StrongHarm

You can get there with the 'originally posted by' link:

Nessuno0505, where VR is concerned, try kicking your PD up to 1.2, it makes a big difference. I wouldn't go over MSAAx2 though. The considerations for the various settings are very different for VR. Drill into quoted post for VR specific quality/performance analysis.

Not only does 4x decrease my performance, but it also introduces visual trails on edges when I turn my head, especially on cockpit objects. I try to avoid this effect at all costs. It will decrease your immersion and cause a lot of eye strain in VR. I've never seen this happen when using 2x, even when turning my head fast in a graphically intensive situation.

There were some discussions about using MSAAx4 instead of X2. Not only does X4 decrease my performance drastically, but it also introduces visual trails on edges when I turn my head, especially on cockpit objects. I try to avoid this effect at all costs. It will decrease your immersion and cause a lot of eye strain in VR.

Valium, this is probably because you're loading textures and objects into memory as you encounter them. I like to speed this along by hitting F11 to change view to a few airfields and spin my point of view around before taking off. This used to be very useful several years ago before ED optimized object load.

With a monitor MSAA is optional. In VR it's essential. Although PD is really important, a mixture of MSAA and PD is required. With MSAA at 2x it's difficult to raise PD above 1.5, but since the MSAA is cleaning up the picture, any settings above 1.4 has diminishing returns. I use 1.2PD and 2xMSAA. See this thread.

MSAAx4 is a no go for me too. I have an i7-5960x extreme, EVGA 1080ti FTW3, 16g RAM, 1tb dedicated SSD. Going from MSAAx2 to MSAAx4 cuts my performance in half.

Outstanding analysis Sideslip! There are a lot of varying facts and opinions on these boards concerning quality and performance gains. This demystifies the situation. I put together a VR specific settings guide: 2.5 Video Settings - VR Quality/Performance Balance in the Virtual Reality thread. Of course, there are different considerations for VR, but I've linked your post in my OP so people can look at base line visual comparisons. Thanks for taking the time to do this for the community!

Edited OP with a new performance step: I've seen a lot of threads about problems being solved by uninstalling the upgrade then reinstalling using the full 2.5 install. I decided to try it and see if there was a difference. I found that I had about 300 files less in \Program Files\ after the reinstall. My theory is that those files were leftovers from a time before ED launched DCSupdater. I haven't cleaned those folders in probably 5 years. It seems that I'm loading missions faster and that I have less stuttering. I acknowledge that this could be a placebo effect, but at the very least I'm glad that I cleaned up the file structure.

Mr_sukebe, the mod files aren't the target when uninstalling the upgrade to install the full. It's files within .\program files. I've read a lot of posts where people have said it made a big difference in 2.5 for them so I tried it out and it had a clear visible effect for me too. Where Saved Games is concerned, I just copied my Missions, Input, and Screenshots folders and changed the name of the root folder so it would create all new files/folders. I then pasted Missions, Input, and Screenshots back in after it was done. I don't use any mods though.. I'm kind of a DCS purist. I would recommend not using hardware monitoring tools at all while flying DCS, unless you're specifically doing testing. Schrödinger's cat gets fragged by the high CPU requirements of DCS.

Gladman, I don't disagree with you. Doing nothing is an option. The other options are to either do extensive testing, or use someone else's extensive testing. I like to share my findings so people with similar hardware don't have to spend several hours collecting their own, if they so choose. See link in my last post for full data. I too respect Jabber's opinions, but his advise comes with a caution that he's targeting high alt MP performance over all else. He didn't need to use new settings because the new terrain doesn't really effect high alt MP. I personally like to fly helos sometimes, and with 2.5 I'll often find myself on simple freeflights to enjoy the amazing work ED has done with the new terrain. If you're using Jabber's settings then you're getting a 4 quality and 10 performance (according to my analysis). If that's a persons goal then they're right on target. If you want 8 quality 8 performance, I've found that 2xMSAA on top of 1.2PD not only increases performance and visual quality with deferred shading on, but it also exposes the full beauty of the new terrain. This is not in disagreement with you or Jabber, it's just what I've found. The one essential change from Jabber's settings for old DCS I would recommend is Preload Radius. Having it all the way up as Jabber does, regardless of your system specs or performance goals, will cause issues. 2.5 creates new considerations for this setting. To your point though Gladman, some people aren't interested in spending a lot of time trying to tweak their settings. Even if someone is sharing their test data, they would rather just get to flying. Nothing wrong with that. Enjoy.

Short answer: VR PD = 1.2 MSAA = 2x Anisotropic Filtering = 2x Deferred Shading = ON Terrain Shadows = FLAT Cockpit Global Illumination = ON Preload Radius = lower can be better Long Answer (with test data and additional info): https://forums.eagle.ru/showthread.php?t=201647

Mr_sukebe, I feel your pain. Your i5 CPU is most likely your bottlekneck since your 1080 has a good GPU. But if the root of your problem is something other than your CPU, I can suggest the following. NOTE: I would try each of these suggestions one at a time so you can gauge which one helped you the most. Replying with your findings might help another vPilot. I noticed that you have 24g of RAM. Do some research on your system/MB and make sure it's rated for uneven increments of 8. Having unequal RAM sticks installed causes some(*most) systems to have really strange memory allocation issues. You'll have to purchase a new stick or simply remove the odd one (i.e. if you have 2 16g and 1 8g, remove the 8 ). Make sure your BIOS is updated to the latest version. I know this sounds trivial, but I assure you that it can cause some seriously weird issues if you're on an old BIOS version while using newer drivers for components that are plugged into your MB. Perform a clean install of the nVidia drivers. When you're doing the install you'll see a checkmark at the bottom of the screen for 'Clean Install'. Be aware, this will delete any 3d profiles you have set in the nVidia control panel, but it also refreshes all DLLs and gets rid of files from previous driver versions. Make sure that your Rift Sensor is connected to a USB 3.0 connection rather than a 2.0. This can cause 'smearing'. Reinstall the Rift Drivers. If you're using Windows 10, make sure you turn off Game Mode in the OS settings. Download and install Razer Cortex. This app has a database of known 'unneeded tasks and services' that may be running on your system that could interfere with game performance. When you left click on the tray icon and engage 'boost' it automatically ends task and stops services that aren't required, freeing resources for games. After you're done flying, simply left click on the tray icon and select 'Restore Now' and it will restart all the tasks and services it stopped. This is a safe application that has never stopped a service that I actually needed, and has never caused me any issues. It frees up system resources. I recommend this because you may have something running in the background that is causing your issue, but why search for a needle in a haystack when you can burn the haystack down. If you have 'Set Affinity' to specific CPU cores, remove that and choose all cores. If you don't know what this is, don't worry about it because you probably didn't mess with it. Delete the files from "..\Saved Games\DCS\FXO and .\metashaders" Bring Preload Radius all the way to low. Test in 10% increments until you receive desired performance If you upgraded stable DCS to beta, do an uninstall. Then reinstall from the full install distribution. I've been flying DCS for a very long time, so my install had a bunch of junk files from before they started using the DCSupdater. DCS ran much smoother after I performed a clean full install of 2.5. Read this post for additional info and 2.5 video test data https://forums.eagle.ru/showthread.php?t=201647

I turn off HeadingAP and AltAP. I then either execute a "nose forward and up" controlled stop, or a "nose turn and pull" immediate sliding stop. In both cases, the important point of my method is that I pump the collective. Fully lower the collective first until you just start to lose altitude (1-2sec), then pull your maneuver and pump the collective in an anti-rhythmic oscillation to maintain altitude. Lowering the collective helps you to bleed energy, while pumping the collective "blows air" in the direction you're trying to stop. From cruise I'll typically use the full range of the collective if I'm initiating a fast deceleration. This can take you from 270 to 0 in about 3-5sec. If you look at it as quickly stopping in front of brick wall using the same method you would use to halt a fast descent toward the ground, it helps to understand it. That's not an easy, or entirely logical, explanation. I wish I had a better one.

ram0506, your stuttering situation may not be due to 1.2PD 2xMSAA. If you have time, try to following and reply with result: - Delete files from FXO and Metashaders as outlined in OP - Preload Radius - down to at least 30k, but preferably min setting - AF - 2x ( I know, this isn't supposed to cause stuttering, but I assure you it does right now ) - Terrain Objects - Flat Also, be aware that some people who are having stuttering issues are deleting their 'updated' installation and performing a reinstall from the full 2.5 distribution. I did this and it made DCS much smoother for me. My theory on this is that people who have had DCS for a long time may have a number of junk files that existed before the DCS installer app, and they're not being deleted even though they're not needed.

dburne, sorry for explaining something that you already understood. You specifically mentioned a 45 FPS limit in your last post, so I thought it was ASW. Maybe someone else will find the explanation useful. If you try 1.2PD 2xMSAA with shadows on (flat) and deferred shading on, reply with your findings and system specs. It's a good balance on my system with a quality of 8 and performance of 8.

Buckeye, I have a monster system which is meticulously maintained (see OP for specs, 25yrs in tech). I spent two 8hr days trying every reasonable combination of PD MSAA settings. The short answer to your question is this: - 1.2PD and 2xMSAA is 8 quality 8 performance - 1.2PD and 4xMSAA is 8+ quality 4 performance According to my personal test results, going over 2xMSAA in VR creates too significant a performance drop to be viable. 1.2PD 2xMSAA was the sweet spot for my system, achieving 8 quality 8 performance. You'll notice that the quality/performance variations are not at all linear or logical. Here's the full analytical breakdown: [TABLE] __PD__ | _MSAA_ | _Quality_ |Performance 1| 0 | 0 | 10 1.2 | 0 | 2 | 10 1.4 | 0 | 4 | 10 1.6 | 0 | 4 | 10 1.8 | 0 | 4 | 8 2.0 | 0 | 6 | 6 1 | 2 | 6 | 10 1.2 | 2 | 8 | 8 1.4 | 2 | 8 | 6 1.6 | 2 | 10 | 4 1.8 | 2 | 10 | 2 2.0 | 2 | 10 | 2 1 | 4 | 8 | 4 1.2 | 4 | 8 | 4 1.4 | 4 | 8 | 2 1.6 | 4 | 10 | 0 1.8 | 4 | 10 | 0 2.0 | 4 | 10 | 0 1 | 8 | 8 | 2 1.2+ | 8 | 10 | 0 [/TABLE]

dburne, forgive me if I'm telling you something that you already know: 45 fps should be considered acceptable at this point in time. With VR's Asynchronous Spacewarp(ASW) you'll typically see 90 FPS, 45 FPS, or a value between 1-44. The reason you see your FPS lock at 45 FPS is because your actual available FPS is higher than 45, but doesn't reach 90. This could mean that your FPS is 87, but ASW is forcing 45 FPS. Locking the frames to 45 instead of allowing a fluctuation between 45 and 90 prevents undesirable effects of VR such as edge tearing and motion sickness. Outside of VR, 45 FPS is not very good. We've been conditioned as gamers to accept nothing less than 60 FPS. With a monitor 60 FPS is good, but 120 FPS is amazing. VR is different; in VR 45 is good but 90 is amazing. My goal with a quality/performance balance is to have a stable 45 FPS and not see it drop to 40. 90, of course would be better, but almost all visual quality has to be compromised to achieve that. A solid 90 FPS may be a reasonable goal in DCS in the future but as of right now, no matter how powerful your computer is, a solid 45 FPS is really good. https://developer.oculus.com/blog/asynchronous-spacewarp/

dburne, did you lower PD to 1.2 when you turned shadows on and you still had issues? You have the same video card that I have and it seems to perform very well on my system. I've also found that turning Preload Radius down helps in 2.5.. I think there's a problem with it. I've been hovering between 0 and 30k PR, and 2x and 8x AF. The best balance I've found is with the settings list in OP: VR PD = 1.2 MSAA = 2x Anisotropic Filtering = 2x Deferred Shading = ON Terrain Shadows = FLAT Cockpit Global Illumination = ON

That works. The name of the game is personal preference. I would encourage you to turn the shadows on sometime and check them out. ED has done amazing things with terrain lighting in 2.5. In my eye, that's half the beauty of the new graphics. I do understand that people have different goals and sometimes only want to ensure only minimally good graphics with maximum performance for multiplayer.

From another post: I bought an Alienware Area51r2. I've always built my own systems, and they're usually in the top systems list on 3dMark and BOINC. Before purchasing this system I had specced out parts on Microcenter.com for purchase to build my own, but decided to go compare the costs of Alienwares. At this point something occurred to me; when buying really big servers for work for developers to bang around on for test servers, I'll sometimes go to Dell's remanufactured site and buy a server that would be $50,000 new for $10,000 used. I've never once received a bad server and the warranty period is the same as a new system. So out of curiosity I went to outlet.dell.com and looked at the Alienware systems for sale. What I found out is that most of the systems on their site are systems that were returned as 'lemons' by spoiled rich kids after they received their first blue screen of death. The QA team at Dell does deep diagnostics and replaces entire subsystems if they throw any errors, and of course they're reimaged. So, back to my buying experience: I ended up finding an Alienware system for less than $3000 that had the same exact parts that I had specced for almost $5000 to build my own(due to intel extreme processor and dual top of the line EVGA vid cards). Of course, it was a different case, but everything else was basically the same (and the case for the Alienware Area51r2 by the way, is hands down the best case I've ever seen). Now, you're probably thinking what I was thinking.. "this thing is going to be full of bloatware and you're going to have to wipe the system and reinstall a nice clean OS". Nope.. it had zero bloatware, and I have to say that the proprietary software it did have was stable and useful. Their Alienware Command Center makes overclocking a breeze, and I was able to quickly create a nice LED color schema that I liked (and I was planning on just turning that flashy crap off). I've been in the computer industry for more than 20 years, and before that I worked on fighter jets, so I'm familiar with technology. I can easily recommend the Alienware Area51r2.. this system SMOKES.. and it has a warranty that will outlive its use.

For the sake of testing, try going from: MSAA-2x, PD-1.8, Shadows-Off,Vis-Low to: MSAA-2x, PD-1.2, Shadows-High,Vis-High .6 points of PD is very expensive, and really doesn't make as big of a difference, since you're using MSAA. Make sure 'Terrain Object Shadows' is flat instead of default or off. You probably won't get 90fps stable, but performance shouldn't be bad. Visual quality should be greatly improved without suffering performance problems.

I believe I did see this once long ago. I'm not sure of the cause, but what I did was checked 'full screen' in settings. After that I was able to ALT+Enter with results, even when 'full screen' wasn't checked.

ED has historically overdelivered on every level with DCS. During Blackshark 1 I was looking forward to DCS going somewhere in the future, but I never imagined I'd be flying in VR with the terrain we're seeing now.. with CA, Normandy, a great ME, and the dozens of other mind blowing things they've done through the years. I have high expectations that we'll see some KA-50 love, and fragmentation physics, etc. I can't imagine how massive their code debt list is considering the high standards they hold along with the mind blowing complexity of this product. They're one of the better software companies where listening to the user base is concerned. The best thing we can do to contribute is write a bug report and be patient.. or if you're really passionate join the test team. It's a lot of hard work and time, but it's pretty rewarding to be part of what they do. I can tell you this; after being a tester you may be a lot more patient with the devs considering what they're able to achieve on a daily basis. I say this not argumentatively, but reflectively.

I edited the OP to add a few additional performance gain methods that I use: Something else I always do when starting a map, and this is going to sound crazy but it works, is hit F11 to switch to various airfields and just spin the mouse 360 to take in the area. I do this on several airfields to preload a lot of ground objects into memory. It sounds crazy, but it cuts down on FPS drops over cities. Lastly, and this sounds crazy too but it works, I hit ALT+ENTER to go full screen. I know that it shouldn't make a difference since I'm using VR, but it makes the mirrored window on the monitor go full screen and it seems to make a difference.