dutchili

Wow, this made a huge difference for me in multi-player, which is now without FPS drops compared to single player! Keep it up ED!

I don't know anything about the internals of DCS, but when my GPU suffers when more planes are 'in game' while i cannot see them, and the CPU has a bore out, i know our experience can be much better. Same thing with FPS being dependent on the network in multiplayer. Vulcan, multi-core, multi-threading, i don't care what is needed as long these 'issues' are addressed. Such a great game, such a pita sometimes

Amy news @Flappie?

@Flappie: thanks for confirming. Please have a look at the manual for comparison. Beside the blurryness, the aiming cross is also an issue. The manual states: "Use the TQS cursor to place the target within the crosshairs on the WPN format, then press TMS forward to designate the target. The crosshairs will close on the target. Confirm that the missile is tracking the correct target, the pointer cross is not flashing, and that the target is in range." It doesn't seem to work like that anymore. The aiming cross is not 'open' so i can't 'place the target within the crosshairs'. Even if i could see it. In previous version, i could acquire the target at about 7 to 10 miles, now it is 2 to 3.

Found the button to make the screenshots. Reproduction steps to create these visuals: start in flight select A-G select WPN on left MFD leave mode in VIS scroll marker on HUD to target area press TMS up ....see almost nothing ....many attempts with TMS up to lock the target. Pointing 'the middle of the screen' works best, but without the crosshairs it is very difficult. Image 'mav mfd' shows the initial situation. 3 horizontal lines. Image 'target not visible' shows how difficult it is to see the target (FOV was pressed) Image 'cross gone' is the result of TMS UP short (top horizontal line gone, no aiming cross, no lock) Image 'misaligned' shows what happens when lock is obtained after repeatedly trying to lock, knowing where the target should be. Track files shows it all mav issue.trk

How can I make a screenshot in VR?

Sinds OB 2.7.10 the image from the Maveric type D camera, as shown in the Viper, is very blurry. It is hard to make out the targets, especially on city streets. The aiming cross is also ....i don't know how to describe it. I see two horizontal lines, but the aiming cross seems to be located a bit higher on the MFD. Only when a target is acquired the cross is clearly visible together with horizontal and vertical lines that extend all the way to the border of the MFD.

This is probably an issue with the oculus software. I switched to the virtual desktop app in combination with stream vr. The issue is gone. Add fholgers performance kit and your fps will go up too. There are still occasional stutters in a scenario like that, but only for a second.

Indeed Mr_blastman. The issue also showed itself in single player with 'everything' set to high. With VD and SteamVR, FPS will consistently low in that scenario (to be expected)

I switched to Virtual Desktop in combination with Steam VR, issue solved.

Thanks for the suggestion @edmuss. I've tried that once but it didn't give any results. A few days ago I switched to the Virtual Desktop app (from Oculus store) in combination with Steam VR and fholgers performance kit. The issue seems gone! So i think it was the Oculus VR stack that choked. In earlier contact with ED support the root cause was not determined through the DCS logs and dxdiag, which now seems to add to the conclusion that it was Oculus, not DCS. I took a while to find the optimal settings in Virtual Desktop, but i ended up with: - VD: select the graphics card you have (rtx 3070 in my case) - VD: set sharpening to 8% (higher = pixelated, lower = just a bit to blurry, but better than a high value) - VD: refresh rate 72 fps - Steam VR: resolution per eye: 2000x2000 (approximately) - DCS: textures medium, terrain textures low, most other list items on low, no msaa or asiotropicalasomething, Pixel density 1.6 - fholger: radius 0.2, quality 0.5, super sampling method: nis end result: 36 fps reported by DCS without drops. I find it 'no fun' below 30 and 'okay' above 35, so this is good enough of a balance between smooth flying and graphics. With Oculus the terrain looks much worse with the same DCS settings (and fps ofter much lower). After fooling around with the VD settings i can replicate the Oculus quality by selection a 'less performing graphics card', which, according to VD, determine the rendering resolution. So I definitely recommend buying VD for EUR 19,99 and intalling fholgers performance kit. I issue i still have is that multiplayer performance varies from one day to the other. yesterday it was 36 fps constant, today between 22 and 36. I don't know why yet, but i suspect the streaming bitrate in VD which seems to be based on the currently available network bandwidth. I need test some more, but on 'the good day', the setting was lower, which is hopeful.

Similar experience here. Textures on medium makes a huge difference. Especially in multi-player, setting main textures to high will give me the 'drop to 10 fps' issue with the 3070 & Oculus Quest 2. I also suspected the compiled textures to be the issue, but none of the mods help. It seems that the cockpit is rendered regardless of the actual Field Of View, otherwise the FPS would be similar to the LOW setting when looking at a blue sky.

cancel my last. It was due to having the autopilot still on without me realizing it.

Today I plugged in my Logitec G29 wheel again. Later on launched DCS and a looked fine until after 45 into a mission the viper went straight up. As cursing didn't help, i unplugged the G29 causing the DCS to hang completely. Something seems really wrong with either my USB hardware....or DCS.

Same here.

I am happy to see voice comms being available in the simulator itself instead of a separate tool. This makes it much easier for newbies and a better sim. As radio comms work differently in modules, it would make sense to me that you just need to bind to the actual switches in the module so that the whole idea of a realistic sim stays as it is....or was. The room option is nice for briefing and mission starts (instead of discord).

I experienced this once more, opened menu and selected 'adjust controls'. It appeared that my Logitec G29 steering wheels was now also configured as an axis control. As both the HOTAS and Logitec where not disconnected/connected while playing DCS, how could DCS possibly go from one setting to two settings for the Y axis?

Here's an example where it goes wrong right from the first second of flight, without touching the Thrustmaster Hotal 4. flcs issue.trk

I have noticed this too in the latest beta, as in some earlier version. It seems more related to the processing of the hotas data. Full up and down movement with the stick send to successfully recalibrate the input. I have seen this multiple times repeated in the same flight, and most of the times not at all. It happened to me right after entering a mission in flight, so no relation to hard banking.

Stable performance in VR.

@BIGNEWYwith FSR and NIS heaving a big positive performance impact, would it be possible to support this in the simulator? I can imaging a general pixel density setting + pixel density setting for the center of the screen + radius setting (like in FSR config). That radius setting is very smart and I would like to see that for oculus vr too. Thanks for considering.

i saved a track after experiencing a FPS drop with the oculus quest 2. When viewing the track, the drop was not visible. I did notice some issues in the dcs log: 2021-11-15 16:34:02.332 ERROR_ONCE DX11BACKEND: texture 'chromic_blur' not found. Asked from '' 2021-11-15 16:34:02.332 ERROR_ONCE DX11BACKEND: texture 'bd3-usk-a_spec' not found. Asked from '' 2021-11-15 16:34:02.332 ERROR_ONCE DX11BACKEND: texture 'bd4-ask-b_spec' not found. Asked from '' 2021-11-15 16:40:53.588 WARNING DX11BACKEND: alive vertex buffers: 1 2021-11-15 16:40:53.588 ERROR DX11BACKEND: Failed assert `DX11VertexBuffer::buffersCount() == 0` at Projects\render\dx11backend_win8\Source\DX11Renderer.cpp:122 2021-11-15 16:40:53.588 WARNING DX11BACKEND: alive index buffers: 1 2021-11-15 16:40:53.588 ERROR DX11BACKEND: Failed assert `DX11IndexBuffer::buffersCount() == 0` at Projects\render\dx11backend_win8\Source\DX11Renderer.cpp:123

The attached track shows the issue with the zoom function. At various moment is switch between Norm, EXT and DBS1 / DBS2. The A/G radar image does not show in the zoomed functions. Also, the NORM picture has a wrong center of origine in the final seconds. ag radar zoom not working plus wrong center of radiation.trk

Thanks @BIGNEWY. Sorry for not providing a track just yet. I was unable to reproduce so far, but i've seen it a few times. I think it is related to using the dogfight overwrite mode (AGM -> dogfight -> AAM (a/g radar image). But these are not the steps to reproduce consistently. For the zoom: I used pixel density 1.7 when upgrading to the latest beta. The zoom never worked. Then i set PD to 1.3 and the zoom did work. Now it no longer matters what the PD setting is. Regardless of the AG rader working, i am unable to make out any objects. The image seems to blurry regardless the zoom level. (sea/land is clear, so are tree lines. Identifying objects is not possible. (might be by design).

Yes, i will. Forgot to save the track. I can say the issue does not always occur, but repeatedly depressing the master mode switches did not help.