set to 120 and capped at half that (60). I removed the cap in Piplay. 80 was the resolution, not refresh rate, that my Crystal Light was set at in Piplay. Sorry for the confusion.

1 Like
19

I unfortunately have the low-rent Quest 3 but I run into similar issues. I set the FR to 36 which meshes with their Q3 internal FG to produce 72 FPS which is their slowest frame rate. I have turned down the MS Dynamic Settings as suggested and use other means to adjust frame rates. Seems I can get away with a TLOD of 100-150 at SFO with FlightBeams detailed airport as well as aircraft insertion. I run my 2D about twice the TLOD but the VR looks pretty good. I do run the Clouds at Ultra in both. PS I run the texture resolution at low in both since I think that adds little to the visual but kills the processor.

20

The Quest 3 isn’t that bad, lol. It has 25 PPD.
I had a Valve Index that was 14 PPD and I couldn’t read some of the gauges.
Went to the PCL with 35 PPD. I have to say it is pretty impressive.
I think you have a pretty beefy PC from what I remember you posting.