HP Reverb G2 and recent IL2 or steam update, improvement?

[RAAF492SQNOz_Steve]

Is anyone else noticing improved VR performance in the last 24 - 48 hours when using their HP G2 Reverb (or even other headsets)?

Been flying my usual Kuban ground attack missions that I use for testing various WMR, Steam and IL2 settings and have observed a significant improvement in HP G2 performance.

Up until yesterday I was pretty much locked into using "Balanced graphics, Steam VR Resolution per eye at 60% (application specific SS at 100%) and no motion smoothing or antialiasing." to maintain 90 Hz. This is with a MSI Ventus 2080 Ti which is a mid level grade 2080 Ti variant.

 

Now I can run High Graphics and FXAA x4 very smoothly at 90 Hz with the G2 which in the past would have resulted in significant juddering and frame drops. I am not sure where the improvement comes from but I sure hope it lasts!

 

Recent changes that may or may not be contributing:

1. Latest 4.504 IL2 patch (but not listed as a fix)

2. New SteamVR release

3. Changes that I have recently made and forgotten but apart from pushing settings higher do not think I have made any other IL2 setting changes.

 

As per my original question, are other IL2 VR users noticing this?

 

EDIT comments added to original post.

Addendum 18/12, was thinking about what else had changed and realised that another possible factor relates to room temperature. My city is currently experiencing a significant and unseasonal cold snap. Ambient room temperatures would have dropped by approx 10 degrees Celsius (18 F) between original tests i.e. 28 C to 18 C. Wondering if is making a difference to my PC's performance?

Improved performance has appeared at the same time as the cold snap arrived.

The cold change will last for another four or so days and then, no doubt, summer will be back with a vengeance.  This should give me a good chance to get a "seat of the pants" bench mark.

 

For those wondering what "seat of the pants" means........ It is a bit of a nod to the comments about perception versus scientific observation  made below.  I consider "seat of the pants" benchmarks to fall midway between the two extremes of perception versus scientific observation but then I am an Aussie!

Edited December 17, 2020 by RAAF492SQNOz_Steve
Additional comments

[I.JG3_CDRSEABEE]

1 hour ago, RAAF492SQNOz_Steve said:

Is anyone else noticing improved VR performance in the last 24 - 48 hours when using their HP G2 Reverb (or even other headsets)?

Been flying my usual Kuban ground attack missions that I use for testing various WMR, Steam and IL2 settings and have observed a significant improvement in HP G2 performance.

Up until yesterday I was pretty much locked into using "Balanced graphics, Steam VR Resolution per eye at 60% (application specific SS at 100%) and no motion smoothing or antialiasing." to maintain 90 Hz. This is with a MSI Ventus 2080 Ti which is a mid level grade 2080 Ti variant.

 

Now I can run High Graphics and FXAA x4 very smoothly at 90 Hz with the G2 which in the past would have resulted in significant juddering and frame drops. I am not sure where the improvement comes from but I sure hope it lasts!

 

Recent changes that may or may not be contributing:

1. Latest 4.504 IL2 patch (but not listed as a fix)

2. New SteamVR release

3. Changes that I have recently made and forgotten but apart from pushing settings higher do not think I have made any other IL2 setting changes.

 

As per my original question, are other IL2 VR users noticing this?

I jumped into a tank yesterday. Normally it would send me into below 90 easily and have red/orange on my GPU in fpsVR. But I did notice yesterday that it seemed less taxing on the GPU. Will have to try later. 

 

Didn't they say something about the ground objects/terrain being changed? Maybe that has something to do with it.

[dburne]

That would certainly be welcome.

I won't get a chance to try it out till the weekend.

[SCG_Fenris_Wolf]

I have seen nothing like that in patch notes... we'll have to bench it. Feelings are no indicator guys.

[firdimigdi]

Haven't had a chance yet to try 4.504 on the Rift S - probably tonight. Given that I don't use SteamVR if it's anything from the game it should show up in the frametimes.

The one thing that does stand out in the changelog for the game engine which sounds like it could have performance impact (either better or worse) is the work they've done on terrain rendering.

[Pekka_Rayha]

I was just wondering this morning how am I getting these framerates now, quite a lot better than few days ago and with SteamVR too. I thought its because I fiddled with some settings but could the update be it at it too? 

 

I'm using Quest 2. Oh and I'm benching with fpsVR. 

Edited December 17, 2020 by Pekka_Rayha

[dburne]

36 minutes ago, SCG_Fenris_Wolf said:

I have seen nothing like that in patch notes... we'll have to bench it. Feelings are no indicator guys.

 

True however how one perceives it to run on one's system is what really counts.

 

[SCG_Fenris_Wolf]

Usually I'd just wink, and not write to it. But in today's times where scientific standards are blatantly dropped left and right, people reason with "their feelings", I can't help to think that's what a horde of monkeys does... let's say I sharply disagree: Perception is deceitful. Only hard numbers are facts. 

 

So it absolutely does count what the benchmarks say, and it is just as absolutely irrelevant how anyone perceives something.

 

You could have a bad time after eating some super spicey Thai food and fly afterwards. Will it subjectively be good? Of course not. Will you have the same framerate? Of course yes.

[firdimigdi]

Had a very very quick look at this in my go-to testing scenario (8 vs 8 QMB at 2500m, clear weather, Autumn Moscow map, north of the city over some lakes), frametimes reported in the Oculus overlay appear identical to what I experienced last time.

Edited December 17, 2020 by Firdimigdi

[dburne]

2 hours ago, SCG_Fenris_Wolf said:

Usually I'd just wink, and not write to it. But in today's times where scientific standards are blatantly dropped left and right, people reason with "their feelings", I can't help to think that's what a horde of monkeys does... let's say I sharply disagree: Perception is deceitful. Only hard numbers are facts. 

 

So it absolutely does count what the benchmarks say, and it is just as absolutely irrelevant how anyone perceives something.

 

You could have a bad time after eating some super spicey Thai food and fly afterwards. Will it subjectively be good? Of course not. Will you have the same framerate? Of course yes.

 

Thanks for the nice reference on "horde of monkeys".

Every individual is different.

What matters to me when I play this game in VR, is how it performs for me while actually playing it in the manner that I prefer to play it with the graphics settings I prefer to play  it with.

 

Yeah it is nice to know how my rig might stack up to others when running the same benchmark at the same settings. Might give me a clue as to a weak point in the rig. I have even considered changing my ram to a higher frequency based upon some of the benchmark results.

More importantly - again to me - is how it performs and how it looks when I am actually flying missions in single player campaigns at my preferred graphic settings.

For others , how it benches in a given track at given settings may be what is most important to them.

 

It's all good, different folks have different tastes. 

 

 

 

 

 

[SCG_Fenris_Wolf]

As long as it runs 90hz everything is fine 

[RAAF492SQNOz_Steve]

Thanks every one for responding, much appreciated.

 

I have made added additional comments to my original post and added another possible cause for my significant improvement in graphics performance.

 

Re: SCG_Fenris_Wolf comments about perception versus scientific standards.

There is a degree of rigour  in my results but as the "environment" is not controlled it does fail scientific standards.

i.e. the test location and flight path is the same (+ or - 100m) and 4 tests are typically conducted with framerates and GPU utilisation being monitored.

Big fail on not monitoring CPU and a few other metrics  Also both IL2, SteamVR and room temp changed at roughly the same time.

 

The big win however is that I am enjoying the improved performance even if I do not yet understand why it has occurred and have my fingers crossed that it lasts.

Edited December 17, 2020 by RAAF492SQNOz_Steve
correction

[dolki]

I confirm with my samsung odyssey and gtx 1080, I have more fps since 2 days, so it's not the last update of the game, but I think it's the last nvidia 460.89 driver.
I play MP with 90Hz, ultra, HDR, msaa off and 110% steam VR. (1860x2324)

Edited December 18, 2020 by NN_Vaucouleurs

[RAAF492SQNOz_Steve]

7 hours ago, NN_Vaucouleurs said:

I confirm with my samsung odyssey and gtx 1080, I have more fps since 2 days, so it's not the last update of the game, but I think it's the last nvidia 460.89 driver.
I play MP with 90Hz, ultra, HDR, msaa off and 110% steam VR. (1860x2324)

 

Thanks for your response, that was not a contributing factor to any performance gain for me as I had not applied that NVidia driver.

Was running a driver that was about four months old.

 

Was however a timely reminder for me to apply the driver update  .

[dolki]

Maybe Microsoft - Holographic - 10.0.19041.2037 install 11/‎12/‎2020 ?

Edited December 18, 2020 by NN_Vaucouleurs

[Panzerlang]

5 minutes ago, NN_Vaucouleurs said:

Maybe Microsoft - Holographic - 10.0.19041.2037 install 11/‎12/‎2020 ?

I saw that one come in, it didn't noticeably affect performance for me. I thought it had improved the whole mouse-cursor crap-shoot but now I'm not so sure.

[dolki]

Intel - System - 2040.100.0.1029 Install ‎11/‎12/‎2020 ?

[SCG_Fenris_Wolf]

Guys, I have now benched it. I had benchmarks (Stan's V6 together with Chiliwili 69) from last week.

 

Edit: The game runs precisely as quick as before on Chiliwili's VR test settings. However, maybe another factor is at play.

 

I noticed the game gains 2-3 FPS when switching SSAO on. I double checked, triple checked, quadrochecked. Confirm that at least.

Edited December 19, 2020 by SCG_Fenris_Wolf

[ITAF_Airone1989]

Somebody has some experience with OBS VR plug in?