Oh is this the update to openxr that @mbucchia teased a while back? I know he wasn’t allowed to say much, but I definitely remember a big update to MR being briefly spoken about.
so MR with shake reduction nullifies the impact of MR and MR without shake reduction is super jumpy/sensitive to movement.
any solution to this issue?
Hello,
With DLSS (DX11 / DX12 ?), should we use MSFS CAS (up to 200% ?) or OXRTK CAS or both ?
(My system = Reverb G2 - GTX3080 - i7-12700 - 32GBRAM)
Yes! A lot better now! 
Hmm yes better but still terrible. I just tried it. If you go to external view and pan around the plane, about 40% of the plane starts jiggling around like crazy. Also some artifacts from cockpit view. Still unusable for me =(
2 Likes
When testing MR, make sure you turn off anti-shake, or whatever it is called now, in the OXRTK first as it affects the MR algorithm.
Reset your environment data in wmr settings. Give it a go. I do it periodically and it never has a negative impact. Only ever positive or neutral.
Hi all! Being still on my mission to pinpoint the bloody thing that destroyed my pretty much decent performance I noticed two other things that bother me:
First thing regards rendering resolution. Not sure it has been like that all the time and I just noticed it right now or is there something I don’t understand… Is it normal that 70% render resolution produces a different output resolution depending on what engine do the scaling? For me (Reverb G2), in-game 100% + openxr runtime 100% + toolkit 70% = appx 2164x2164-ish while in-game 100% + runtime 70% + toolkit 100% = 2640x2500-ish. Shouldn’t the result be exactly the same?
Second thing is lack of any impact on performance of FFR. Zero. I see its working on the screen and it used to give a solid boost to my fps before. Now literally zero noticeably benefit in fps nor frame times. Madness…
Do these sound familiar to anyone?
The percentage is expressed in different units.
Percentage in OpenXR Tools for WMR is in total pixel count.
Percentage in MSFS and OpenXR Toolkit is a fraction of the size (the width in pixels).
You might no longer be GPU-bound, where FFR helps. Take a look at your Performance overlay to see where your bottleneck is (CPU or GPU).
Ah, at least one headache less, thanks!
I think I still am. Overlay readings are as follows:
app CPU 20-ish ms
rdr CPU 15-ish ms
app GPU 40-ish ms
Can you try the following:
- Under Menu in OpenXR Toolkit, enable Show Expert Setting
- Under Performance, enable Overlay as Developer
- What value of
VRS RTVdo you read?
VRS RTV: around 50 (45-65)
VRSw: 2212
This means it’s working properly. Most likely you are no longer Pixel Shader-bound, and therefore FFR (which reduces the amount of Pixel Shader invocation) is not helping.
2212x2212-ish isn’t a very high resolution, so it’s not entirely surprising to not be Pixel Shader-bound.
is this shaking in VR a reverb G2 thing or a WMR thing or a specific headset thing or a tracking thing?. can i eliminate it in any way without using shake reduction because then i cannot use MR. does this issue exist in steam VR and/or on other headsets as well. would getting a G2 V2 help?
Im using G1, after this G2 from years. I never noticed any strange shaking. If yes, it was my head only 
.
Most likely, it may be related to the fact that someone has not cleaned the environment (of WMR) for a long time and has not centered the view in the room / at the desk from long time too. It is worth doing it at least once a month, in the light at which you usually fly (the room must be well lit, you cannot fly in the dark with Reverb). A shaky image is most often related to external factors, it’s the user’s fault most often.
The question is what could cause the change in the balance then? I literally didn’t touch anything in my hardware configuration and previously FFR had tremendous impact on perf. Could this happen because of one of the recent sim updates? I think I noticed the issue for the very first time soon after the SU with the new in-game settings like, FSR, motion repro, etc. or about the time Turbo mode arrived to Toolkit.
turn of motiom reprojection?
well the whole point is to get MR working. if not using MR then everything i said is irrelevant
Just FYI I saw the same. But I am not sure what the reason is. With my 3080Ti I got about 32 FPS at airports. Switching on or off FFR did not change anything. Now I have a 7900XTX, which gives about 46 FPS (the PC is also stronger now). But switching FFR on or off changes the FPS by about 10%. That is huge and new for me.
Thanks for your input. You changed graphic board so your GPU to CPU balance could change somewhat as well which would explain different result of the FFR on performance and as Matt mentioned, if the load is more on the CPU side the FFR benefit would become lesser as it unloads mainly GPU. My scenario is more mysterious though, same rig before and after. Might be I’m not aware my wife changed hardware config by spilling some water and burning a few circuits 
1 Like