Jump to content

VR Test Discussion


RogueSqdn

Recommended Posts

One thing I noticed on my test was that the game really wanted to be at 45fps. Starting the scenario was at 45fps, and it would go up to 60s or 70's, at t times 90 when looking at sky. And that it seemed to want to go back to 45. I.e. at 45, look up at sky till you get 90, then look at the same scene, and its a 60-70 for a bit (5-10sec), and then it goes back down to 45, look up again. Its almost like there is a buffer or something that you are unloading when looking up, that fills back up.

New hotness: I7 9700k 4.8ghz, 32gb ddr4, 2080ti, :joystick: TM Warthog. TrackIR, HP Reverb (formermly CV1)

Old-N-busted: i7 4720HQ ~3.5GHZ, +32GB DDR3 + Nvidia GTX980m (4GB VRAM) :joystick: TM Warthog. TrackIR, Rift CV1 (yes really).

Link to comment
Share on other sites

  • Replies 159
  • Created
  • Last Reply

Top Posters In This Topic

One thing I noticed on my test was that the game really wanted to be at 45fps. Starting the scenario was at 45fps, and it would go up to 60s or 70's, at t times 90 when looking at sky. And that it seemed to want to go back to 45. I.e. at 45, look up at sky till you get 90, then look at the same scene, and its a 60-70 for a bit (5-10sec), and then it goes back down to 45, look up again. Its almost like there is a buffer or something that you are unloading when looking up, that fills back up.

 

Perhaps it is averaging the FPS? so when you look up it locks to 90 but it takes a moment for the average to get there, then when you look down the same in reverse. don't know really.

Link to comment
Share on other sites

Okay - I haven't done an extensive test or anything, but last night after doing some clean-ups and updates, the Persian Gulf really seemed to scream FPS-wise, unlike the Caucuses. If I didn't know better, I'd say they fixed PG but hadn't fixed Caucuses.

 

I also do not skimp too much on trees (I tend to set them to be somewhere between 60%-75%), so maybe that accounts for the differences. And maybe I was imagining it. It seemed like Dubai seemed to fly more smoothly and at a higher max frame rate (I'd forgot to switch my Pimax back down to 72 FPS, so it was running 90 FPS + Smart Smoothing), and I was even able to improve some settings. Going back to 72 FPS + Smart Smoothing in the TF-51 Free Flight scenario they had us testing with, it struggled to maintain a consistent 36 FPS with the same options.

Link to comment
Share on other sites

Just re-iteration of some things that have been proven by others' testing:

 

In VR:

 

1. MSAA has a serious negative effect on FPS.

2. Shadows on the ground and in the cockpit have a serious negative effect on FPS.

3. PD settings above 1.2 on Rift S or 1.5 on Rift CV1 have a serious negative effect on FPS (can't speak to all the others)

 

On my system with MSAA and Shadows OFF I can stay pinned at 80fps mode in nearly all situations. As soon as I enable either, I drop into 40fps ASW mode. With both enabled I begin to see some stutters. While the various other graphics settings do affect FPS they seem to be in the 1-3fps each. Turning ASW off makes it a stutter fest unless FPS are pinned on 80. Setting PD above 1.2 on my rift s kills FPS without any noticeable visual clarity benefit.

 

some other settings:

vsync: off

Civ traffic: off

ground clutter: 150

trees: 70%

Visibility range: High

Anis. Filtering: 16x*keeps the lines on the runway from fading out

Preload radius: 120000

Chimney smoke: 0

 

***Also enabling Windows game mode gave me a significant FPS bump (windows 1809)

 

Check your PD, shadows and MSAA settings.


Edited by unltd

Ryzen 5950x PBO | MSI RTX 3090 | 32gb 3600 RAM | 2 x SSD | Quest 3 | TM Warthog

Link to comment
Share on other sites

It looks to me like folks that aren't seeing any FPS improvement from the beta patch have graphics settings that are keeping their systems bogged down at or below 40-45 FPS. Check your PD, shadows and MSAA settings people.

Not true in my case. Also others have reported no improvements with FPS>45.

A warrior's mission is to foster the success of others.

i9-12900K | MSI RTX 3080Ti Suprim X | 128 GB Ram 3200 MHz DDR-4 | MSI MPG Edge Z690 | Samung EVO 980 Pro SSD | Virpil Stick, Throttle and Collective | MFG Crosswind | HP Reverb G2

RAT - On the Range - Rescue Helo - Recovery Tanker - Warehouse - Airboss

Link to comment
Share on other sites

Just re-iteration of some things that have been proven by others' testing:

 

In VR:

 

1. MSAA has a serious negative effect on FPS.

2. Shadows on the ground and in the cockpit have a serious negative effect on FPS.

3. PD settings above 1.2 on Rift S or 1.5 on Rift CV1 have a serious negative effect on FPS (can't speak to all the others)

.

 

Agreed...but more than anything else the thing that kills frames for me is looking at another hornet. The object count goes way up and my frames go the opposite direction. Try the instant action caucsaus Hornet ' ready on the ramp' mission and taxi down towards the parked hornets.

Check out my 'real world'

video series

 

 

[sIGPIC][/sIGPIC]

Link to comment
Share on other sites

Agreed...but more than anything else the thing that kills frames for me is looking at another hornet. The object count goes way up and my frames go the opposite direction. Try the instant action caucsaus Hornet ' ready on the ramp' mission and taxi down towards the parked hornets.

 

Experienced the same situation when in the "Cold and Dark" instant action, where there are Hornets parking at each side of me...

i9-9900K, G.Skill 3200 32GB RAM, AORUS Z390 Pro Wifi, Gigabyte Windforce RTX 2080 Ti, Samsung 960 Pro NVMe 512G + 860 Pro 1T, TM Warthog HOTAS, VKB T-Rudder, Samsung O+

F/A-18C, F-16C, A-10C, UH-1, AV-8B, F-14, JF-17, FC3, SA342 Gazelle, L-39, KA-50, CEII, Supercarrier Preordered. (Almost abandoned: CA - VR support please?)

PG, NTTR

Link to comment
Share on other sites

Agreed...but more than anything else the thing that kills frames for me is looking at another hornet. The object count goes way up and my frames go the opposite direction. Try the instant action caucsaus Hornet ' ready on the ramp' mission and taxi down towards the parked hornets.

 

That's not really ED fault, MSAA is brutally resource intensive. It's better than straight super sampling, but it's essentially executing 4x shaders per pixel where the coverage mask is non-zero. It produces great visuals but with deferred shading/lighting that DCS uses its going to kill fps, nothing they can do to change that. Post processing techniques are usually much easier on the GPU, but don't look very good in DCS. Perhaps that is something they could improve?

 

Honestly I don't think we are going to see any big rasterization changes until vulkan, as it doesn't make sense to put resources into it. Even then early ray tracing implementations in vulkan give a lot of reason to be optomistic, and open up a number of more modern approaches

just a dude who probably doesn't know what he's talking about

Link to comment
Share on other sites

NineLine was right. A problem with the oculus keyboard shortcuts (LCTR + KP1, kp2...) so all previous tests would be in ASW AUTO. :doh:

 

NEW TEST VERIFYING THAT ASW IS OFF. :music_whistling: (Using OculusDebugTools).

 

VR FPS:

 

Using the TF-51D Instant Action mission Flight Over Tbilisi the first few seconds

 

DCS BETA 2.5.5 = 81 - 85 FPS

DCS Stable 2.5.4 = 59 - 62 FPS

 

https://forums.eagle.ru/showpost.php?p=3955759&postcount=47

 

Yes VR improvements!!!, BUT.... I created a mission with an F18 in the same config, same position, altitude and direction as the tf51 on tiblisi, result:

 

DCS BETA 2.5.5 = 45 - 54 FPS

DCS Stable 2.5.4 = 45 - 48 FPS

 

:huh: :noexpression:

 

it goes better, but these improvements are marginal, ????????


Edited by Werewolf_fs
Link to comment
Share on other sites

it goes better, but these improvements are marginal, ????????

 

I don't think the improvements are marginal, but I'd say that the performance impact of the modules themselves is larger than that of the core engine, and as such the *perceived* performance difference is only minor when you're flying a full-fidelity module.

That's how I read those results, at least...

 

You can easily see that DCS runs far smoother when flying a FC3 module compared to say an F-18 or F-14.

 

Any performance gain is more than welcome, but I think the largest gain when it comes to the core engine itself, especially for us VR users, would come from support for a post-process anti-aliasing technique that is compatible with PBR/deferred shading.

That's not to say that a poorly-optimised map can't have a major performance impact. Look at Normandy...

Spoiler

Ryzen 9 5900X | 64GB G.Skill TridentZ 3600 | Gigabyte RX6900XT | ASUS ROG Strix X570-E GAMING | Samsung 990Pro 2TB + 960Pro 1TB NMVe | HP Reverb G2
Pro Flight Trainer Puma | VIRPIL MT-50CM2+3 base / CM2 x2 grip with 200 mm S-curve extension + CM3 throttle + CP2/3 + FSSB R3L + VPC Rotor TCS Plus base with SharKa-50 grip mounted on Monstertech MFC-1 | TPR rudder pedals

OpenXR | PD 1.0 | 100% render resolution | DCS "HIGH" preset

 

Link to comment
Share on other sites

This is quite confusing. So in the end, the 50% optimization is included. I thought it was coming and had not gotten into the update because of a problem in the build.

 

Now it seems that it is included.

 

I do not understand anything.

 

I say this because I have not noticed this 50% improvement in VR performance. 50% is a huge improvement and I am the same.

 

Well yes, I have improved because I have gone from the Oculus Rift CV1 to the Oculus Rift S. But this has nothing to do with the 50% improvement mentioned in the VR.


Edited by Fisu_MAD

YouTube Channel


Update: MSI Z790 Tomahawk, i9 13900k, DDR5 64GB 640 MHz, MSI 4090 Gaming X Trio, 970 EVO Plus 1TB SSD NVMe M.2 and 4 more, HOTAS TM Warthog, Meta Quest Pro

Link to comment
Share on other sites

@A2597 - that is similar to my results. The update made things slightly more smooth in repro but not much on overall FPS.

 

I think my biggest revelation from all of this is not using PD to change the SS in DCS. I have a WMR headset and I use Steam to address both my repro and super sampling. That has resulted in a much better experience for me. My only config in DCS for clarity is post processing the image via AF and MSAA, texture quality, and number of objects to render.

 

The question is - did we get all of the improvements, or just some? Are the improvements broken and when fixed will they provide even better performance that what is on the table now? Lots of questions, and waiting patiently for devs to work out some answers...

:pilotfly: Specs: I9-9900k; ROG Strix RTX 2080ti; Valve Index HMD; 32GB DDR4 3200 Ram; Samsung 970 EVO 1TB SSD; TM Warthog with pedals, 3 TM MFDs

Link to comment
Share on other sites

I don't think the improvements are marginal, but I'd say that the performance impact of the modules themselves is larger than that of the core engine, and as such the *perceived* performance difference is only minor when you're flying a full-fidelity module.

That's how I read those results, at least...

 

You can easily see that DCS runs far smoother when flying a FC3 module compared to say an F-18 or F-14.

 

Any performance gain is more than welcome, but I think the largest gain when it comes to the core engine itself, especially for us VR users, would come from support for a post-process anti-aliasing technique that is compatible with PBR/deferred shading.

That's not to say that a poorly-optimised map can't have a major performance impact. Look at Normandy...

 

I agree. The high-fidelity third-party modules would explain this discrepancy with the TF51 of the DCS core. Is true that the performance of my Hornet and Tomcat module has improved a bit, but I would like more improvement :(

Link to comment
Share on other sites

This is quite confusing. So in the end, the 50% optimization is included. I thought it was coming and had not gotten into the update because of a problem in the build.

 

Now it seems that it is included.

 

I do not understand anything.

 

I say this because I have not noticed this 50% improvement in VR performance. 50% is a huge improvement and I am the same.

 

Well yes, I have improved because I have gone from the Oculus Rift CV1 to the Oculus Rift S. But this has nothing to do with the 50% improvement mentioned in the VR.

 

Not all systems are seeing improvement. If you look at the test request thread I'd say 70-80% of users see no improvement or a tiny improvement which may be more of a figment of their imagination. But there are a few guys reporting differences in the 10s-30s of FPS which are likely real and it means its working for them for some reason.

New hotness: I7 9700k 4.8ghz, 32gb ddr4, 2080ti, :joystick: TM Warthog. TrackIR, HP Reverb (formermly CV1)

Old-N-busted: i7 4720HQ ~3.5GHZ, +32GB DDR3 + Nvidia GTX980m (4GB VRAM) :joystick: TM Warthog. TrackIR, Rift CV1 (yes really).

Link to comment
Share on other sites

... of their imagination. But there are a few guys reporting differences in the 10s-30s of FPS which are likely real and it means its working for them for some reason.

 

I read post from ED Staff 3 days later from patch release saying that this optimization wasnt included. Now seems that this is. This is confusing i said.


YouTube Channel


Update: MSI Z790 Tomahawk, i9 13900k, DDR5 64GB 640 MHz, MSI 4090 Gaming X Trio, 970 EVO Plus 1TB SSD NVMe M.2 and 4 more, HOTAS TM Warthog, Meta Quest Pro

Link to comment
Share on other sites

Agreed...but more than anything else the thing that kills frames for me is looking at another hornet. The object count goes way up and my frames go the opposite direction. Try the instant action caucsaus Hornet ' ready on the ramp' mission and taxi down towards the parked hornets.

 

In my OB (not the one that was released today, but the previous), using HP Reverb, it starts at 45FPS. But doesn't dip any lower. And as soon as i leave the ramp, it shoots up to 80-90 FPS.

hsb

HW Spec in Spoiler

---

 

i7-10700K Direct-To-Die/OC'ed to 5.1GHz, MSI Z490 MB, 32GB DDR4 3200MHz, EVGA 2080 Ti FTW3, NVMe+SSD, Win 10 x64 Pro, MFG, Warthog, TM MFDs, Komodo Huey set, Rverbe G1

 

Link to comment
Share on other sites

With these settings in my Odyssey I'm actually getting 90 fps over Caucasus instant action missions, TF-51 over Tiblisi, Harrier weapons practice, Hornet weapons practice.

 

vrreso.JPG.26cc0fcc1856aab1c43e340c45d4c6bf.JPG

 

Screen_190622_005246.thumb.png.30652d8c2a235530aa4fe9eed31a9dc4.png

 

Screen_190622_005241.thumb.png.0f98fb967171bfedeb20cdc601e9222e.png

 

Can't really be bothered to install stable release, but these framerates were impossible to achieve before the VR optimizations.

 

It's a compromise to lower my pixel density and fly without shadows.. but at the same time it's kind of like "OMG 90 FPS VR in Dcs World".. finally all this expensive hardware paid off lol

 

raising PD to even 1.2 i get lows in the 70's.. 1.5PD im seeing around 60-75 fps.

 

I like the higher pixel density but to be frank motion reprojection makes aircraft flying around me look weird sometimes.. so I think Im going to stick with 1.0 even though it's not quite as sharp a picture as higher pixel density levels.

 

*Edit- now I'm sitting here wondering if we're also going to see a nice performance bump from Vulkan whenever that gets here. Might be able to get away with higher pixel density values with my setup or a higher res headset.

 

*edit* - just did a campaign mission for the A-10C, which I won't name the campaign or mission as not to spoil anything but flying over areas with a nice amount of units on the ground tanked my framerate even at the settings described above.. so going back to motion reprojection and higher pixel density for now.


Edited by Headwarp
Spoiler

Win 11 Pro, z790 i9 13900k, RTX 4090 , 64GB DDR 6400GB, OS and DCS are on separate pci-e 4.0 drives 

Sim hardware - VKB MCG Ultimate with 200mm extension, Virpil T-50CM3 Dual throttles.   Blackhog B-explorer (A), TM Cougar MFD's (two), MFG Crosswinds with dampener.   Obutto R3volution gaming pit.  

 

Link to comment
Share on other sites

performance wise this latest patch introduced the worst performance i ever had with DCS, especially in VR. i'm on HTC Vive and flown the last 2 years more or less with stable frames between 40-60. my settings are already dialed down for VR, no shadows, medium view distance etc. with the latest patch fps suffer a lot with rising object count. having a few static objects (2-3 f18c, hawkeye, s3b) on the stanton and looking around in the hornet cockpit on the ramp gives me fps in teens and loads of reprojection. especially i feel the lower parts of the cockpit suffer from fps loss. also some black screen flashes and massiv stutter. it gets worse with several contacts on datalink SA screen. i changed nothing on my system since the last 2.5.4 OB patch. cleaning metashaders and fxo changed nothing. would love to report a change to the better, but unfortunatly it's the opposite.

nvidia 430.86, i7-4.8Ghz, 32GB RAM, GTX 1080Ti, htc vive

Link to comment
Share on other sites

performance wise this latest patch introduced the worst performance i ever had with DCS, especially in VR. i'm on HTC Vive and flown the last 2 years more or less with stable frames between 40-60. my settings are already dialed down for VR, no shadows, medium view distance etc. with the latest patch fps suffer a lot with rising object count. having a few static objects (2-3 f18c, hawkeye, s3b) on the stanton and looking around in the hornet cockpit on the ramp gives me fps in teens and loads of reprojection. especially i feel the lower parts of the cockpit suffer from fps loss. also some black screen flashes and massiv stutter. it gets worse with several contacts on datalink SA screen. i changed nothing on my system since the last 2.5.4 OB patch. cleaning metashaders and fxo changed nothing. would love to report a change to the better, but unfortunatly it's the opposite.

nvidia 430.86, i7-4.8Ghz, 32GB RAM, GTX 1080Ti, htc vive

 

did you try running "dcs_updater.exe repair" from cmd prompt or using the start menu shortcut?

Spoiler

Win 11 Pro, z790 i9 13900k, RTX 4090 , 64GB DDR 6400GB, OS and DCS are on separate pci-e 4.0 drives 

Sim hardware - VKB MCG Ultimate with 200mm extension, Virpil T-50CM3 Dual throttles.   Blackhog B-explorer (A), TM Cougar MFD's (two), MFG Crosswinds with dampener.   Obutto R3volution gaming pit.  

 

Link to comment
Share on other sites

... cleaning metashaders and fxo changed nothing...

 

 

I noticed a crapload of files now in metashaders2 folder after flying in Caucasus map. The first few flights I made after this update were in the Persian Gulf area and metashaders2 folder had only one file in it. After flying a mission in Caucasus the folder filled up with many .meta2 files.

 

Is this part of the optimization for Caucasus?

 

I did notice an improvement in smoothness when flying at tree tops in Caucasus. On the second flight over the same area basically there were no stutters... there was one "blink" at one point and that was it:) The FPS didn't change a lot. I usually fly with Fraps running at 45 and it acts as a frame limiter so the fps never goes above but... my settings are maxed. PD for my Rift is set to 1.5 and MSAA x2:noexpression: (anxiously waiting to upgrade to new VR... Pimax5+? Reverb? I don't know yet)

I9-7900x @4.1 32gigs @3600, 2 ssd's (raid0... for other purposes:smilewink:)

Link to comment
Share on other sites

New to VR here on the Rift S. Been following this thread and the VR test results thread trying various settings for a couple of weeks here. I’ve disabled ASW via the Oculus Tray Tool, the Oculus Debug Tool, and NVIDIA settings... I’ve experimented with the hotkeys in game to disable them as well. I am locked at 40 FPS no matter the settings, unless I go above 1.6PD, then the FPS goes below 40. Still testing over here and monitoring these threads.

 

 

Sent from my iPhone using Tapatalk

sendaero

twitch

youtube

Link to comment
Share on other sites

New to VR here on the Rift S. Been following this thread and the VR test results thread trying various settings for a couple of weeks here. I’ve disabled ASW via the Oculus Tray Tool, the Oculus Debug Tool, and NVIDIA settings... I’ve experimented with the hotkeys in game to disable them as well. I am locked at 40 FPS no matter the settings, unless I go above 1.6PD, then the FPS goes below 40. Still testing over here and monitoring these threads.

 

 

Sent from my iPhone using Tapatalk

 

Hi mrushforth,

 

Without more details on your system specifications, GPU, and DCS setting ist is prttey hard to suggest anything specific.

 

The link below is somewhat dated but does a pretty good job of explaining what settings have the most impact on fps.

https://www.bing.com/videos/search?q=Youtube+All+things+VR&view=detail&mid=5645616A5D472B510C9A5645616A5D472B510C9A&FORM=VIRE

 

Also, I would suggest looking through the thread below to find a system similar to yours and try those settings.

https://forums.eagle.ru/showthread.php?t=243521

 

S!

----------------

AKA_Clutter

 

Win 10 Pro, Intel i7 8700k @4.6 GHz, EVGA RTX 3080  FTW, Ultra 64 GB G.Skill DDR4 3600 RAM, Acer 27" flat screen, Oculus Rift S, HP Reverb G2, CH Fighterstick, Pro Throttle and Pro Rudder Pedals HOTAS, TM Warthog HOTAS, MFG Rudder Pedals, TrackIR 5 Pro w/Vector Expansion, PointCTRL.

Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...