dlder

Just came across this; shame on me for not getting to know this trick sooner. Anyhow, whilst reading through the INI, there are a couple of things I wanted to ask you about; maybe someone knows this: it seems to me that "OverrideAutoExposure" could also be quite helpful or the only setting that's needed? The linked video looks promising at least! I AM gonna try this myself too, but was wondering what you are seeing in DCS? if one uses a newer DLSS version - as most of you apparently do, did you force sharpening via the nvCPL ? (nvidia control panel) Because it says "DLSS v2.5.1 and later versions removed sharpening from DLSS (use sharpening available in NV control panel instead)" -> or does that only apply to the "OverrideSharpening" and the "ingame sharpening" setting still applies? Well, I guess that's all for now. Hopefully I can give this a go soon. Would be awesome to play DCS without the ghosting artifacts

Yep; that's the strategic and tactical side from it. But even if we'd had those coordination capabilities ingame, doing that in the ME... well, that would be a whole different ball game.

It would be really cool if you could make this campaign compatible with the Flaming-Cliffs version of these jets! Easiest would probably be straight copy and just exchange the Jet-IDs. Or you could maybe make it able for the player to choose (like setting the planes to Client), but I don't know if that would mess with triggers and such. I mean, it could be done by anyone personally, but not since all campaigns are encrypted... so no one can change a thing, which is both a blessing and a curse^^ Thanks!

So, after analyzing a lot of hwinfo logs: no pagefile is ever in use (as it should be, as long as RAM is available). But rarely the GPU memory (24 GB! What?) seems to fill up, causing a crash. This usually only happens when switching modules and maps, without quitting DCS. -> for me, this looks like a buffer overflow issue of the game. Old cached data isn't removed from the GPU Ram (fast enough). I mean, if this happens with a 4090, this should be a huge issue for smaller cards, one might think? Or is it, that this is so much memory, that the game says: Ah, it's fine, enough memory here!

Using Pimax Offical OpenXR + their weirdly named "Pimax Central Priority Rendering" (when will Pimax get someone that speaks English to go over the text and the tooltips!?). No more issues since then!

No worries, thanks for your reply! Yeah, I still know, as I'm still testing the issue and no, while using Mbuchia QuadViews, I've turned both Pimax settings off. I've also reported the problems regarding Pimax-QuadViews to Pimax support. For now though, we are still in the "have-you-tried-turning-it-off-and-on-again" phase (meaning bull<profanity> answers from support. I hope my ticket will get pushed to the higher tech support soon). edit: but I gotta say, now that I no longer use the unfortunately old Mbucchia Pimax-OpenXR runtime and -QuadViews, I have no problems with the game anylonger. So I guess it really is true what the developer posted himself: to no longer use his tools

That's the idea. Would be absolute bonkers if possible in DCS

Well, as my logs show: pagefile is never in use. Sure, I gotta check with Syria too, then I can be certain. Hm, yeah, tried that. Waited quite some time too. I'll give it another go if it crashes again.

I will from now on start DCS with hwinfo64 in logging mode to get to the bottom of this. I do have some more info because of the first 2 logs: I currently have the i7-14700, that means 8 pCores / 16 pThreads + 12 eCores/Threads and all threads will be used by the game! But for whatever reason, the eCores more then the pCores, except for the 1 main thread, which for me is running on pCores 4, Thread 0 Yes, the GPU memory will get filled up until..., well, its full. But that's by design, right? After loading the game, about 15 GB of GPU Mem are used. This will gradually fill to 22.9 - 23.7 GB of the 24 GB available. So, if the game erroneously fills up over this point, then that would point to a potential ingame problem of not removing old data from the video memory (if the game really would need more RAM, because of my chosen settings, it would fill up pretty much when loading a mission -> this happens when playing No Man's Sky in VR with the Pimax Crystal without DLSS, so I know what that looks like^^). What I do see though, that in the 2 gameplay sessions I logged, is that the Physical Memory (RAM) is used 46-59% (of 64 GB). That's no where near overfilling. -> that's also, why the page file isn't used at all, just as it should be! BUT: what does get filled up pretty high is the Virtual Memory: the 1st session isn't that wild, Vmem sits at up to 71%. But on the 2nd session, this reached a whooping 94%! It is rather easy to circumvent any such memory overflows, if one exits the game every few hours or at least, when changing the map. What I remember, Syria especially needs a lot of RAM, so I'm gonna play around that map next time and see what the numbers are then.

Unfortunately, I have to replace 1 mountpoint to add that 1 mountpoint + whatever I want to add. By that I mean: to get the outer pylons, I have to add these shapes & textures to an existing mountpoint. That means, I would need to create 2 extra shapes (one for each side) and add these to the default shapes (the LAU-115). That's quite a lot of work for... less missiles^^ "Should" be possible if I can get the default shape, but unfeasible, sorry. edit: for better understanding, the marked missiles / mounts are 1 mountpoint!

Version 1.06 is still valid for DCS 2.9.9!

You have to look inside the ZIP files... Yes, v2.84.1 But, I didn't try it ingame. Too much of a hassle to make the same screenshots from mid and low altitude and from the ground. I just go after the size/date for each picture (inside the ZIP) So yeah, maybe GTC actually does look better, subjectively at least. The biggest issue I have with both is: terrain.cfg is now encrypted. So... what does that mean for such mods? Thanks! Could you please add which picture is what?^^

Just a quick question: just for this reason, I've joined the DCS discord, dropped the log into the "dcs-log-analyzer" chat and... nothing happened. How does this work, so that I can help myself next time?

Thanks, I'll try to get some more info if/why the GPU Ram is really filling up. Seems weird. But I did fear that I should only set 1 pagefile, though it shouldn't matter in any way.

This is a bit strange imho... I have 64 GB RAM 32 GB Pagefile 24 GB GPU memory And after changing from one map to the other (I'm not sure which; from Nevada to Persian Gulf maybe?) the game crashed. The log is attached. (line 2.680) How is that possible? Could it be the pagefile layout? C = 256mb (minimum for the small Windows dump file, if an error accurs) D = 32 GB Any ideas, tips welcome! I only have 26GB of free SSD space on C, so I can't use the main drive for the page file.... dcs.log

Here's the full DCS.log dcs.log Will try! 1st: Mbucchia OpenXR-Pimax + OpenQuadViews 2nd: PimaxXR + OpenQuadViews (NOT PimaxQuadViews) Today I tried PimaxXR + PimaxQV and haven't had any (XR runtime) crashes so far. I'm now using the second option in the Pimax-Client. Not "Pimax OpenXR Quadviews", but "Pimax Central Priority Rendering". They really need to start providing tooltips and/or better option names, but I'm guessing this is Fixed-Foveated-Rendering? -> because PimaxQV has two 3 problems Settings cannot be adjusted and the quality is too low for my taste the 2D mirror from DCS is ultra-low-resolution which is unacceptable, as I'm recording from that XrNeckSafer is incompatible

Just checked the other logs: Mbucchia Pimax OpenXR Mbucchia QuadViews Log (this is the "newest", where I played with the official PimaxXR, thus no log file from the runtime like the one above): Maybe I should add, that I'm currently playing No Man's Sky in VR with the Crystal. No crashes whatsoever. None. And that's a 150+ hours... and this game REALLY stresses the 4090 because if DLSS is not enabled, you run out of video memory.

Hey, so, I haven't played DCS since the summer (stability issues and Su-25 TV bug). I've tried it again today and... there are a few things different now: Pimax finally has their own OpenXR it also includes QuadViews with DFR or VFR DCS too has QuadViews and EyeTracking options (for either DFR or VFR) Now, I have both options enabled in DCS, as this seems necessary for DFR to work with either Mbucchia or Pimax. But because the quality of Pimax's DFR-settings suck and - for now - cannot be changed, I am using Mbucchia's QuadViews (as well as Pimax-OpenXR). It works, though the performance seems "off". And, then this happened two times in different missions / maps / jets: As you can hear, the game itself doesn't crash, but continues. It's just the output that's borked. I guess the OpenXR runtime crashes and the HMD stops working and looses track. In "dcs.log" are quite a few of those errors: ERROR VISUALIZER (26880): xrEndFrame failed with error: XR_ERROR_RUNTIME_FAILURE There are no VR / OpenXR / nVidia specific errors in the Windows logs though! And now crashdumps of anything! If I remember correctly, Mbucchia himself said to stop using his OpenXR runtime, because it's out of date and he stopped working on it, so I switched to Pimax's own, official OpenXR. But the same thing happened again... Here's my DCS options: Here's my OpenXR setup: and with PimaxXR: And here are my Pimax settings: v1.33.01.01 Lighthouse, 120Hz, Eye Tracking enabled, Sound Standard RQ = 1.0, DFR via QuadViews and Pimax both disabled, Smart Smoothing disabled, Lock to Half Framerate enabled, Hidden Area Mask enabled

I just work in IT, which doesn't actually mean anything other then that it allows me to make educated guesses and allows me to do some modding if I like. BUT: I don't know nothing about textures and stuff! So, I can only manually check every file from mods vs. the games files, which I do on pretty much every update. The gist is this: CGTC has lower quality textures and also fewer textures then Barthek. That's why I prefer this texture pack. And again: I cannot recommend to mix both, but if you do, install CGTC first and then Bartheks. Caucassus Terrain hasn't been changed by ED in quite a few years, so this pack is still very relevant. Only the now encoded terrain.cfg remains a problem... That is true! But it shouldn't really affect you, if you have texture and visiblity set to max. ALTHOUGH: I have no idea how the LOD system works in this game... maybe it would be beneficial to mix both mods, but then you'd have the nasty job of manually unpacking all ZIPs of both mods and use the "best texture". And even then you have the problem, that maybe the darkness/shading isn't the same, so you have different kinds of textures between low- and high-quality. -> try both and stick with one

Updated to DCS 2.9.8.1214 As last time, Aim-9 and -120 code changed. hf!

I'd really like to know: now that the "terrain.cfg.lua" is encrypted, what do we do? Can we just delete the encrypted one? Does somebody have a de-crypter, to merge the changes by ED into the mods file? Because as it stands, this file is no longer relevant.... Also: I at least wouldn't install the terrain.cfg.lua I mean, I wouldn't mix those 2 anyways, you'll never know which texture would be out of place. But to each their own. I've also been using all the tree-mods from Taz, but as it is currently, they are out of date and imho should no longer be used (just compare the files they use; ED updated some and are thus newer / don't exist with Taz's mods).

gotcha! Thanks again!

First off: thanks! Those settings get deleted, when you start DCS, but get added by the launcher (just open the launcher and don't start the game, and you'll these options in the LUA)

I've just flown a night mission in the Su-25T and I was not able to lock on to any ground targets. I also couldn't see any targets with LLTV (night time TV!), as the screen was completely dark. And I have "symbols" enabled, so I know where the target is located; meaning, I'm looking right at it in the TV. Only in lit areas (from flares/light bombs for instance) I'm: a) seeing the target on screen b) able to lock onto tanks But even that doesn't work all the time; I've made a quick test mission, where I also made sure the tanks a "warm" by moving them, and they aren't visible with or without lighting (fires and flare). Here's a video about that; track is also attached. The daytime TV is also way too dark and it's hard to find targets, but at least the Skval is able to lock onto targets... Could someone confirm or deny that? How are you guys doing the past 2 weeks or so? (I'm flying in VR, if that matters) su25.trk

There were changes for the Aim-9 and Aim-120 types of missiles. Mod on GDrive was updated to 2.9.6.57650