Jump to content

Wali763

Members
 View Profile  See their activity

  • Posts

    418

  • Joined

  • Last visited

 Content Type 

Everything posted by Wali763

  1. Wali763
    Did not read through the thread, only parts. Sry. I dont like DLSS, makes gauges rather hard to read. Also ghosting. DLAA removes all the shimmering but makes the image blurry. Still running lower settings with 2xMSAA to achieve 90fps. No DLxx for me. For me the best compromise.
  2. Wali763
    Ok thx for your input.
  3. Wali763
    I got a question: Many have stated, that the Aero´s sweetspot is much higher than on the G2. But when in my G2 I look to the edges of the image, I can see the overlap-edges of the lenses. I very much notice the edge between the part, where only one screens is visible vs the part, where both screens overlap. So it does - at least for me - not matter much, that outer parts of the image are blurrier, since those lens-edges are rather irritating to me and I like to stay focused on the center. Is this the same on the Aero? Or does it have better binocular overlap without those edges visible to a lesser degree?
  4. Wali763
    Sorry for replying so late. Did not see your post earlier. I had this problem on every map I tried. Was not related to PG only. Since I removed the OTK, the problem is gone. Still have to say, that this might only be a few hours, since I dont have too much time for DCS lately.
  5. Wali763
    Hi. I had occasional flickers/flashing with 2.9 until I disabled the Toolkit. Also seems my frametimes went down a bit too. I had only Foved Rendering and the fps-counter on. At first I tried to change other settings like clouds, shadows, etc and also went to general low settings, but nothing changed; still had the occasional white flashes. Only disabling the Toolkit did help. And the problem has not come back yet. Will keep it disabled.
  6. Wali763
    I did not use MSAA up to now, only 75% SS in OpenXR with my G2. Im kind of ok with the shimmering/aliasing I get. DLSS, DLAA and TAA all give me a blurry mess, but all greatly reduce shimmering.. DLAA also reduces perf noticably (~1ms higher frametime). 2x MSAA retains most the the non-AA clarity, but reduces shimmering to a lesser degree than the three above alternatives. Perf impact of 2x MSAA is way too high, so I did not bother testing 4x. As the original poster, for now Im staying with my "old" settings. Will do a retest at a later time.
  7. Wali763
    If you go over to the amd.com support-forum, there are several threads about VR-performance of the 7000-GPUs, this being the longest I think: 7900xtx Low Performance in SteamVR and VR in gene... - AMD Community 7000card dpmt seem to be that good in VR. Some report, that their 7900XRX is as good as 6800XT in VR. I have a 6950XT and its fine. Wanted at first to get a 7900XTX but did not because of the reports of its poor VR-performance. Right now Im very happy with my decision. If my card were 20% faster though, that would be great. I would not bet on AMD getting their issues with the 7000-cards sorted. Its already been 5 months since its release.
  8. Wali763
    Hi. Just need to rant some, since I do not understand this. Probably its just my lack of understanding how PC-games work..... Between water quality on low (which I usually have to get max fps) and quality on high, there is a huge perf difference. Its maybe like 15 or 20%, going (in a caucasus mission I tested this) from ~9ms GPUframtime to something like 11 or 12ms. Well thats ok, since you get something for the extra-Perf. But the thing is, this fps-drop is also over land, where there is no water around. I do understand, that water is rendered below the ground textures. But to me this seems like a big waste of GPU-power. Isnt there a way to tell DCS, to not render water, if there is no water visible? I really could use that extra GPU-performance....
  9. Wali763
    Ok, thx. I was for some time thinking of upgrading to a 4090 or at least a 4080. But since I am already quite used to my lowish settings but 90fps in most situations I fly in on my AMD-GPU, Im not sure, if its worth paying so much for a bit more eyecandy. Anyway, Im not in a rush and Im pretty sure, prices will come down much more. At least for the 4080.
  10. Wali763
    Maybe a stupid question: Are many with a G2 (or other 90-Hz-HMDs) running DCS in 90Hz? Especially with MT and considering 4090-owners for sure have fast CPUs, this should be easily doable. Yet when browsing this thread it seems, many people even with 4090s are below 90fps. Do they prefer to use higher settings instead of fluid 90fps?
  11. Wali763
    I have the same once in a while; maybe 1 out for 4 times. Then I have to restart DCS and usually its ok.
  12. Wali763
    I had it recently, that when there was a call from "Overlord", the correct message was written at the top, but the vocals were those of "request bogey dope"
  13. Wali763
    Hi. I have Ryzenmaster. But since I cant oc my 5800x3D, I dont use it. Chipset drivers were updated 1 month ago. Mobo Bios is a bit dated, 03/22. Did not update, since there were no desireable new things to potentially upset a working system. No gamemode, no gamebar. See no use in those. As I wrote above MT brings very little change for me. Maybe 1ms GPU frametime. In CPU-limited scenarios MT helps a bit, but stutters still occur.
  14. Wali763
    I have the strange situation, that in ST (non-MT) in normal (lighter) missions my cpu-frametimes are lower than in MT. At the same time Fps-tracker in MT constantly says "CPU bound rendering thread", whereas in ST there is no message. I get the frametimes from Frame-Timing-Overlay in the OpenXRTools (not toolkit), but the fps-counter in DCS seems to support those numbers. It seems, that CPU-frametimes in MT try to stay around 10,5ms. So, where in ST I get eg 8ms CPU-frametimes in MT I get 10-10,5ms. Thats actually in contrast to what MT should do, meaning lowering CPU-frametimes. At the same time the GPU-frametimes drop a bit by ~1ms, which gives some more headroom for higher video-settings. In larger missions MT is working to some degree. Its lowering the CPU-frametimes in situations where in ST I got 17ms CPU-frametime to 11-12ms in MT. But since (in my case im running my G2 without reprojection to get to full 90fps) this is too high for real 90fps, I get stutters. If those were arount 10ms or lower all would be good. Even those 10,5ms are pretty close to the needed 11,1ms for 90fps, overhead in lighter missions is not that big so I get an almost smooth 90fps ride in lighter missions. But in heavier missions, this overhead might to get too big, so it causes fps to often drop below 90fps which gives visible stutter and spikes in the DCS-framecounter.
  15. Wali763
    Just a brief feedback about my findings about MT. This morning I posted, that I dont see any difference MT on or off. I now found some differences for me, but Im not sure, if MT is making it better for me. My settings are directed at getting full 90fps in my G2 since AMD-GPUs and reprojection dont work well together IMO. So I only use 45% OpenXR resolution and most things set to low (except textures and visib range). Its a tradeoff between eyecandy (eg little shimmering) and smoothness. I prefer second. I have a 5800X3D and a RX6950XT. So I guess Im mostly GPU-bound anyway. Usually when frametimes (either CPU or GPU) jump over ~10,5ms fps drop below 90 and I get visible stutter. The difference for me in MT is, that in really CPU-heavy missions CPU-frametimes are better. They are almost the same as in smaller missions and hover around 10ms. Problem is, they often get spikes over 10,5ms, which causes stutter. GPU-frametimes mostly are around 9ms in my tested scenarios and no problem. The problem is, that also in light missions I get those 10ms CPU-frametimes but far lesser spikes to 10,5+ms, whereas in non-MT I have lower CPU-frametimes between around 8ms in comparable situations. So in short it seems that MT in my case seems to level all CPU-frametimes around 10-10,5ms, now matter if a mission is big or light. But since those 10-10,5ms are close to my "stutter-threshold" and those stutters also happen, its not all good for me. On the other side, I do have the feeling, that GPU-frametimes may have dropped by maybe 1ms. I did comparisons back and forth and Im pretty confident saying that. That allowed me to set ground-textures to high again without lowering fps. So my overall finding is, that with my current CPU MT does not add much and differences are rather small. Only in heavy missions, that I usually dont fly anyway. I guess, if current CPU-frametimes would be a bit lower (like 9ms instead of 10ms), so spikes would not let fps drop below 90, Id be very happy with the update. But so my feelings are mixed and Im not sure, if I should go with MT or stay with non-MT.
  16. Wali763
    Hi. Did some testing with your missions, especially the PG one. I cannot confirm your findings. Load times are quite variable for me and overall there is consistent difference. But I now found that, MT also works on my system, since with off-MT I get CPU frametimes up to ~16ms, whereas with MT I get mostly 11ms. But I think this is due the fact, that I already have a good CPU (5800X3D).
  17. Wali763
    Im on a 5800x3D CPU in OpenXR. I dont see any improvement. Im using the frametime-overlay in the MS OpenXRTools and it seems CPU frametimes have gone up. I tested a large InstantActionMission with max units back and forth and CPU-frametimes are now mostly at 10,5 whereas before I saw lower numbers. Looking at the Taskmanager-performance tab in Non-MT I have at least 5-6 threads at idle, while in MT all seem to get used somehow. But ony thread is always at the max. Also the other mission I tested showed the same result. Anyway I also dont feel any difference when not looking at the numbers. The biggest difference for me is visible when jumping into a cockpit. With MT on it now takes 5-10sec to fully load the image. So at first you hoover over the sea and have only the frame of the cockpit with no textures before you. In non-MT I did see this almost never, maybe only on very large MP-missions. Also the lobby (hangar-floor) has a red tint now, but thats no problem. Colors in the cockpit look the same.
  18. Wali763
    Tested 23.3 yesterday. No change. Still flicker. Only 22.5 works for me without flicker. Had a 6800XT before and did not experience the flicker-problems with newer drivers. But with the 6800XT I (other than testing) mostly used the 22.3 driver (that does not support the 6x50-cards) since it felt the smootest when capping fps to 45/47. Going to full 90fps did not work for me with the 6800XT as it does now with the 6950.
  19. Wali763
    Is anyone with an AMD 6x50-card (mine is a 6950) having the same problems as I do? Unless I use the 22.5 driver (actually the oldest to support my 6950) I get grey flickering on one eye. Tested almost all newer driver since 22.8 up to 23.3 but all had this problem. Using the Opencomposite-way, I dont get the flicker, but instead at times bad stuttering. SteamVR has no problems but works as bad as always before. Win10 caused those flickers to come back recently since it updated the driver to a newer version. So I had to reinstall 22.5 and block windows from updating drivers in the background. Other than that, Im pretty happy with DCS right now. I lowered my settings and set OpenXR-resolution to 45% (which is around the native res if my G2). Now I can enjoy DCS in most situations (usually A-10 on Caucasus, PG or Syria in singleplayer) with 90Hz. Aliasing is quite visible now, but thats a ok-tradeoff for me.
  20. Wali763
    I think the 'optimizer' is meant (also in conjunction with xbox-app) to detect if the running app is a game, in which it will park the non-Vcached CCD or if its a non-game-app, where it will use both CCDs. I dont think this will have relevance for current 5800x3D-owners.
  21. Wali763
    Hi. Just a brief report from my side. @roboHackfeld Using any kind of repro jection like ASW will give you some kind of jitter, mostly visible on planes flying past you close by. Nevertheless: A month ago I changed my 6800Xt for a 6950XT, that I got a great price on. I expected something like a 15% increase in fps, that did not materialize in benachmarks like 3dmark. Its more 10%. That at first made me a bit sad tbo. But then I tried to go for full 90fps in DCS and lowered all my settings. Now I only have vis range to high and textures to med. Shadows on low and OpenXR set to only 45% (which is kind of the resolution of my G2 anyway). That said, Im very positivly surprised, since Im getting (maybe with the help of my 5800X3D) for the most part those 90fps and therefore a really smooth ride. Especially in the scenarios I usually play (A-10C on Caucasus and PG) I only drop below 90fps over larger cities and on the carrier. Syria is also ok. Visibly, its of course a step back and the shimmering, that OpenXR "solved" for me is back. But during gameplay, this actually does not matter whereas the smooth 90fps matter big time. Before I had nice visuals with little alialsing/shimmering at capped 45-47fps, but I still had jitters and double-vision on 3/9. But now its mostly really smooth and for me more enjoyful than before. Asked for some rating, Id say SteamVR is a 3/10, in which case I for long preferred 2D. OpenXR capped at 45fps was a 6/10. Nice visually with some stuttering. Now Id give it at 7/10. Poor to moderate, but acceptable visuals with 98% smooth 90fps. I not too often do MP, but the last time the experience was quite good too and maybe once multihreading comes along, this could even get better. I really encourage people to give it a try like I did. Im running my 6950XT with a moderate 4-5% OC (22600pts in Timespy-graphicscore in 3dmark), which should also be doable for a 6900XT.
  22. Wali763
    Big Thank you for your elaborate answers!!! They really helped me s lot, also in understanding. Thx
  23. Wali763
    Hi. Probably a problem, that does not touch many people. Still I cant figure this out for myself. In short: Making a new Markpoint blocks the way back to the previously set WP-page on the CDU. I only have a workaround, that seems odd. Anyone know the correct procedere? Or is this just a bug. Long: Usually for a mission I prepare a new waypoint, so I only need to push in the UTM-coordinates and be able to use the WP. But when creating a Markpoint, the CDU jumps to the markpoint-page and I dont seem to be able to go back to the last WP-page to further edit my prepared WP. Pushing "prev" wont do it. I can set the AAP-SteerPt-switch to "other" (not Mark) and then I can circle the Markpoints (like I were on Mark on the AAP-SteerPt-switch) but not the rest of the WPs. Even going to some other page in the CDU and then pushing the "WP"-button will bring me back to the Markpoint-page. The only way out for me right now is to create a new WP, then I can circle the previous ones too. But then I do have to edit the new created WP or circle back to the previously. "My" procedere of going back to the last WP seems odd and wrong to me. Anyone know the correct way?? Thx.
  24. Wali763
    Hi. I recently heard about those Aernime-drivers and will give a try. Thx.
  25. Wali763
    Hi. Just a brief feedback to my problem with a flickering headset. It seems going back to an older driver (AMD Adrenalin 22.5) solved this and some other problems with GoogleEarthVR. So its not always good to follow the recommendation of installing the newest driver. Adrenalin 22.3 worked best for me, but this does not support my current GPU anymore.
×
×
  • Create New...