Jump to content

Thinder

Members
  • Posts

    1414
  • Joined

  • Last visited

Everything posted by Thinder

  1. OK done. Thanks again. No, it doesn't work. I didn't touch the line before it first failed to start. Now I got two .exe on this line.
  2. Thanks. I copied/pasted from the email I received from support. I'm correcting immediately, but it seems that the line given by support ias correct, something is adding .exe to the line but it doesn't show in the shortcut.
  3. Did anything changed? I got this error message now and I didn't change anything since I last was able to log-in... i guess it's back to writing a ticket to support...
  4. 100%, we need something that would actually work for us and dispense with the extra apps...
  5. The main problem was the duplication of folders, I had to rename the DCS.openbeta folder to Backup, it took forever to load and apparently started the game on another map/mission, I guess interception looking at the sea.
  6. Corrected. Thanks to support who responded quickly, I can play this release and it's great, I gained <> 15FPS flying the Mirage F1 at very low level Nellis AFB map in high settings, I'm sure I can improve on this with better drivers and a few tricks...
  7. Tested in 2D. High settings, it runs real smooth, I believe my issues are coming from the VR implementation, not Multithreading at all, and I'm sure it can run with the same high settings in VR once those issues and AMD drivers are sorted. This is the first few seconds of the game, Mirage F1 EE, Free flight Caucasus, after that the FPS is real low, so there is a fundamental problem with the way VR runs VR with my Pico, I will submit a ticket to support to try to figure out what's wrong...
  8. I had ONE good run using the M-T release then it really got bad but I don't think it is coming from DCS, Windows 11 Pro for example doesn't do scheduled clean up despite me setting it for daily, I had to get rid of 240MB of junk files manually. It went some way to recover the Visual quality and FPS I had in January but it's not totally there yet, I think the implementation of VR itself needs extra work, something changed since and it can be improved further, GPU usage was 100% but running cool, CPU usage was <> 16% at around 76°, I might be able to use a more demanding map for the CPU. After that, the game started in a grey, unmapped mesh in a blue environment, it takes <> 10 sec before everything else falls into place, mapping, cockpit instruments, terrain and then it is real slow and jerky, I tested RAM sticks latency and CPU multithreading performances with CPU-ID CPU-Z and it's fine. I tried testing with FPS VR before that in the "normal" version, I had a low 40 FPS with lower settings than I used in January, Frametime was sky-high but I keep in mind that now I have to deal with bad AMD drivers too, so I'm reinstalling it and Drivers and if I can recover what I experienced during the first run it will be better.
  9. That's why B.Die is recommenced by G.Skill support, lower latency and Ryzen3/3D are designed for lower latency, it works with ANY application, now the latest I get from them is this: I was worried about the CPU IMC support for 64GB because those sticks are two ranks instead of 4 and in theory, Ryzen IMC are limited to 4 but after testing it doesn't seems to be the case with this kit and CPU, on the other hand I quote eurotech replying to my last email: So considering the gain in performance, I don't see where the "overpricing" is, people spend much more than this premium in CPU water cooling only to see a lower gain and lose their manufacturer warranty, from my PoV, it is optimization. Latency related to cache or RAM is independent from the game engine, and in fact, the reason behind the Ryzen 7 5800X 3D cache design, it's all about data transfer times, it is only logical that a lower latency RAM will add to the gain obtained by the CPU cache. In my case, going from 32GB to 64GB is motivated by fact that I do video editing and also play around with 3D Studio Max, I am planning to design some concepts ships based on Elite Dangerous, not that I hope they would use them but since this sort of activity was my job in the past it became a pleasant hobby. Also on top of capacity, according to eurotech, there is a "a slightly better memory bandwidth performance".
  10. Yeah, this too, after a fresh windows install...
  11. It really doesn't need Open XR, this is one of the half bum solutions introduced it an attempt to recover the loss of performances that came with the "VR" update, I had much better performances before and adding layers after layers of apps isn't the solution, it just contribute to more loss of performances even if it allows the headset to run with a disabling VR update. Just to give you an idea, even after deleting and uninstalling all the stuff I don't need running in Windows 11 Pro background, I still have more than 55 apps doing so, now, the more apps, the heavier the load on the system, the lower the visual quality, it was working great before the update, why would anyone want to add more apps for playing VR? It looks like this situation have been corrected since, I still don't use Open XR and it's better now, even if not as good as it was in early January, it's a progress, not telling you not to use Open XR, just saying it really doesn't need it. I can tell you the procedure I use to launch a game and DCS in VR, 1) Streaming assistant; Steam VR kicks in, once in the room it presents you with the options you chose, I have Elite Dangerous and Virtual Desktop, to make sure I don't associate DCS with Steam VR directly, I launch it from Virtual Desktop. I had better results improving visual using a simple trick; setting the Resolution in the / Saved Games / DCS / Config/ Options.lua ["graphics"], 800 to 100 and 10280 to 320, the load on the GPU is reduced since it doesn't have to fill up the screen you don't see with the headset anyway, I noticed a clear improvement in cockpit instrument readability. When it comes to environment, with the way I test DCS, flying at very low levels where redraw is much more demanding, the change is much less noticeable and I still can't see the electric lines before I got my nose right on them, as opposed to what it was before, the FPS is also stocked in the low 40s. I estimate the loss to <> 20FPS just looking at how smooth the game was with much higher settings, so the increased load on GPUs since this update is important, with no improvement in visual quality whatsoever but the opposite, but to be fair, part of it also comes from mediocre AMD drivers introduced after the update... Reducing GPU load in VR.
  12. I had a G2, now a Pico 4, the difference in image clarity is minimal, but the Pico has a much larger "sweet Spot", nearly 100% of the lenses. When you'll struggle to get this level of clarity working for you with the G2 because you have to aim the center of the lense to what you want to see, with the pico, I can read my instruments while looking at the runway in front of me for example. In the Mirage F1 you can see your instrument panel and all the information the HUD doesn't show or much too small (old tech HUD) without the HUD mod. Tracking is excellent, never had an issue with it, the fact that I always play in a dark room with a 4W red Led behind the screen might help (high contrast). As a mid-range I'd say it is the ideal headset for DCS because it is light (295g vs 550 g), it is well balanced thanks to the battery, has very good visual, and is fairly comfortable. You also can enjoy a growing library of free and payable apps and games which is something I came to appreciate. In passing, controllers are good too, easy to use. >>> On the minus side, the USB cable doesn't recharge the battery but if you can play for longer than it last you'll probably shot you eyes anyway. It's new so its support and range of apps and game is still not on par with the competition but they are making a lot of progresses at this level. It will be supported for much longer than the G2 which is now out of production. DCS support for the Pico have been screwed by their "VR" update but they will correct this, it is already way better than a month ago. You'll need a strong GPU. Talking of price range... £622.80 for a G2, £449.00 for my 256GB Pico 4, you don't need 256 GB of Pico storage to play PC-based games, the 128GB is cheaper. Resolution: 4,320 X 2,160 X 2160 per eye. Refresh rate: 72/90Hz. Pancake Lens 105° FOV To summarize, I wouldn't swap my Pico for a G2 today. ps, the red Led trick could work for other headsets too
  13. I had similar issue, try to use the repair tools, it worked for me...
  14. I flew 3 different aircraft, the Mirage F1 and 2000C and the F-15, what this first Pixel Density setting did was to give me a good clarity at the cost of more jerky play, in fact it was the first setting I started to lower to get a smoother game, then everything not affecting the clarity and cockpit instruments readability too much. From 16 to 13/14 there is little difference visible with the Pico4, there could more noticeable differences with a high end headset, I can't tell. Setting the screen resolution using: / Saved Games / DCS / Config/ Options.lua ["graphics"] 800 to 100 and 10280 to 320 leaves me with a very small repeater in the middle of the screen, it did reduce the load on the card enough for me to notice an improvement in the readability of the instruments after all the losses with Driver and VR update, not so much with environment... It's OK as long as you start the game with your mouse cursor inside the screen and disable the option "cursor confined to game windows", or the cursor can go AWOL.
  15. We all hope for AMD to pull their finger out and provide with a proper driver (unless they are committing commercial suicide)... From my 3DMark Pro tests, I lost minimum 9.3% in Combine Score at 4K 2 X MSAA Firestrike Ultra, and when I say minimum it is because I lost the previous tests results but I'm sure it was gone well above 20 000. The best result I obtained after driver update is with manually under-volt with 19 040, in default settings, it's more like 18 495, then again, the last two drivers causes the card to crash when under-volt is selected, manual of automatic. there is absolutely no improvement with the last driver release, some people are sending their GPU back which is perhaps a bit overdone but I understand why. This card has a huge potential but is hold back by bad drivers, especially with 10% loss plus <> 3% you can expect from under-volting it, the RX 7900 XTX Nitro+ Vapor-X runs cool under load and never did throttle back due to thermal limits, if it haven't been for those issues, I would have enjoyed my VR experience a lot more and be able to provide with proper tests including FPS. So I'm not blaming the GPU but the bozos who released 3 drivers in two month, made the one that worked unavailable to players, all of which with no positive results to show for their (and our) troubles...
  16. Well, actually I did when I started my first tests, for the story, I used (or tried to) afterburner and the FPS you see in one or two of those videos is not that of the actual game but screen recording of the replay... I didn't know about FpsVR despite the fact that it is already in my Steam library, thanks for the tip. The goal was to push the card to its maximum and see how to improve the visual, to achieve this, I flew very low level where the terrain details redraw is more demanding than mid or high altitude, I came close to something very smooth with high level of details in the last video before I was able to find Afterburner settings so I don't know the FPS but it was smooth. With the Pico4, the clarity of the visual obtained was indiscernible to that of my screen, the difference is just that you can see the grid, like playing with your nose on the screen, resolution was at its maximum, Display set to Ultra, Bitrate 150Mbps, Codec HEVC and Refresh Rate 90Hz, I was able to see the electric line from much further away than today. Sapphire Radeon RX 7900 XTX Nitro+ Vapor-X 24GB GDDR6 first run. I was also setting my BIOS, made back-to-back performance comparative between my PC configurations (upgrades) and the PCI_E1 slot was causing trouble, setting itself back to default causing the card not to output any signal. With the help of MSI support, I was able to solve the problem by using PCI_E2 and setting the BIOS back to PCI_E1 Gen 1 but it happened after I made the last videos, so all I can say right now is that with the right driver, the card seems to be very strong at 4K.
  17. AMD Drivers and Support Gonna try it, hopefully it will sort out a few issues. After test in 3DMark Pro, 4K 2 X MSAA it doesn't improved on the previous driver...
  18. As a Pico 4 user, there have been much more damage done for VR by the DCS update than AMD drivers. Those drivers are holding the performance of the GPU by about 9.3% at 4K MSAA in 3DMark test, but this issue appeared after the DCS update and the image quality + FPS I obtained when I tested my GPU before the update were really good, I managed 4 test videos before the DCS update destroyed my VR performance. It is possible that people with slower AMD GPUs can experience issues in VR because of drivers that I don't see, from my PoV, on top of the loss of performances, we have some mesh-maps distortion in Elite Dangerous, making the moons where the engineers are based look cracked like nuts, making it impossible to access them and engineer our ships/systems. Now my visuals are of much lower quality, better than a month ago but I can't see the electric lines as I did when I flew this test on the same map (Nellis AFB) with the same A-C, and much lower settings, the side image is jerky, I pay particular attention to this and it is the reason why I fly at three top then look on the side. So both RX7000 are suffering from AMD drivers but in DCS also from the update, particularly those using the Pico4. Not only, I had issues in VR shortly after I uploaded this video, before I updated my drivers. That's got nothing to do with the 7900XTX itself, but in the case of Picoi 4 users, the DCS update first and foremost, AMD released 3 drivers in 2 month, proof that they know there is a problem and it didn't make things any better, by your reasoning I wouldn't recommend Pico 4 either but both are great. Since I was lucky enough to start testing my GPU before both the DCS and AMD updates I know exactly what it is capable of and it's a great card let down by terrible drivers, as for the DCS update, well we all expect a good, solid update bringing a significant boost in performance, even if it's late, it will do. There are some solutions for AMD users, and also those with lower specs PCs HERE I now have a much improved cockpit view with readable, sharps instruments, but no real improvement in environment, for example I can't see the electric lines before I got my nose on them, now good for three top flight and side image is jerky. Just in passing, in any serious (and not drivers biased) tests, the RX 7900 XTX is faster than the 4080 at 4K, that's enough to justify my choice for this card since my requirements were VR at 4K 2 X MSAA. I forgot: At the time of the last video, my GPU bandwidth was reduced by 50%, because I couldn't use the PCI_E1 slot, issue resolved now but it says a lot about the capability of the card.
  19. If you run it on Cl16 RAM, it won't actually be as fast as it can especially at 4K, that's what the problem is with AMD Zen3, they are almost never bounded to the RAM that actually make them work properly.
  20. For those interested, we have a few solutions for low specs PCs and AMD users (due to drivers holding performances) HERE it should help a little, personally, I have now an excellent cockpit instruments and logos readability, not as much improvements in terms of terrain details and FPS but it's a start. Here is the whole team at work, and Vanellope von Schweetz turns out to be one of the good guys...
  21. True, same for DDR4 but at 64GB, interleaving becomes near irrelevant, there is enough capacity per stick to compensate and the controllers can manage pretty much the same way...
  22. Better late than buggy, we had a "VR" update that screwed Pico users, AMD released 3 drivers in two month that screwed our GPU performances, so if DCS come out with a good, solid update later than expected I'll be happier than what I saw lately. I think they know that the latest updates weren't up to what can be expected from them, I don't think they want a repeat and @BIGNEWY is only doing his job, passing us the info he can when he can.
  23. Thanks for your insight, I chose not to reduce the mirror if I can find other solutions, in any case this is a stop gap set of solutions while I wait for AMD to release better drivers and DCS support for multi-core processors which hopefully will give a noticeable performance boost. Your solution is valid for people with slower PCs wanting to gain some extra FPS, in my case, I try to recover the image quality and FPS I had in January, before the update, but when it comes to raw GPU performances, I also have to deal with AMD drivers issues, so every little help is appreciated.
  24. Excellent! Thanks for your reply. I'll try this. This is for the mirrors. I tried on ["graphics"] 800 to 200 and 10280 to 320, original file is backed up, we never know. = It works, You have to change the option cursor confined to game windows or it still limits its movements, so I have a very small 800 x 320 screen repeater telling me DCS has started but not such a large area for the GPU to fill with pixels... I've done a very low level test sortie from Nellis, the F1 is a b!tch to trim but I noticed a very slight improvement in image clarity (particularly cockpit view where instruments are much sharper), I still can't see the electric lines until I got my nose on them but overall, terrain looks better and side view a little less jerky. Work in progress.
  25. Most players doesn't know what to do with their Zen 3 in order to increase their performances in gaming. That's because the misconception about high street RAM kits (non-B.die) working just as well as the premium Cl14, that's completely false kind of forum disneylandish stories. With a Cl14 RAM kit, the Ryzen 7 5800X 3D runs circles around a 5600X at 4K 2 X MSAA in 3DMark test, while it clocks faster 4.6GHz vs 4.5GHz, the only difference was 3200MHz vs 3600MHz, but the 5600X and the GPU (EVGA 1080Ti) were boosted with Ryzen Master and Afterburner. NO boost for either in the 7 5800X 3D comparative test, test conducted back to back during my upgrade. Yet I gained 32.34% Combined Score, the CPU was running 14.09% faster, the GPU 18.91%, so they can brag about the i5 13600k, I say bring them on. There is a reason why the 5800X 3D has a 96MB L3 cache and it's latency, they stated that very clearly when it came out, so bounded with Cl14 RAM it doesn't throttle down under load, all channels stays fully opened, there is no bottleneck, and low latency RAM + low latency CPU combined to give much higher gaming performances. Now those AMD bashers can bring all their Youtube videos as well, so far I haven't seen ONE which gives the RAM/CPU details and when they do, it's showing a bottleneck (Cl14 removes it), no wonder Intel CPU compare well... You want your Ryzen 3 to run fast? = Cl14, 4 X 1 rank for interleaving, from 3200MHz, you can easily O.C it to 3600MHz or fit a 3600MHz kit as I did to avoid O.Cing but the 5800X 3D will really fly with this RAM. ps, the 5800X 3D also runs faster than the 5800X in multithreading, by memory, something like 25%... In a mainly CPU based game, the cache brings 20% gain, the Cl14 RAM 12.34%, multithreading isn't gonna be an issue with this bound...
×
×
  • Create New...