cSonora

Hey team, I just wanted to let you know that after the latest update I've been able to play several long sessions without having a single CTD or BSOD, so in my case the issue seems to be resolved. Having a blast with the module so far. Congrats for such an amazing work and thanks for your support! Carlos.

I just played a couple of sessions, in the first one I was able to play 30 minutes until I got a crash: dcs.log-20240526-125750.zip The second one lasted 10 minutes and crashed my system completely resulting in a BSOD: dcs.log I will try to play a couple of sessions more this evening. Thanks for the support.

Hey Nicholas, thanks so much for your reply. Sure, I work as a tech artist in an AAA game studio, and I personally know pretty well how hard and time-consuming bug-hunting processes are. I will keep testing and posting my crash logs here, and if you need anything else, don't hesitate to ask. I will be more than happy to help fellow devs.

I tried reinstalling the module and disabling the AV, and crashes keep happening every single time. I also played several hours with other modules like the AH-64D or the F-14 with no issues, so I can confirm that this problem is isolated to the F-4. I haven't been able to play a single full mission yet with the F-4, which, after waiting so long for the module release, is honestly quite frustrating. It would be nice to have at least some answers from the team giving us some ETA on the fix. Here's my latest full crash log. dcs.log-20240526-071731.zip

Getting also DCS crashes, only when flying the F4. My system : i9-13900K 4090 64 gb ram. I can fly from 5 to 15 minutes until the game crashes. The last lines of my logs are always similar to these ones: 2024-05-24 15:59:30.882 ERROR ED_SOUND (Main): invalid source_params(woLA-16777472:aircrafts/engines/j79/rearclosehighpower): gain 2024-05-24 15:59:30.882 ERROR ED_SOUND (Main): invalid source_params(woLA-16777472:aircrafts/engines/j79/distantmil): gain 2024-05-24 15:59:30.882 ERROR ED_SOUND (Main): invalid source_params(woLA-16777472:aircrafts/engines/j79/frontdistant): gain 2024-05-24 16:01:47.775 ERROR ED_SOUND (Main): invalid source_params(F4E_PILOT_COCKPIT_2D:aircrafts/f-4e/cockpit/windforce): gain I tried repairing the files, reinstalling the GPU driver, and the problem persists. Both DCS and User Files folders are in the AV exclusion list.

Makes sense, the thing is that SteamVR's MR doesn't seem to have that overhead cost, as FPS keep exactly the same when I disable/enable it (with SteamVR I can disable/eneable it in real time via its video menu, so it's pretty easy to spot the difference). I've still got my trusty 1080ti. I need as much VRAM as I can get for 3D modeling stuff, so I have been reticent to upgrade and loose VRAM (the 3090 is the only option but, its price is definitely not something I'm willing to pay for a GPU )

Not only that, but in my case enabling motion reprojection causes my GPU frametimes to increase. I don't know if this should be the expected behaviour with OpenXR, but in my system, enabling SteamVR's motion reprojection doesn't impact frametimes at all. I have continued my testing with many aircrafts/scenarios combinations and the result is always the same, even in the DCS menu screen. Examples with average APP frametimes: F18 Caucasus free flight: MR disabled: CPU: 11ms GPU: 16ms Render Refresh Rate: 55hz MR enabled: CPU: 11ms GPU: 21ms Render Refresh Rate: 30hz DCS menu: MR disabled: CPU: 2.5ms GPU: 11ms Render Refresh Rate: 90hz MR enabled: CPU: 2.5ms GPU: 14ms Render Refresh Rate: 45hz

More of my testing insights, just in case this is useful for anyone. I've been testing which are the results of changing the resolution scale in the OpenXR Tools for Windows panel VS doing it in the OpenComposite's ini file, using the supersampleRatio option. I'm getting like 10fps more by adjusting the scale via OpenComposite and letting the custom scaling disabled in OpenXR Tools. Regarding the motion reprojection issue, definitely in my case is due to that enabling it causes my FPS to take a considerable hit. So with MR disabled I get 50 to 70 FPS on F18 free flight over caucasus, but with it enabled it can't surpass the 30 FPS (unless I'm looking at the sky), hence I'm getting 30hz reprojected to 90hz, wich I think it what makes this MR looks worse than SteamVR's MR, wich always managed to give 45hz reprojected to 90hz. The benefits alone of being able to increase the resolution so much with OpenXR make me stick with it, if just this little MR issue could be solved to achive smooth reprojection would make this the definitive VR API to use in DCS IMHO. PD: Getting rid of SteamVR's spamming flashing screens during loading times is a god sent gift also

Hi everybody! I've been testing OpenXR today and I wanted to share my results. With an 1080ti I'm managing to get my Reverb G2 to work at its native resolution with having an average of 50 to 70fps and a really smooth tracking response, with some occasional stutters, specially on flyby views or when the terrain is moving fast. In order to have the same smoothness with SteamVR I have to lower a lot the resolution and to enable motion reprojection, so a big improvement for me. Which I'm not able to get to work properly is OpenXR's motion reprojection. I wanted to activate it to see If that helps to smooth the stutters, but in my case, setting it to "always on" makes my framerate to go down to 20-30fps, completely killing the smoothness. I did what @edmuss sugested to try to fix it, but it doesn't work in my case. Is anyone else getting this behaviour?