Jump to content

Another win for Fholger


Sr.

Recommended Posts

I've been testing this a bit both with NIS and without and the only meaningful filter that makes a difference for me remains Reshade VR using AMD CAS technology, none of this FSR stuff provide adequate image quality for my personal use.  With Reshade VR, the texts on F/A-18's AMPCD on are very crisp and completely legible on the Vive Pro 2 and so are the distant ground details especially buildings and bridges.  Without Reshade the AMPCD texts are a bit blurry and distant buildings and bridges really lack definition (with or without FSR).  I'm going to try more various combinations of FSR with Reshade to see if it can shave off some frametime while retaining similar graphic fidelity, but I doubt I'll see anything groundbreaking at this point (I consider Reshade VR groundbreaking since it completely changes the way I look at DCS VR).

PC: 5800X3D/4090, 11700K/3090, 9900K/2080Ti.

Joystick bases: TMW, VPC WarBRD, MT50CM2, VKB GFII, FSSB R3L

Joystick grips: TM (Warthog, F/A-18C), Realsimulator (F-16SGRH, F-18CGRH), VKB (Kosmosima LH, MCG, MCG Pro), VPC MongoosT50-CM2

Throttles: TMW, Winwing Super Taurus, Logitech Throttle Quadrant, Realsimulator Throttle (soon)

VR: HTC Vive/Pro, Oculus Rift/Quest 2, Valve Index, Varjo Aero, https://forum.dcs.world/topic/300065-varjo-aero-general-guide-for-new-owners/

Link to comment
Share on other sites

2 hours ago, Supmua said:

I've been testing this a bit both with NIS and without and the only meaningful filter that makes a difference for me remains Reshade VR using AMD CAS technology, none of this FSR stuff provide adequate image quality for my personal use.  With Reshade VR, the texts on F/A-18's AMPCD on are very crisp and completely legible on the Vive Pro 2 and so are the distant ground details especially buildings and bridges.  Without Reshade the AMPCD texts are a bit blurry and distant buildings and bridges really lack definition (with or without FSR).  I'm going to try more various combinations of FSR with Reshade to see if it can shave off some frametime while retaining similar graphic fidelity, but I doubt I'll see anything groundbreaking at this point (I consider Reshade VR groundbreaking since it completely changes the way I look at DCS VR).

I agree. FSR introduces some awful terrain and cloud shimmering for me that makes the performance boost not worth it. Reshade VR is amazing though.

PC Hardware: Asus Maximus XIII Hero / i9-10900K / 64 GB DDR4 3600 MHz / NVIDIA RTX 3090 FE / Samsung 960 NVMe / LG OLED48CX / Reverb G2

Flight Controls: Thrustmaster HOTAS Warthog / VKB T-RUDDER MKIV / Cougar MFD Set

DCS Modules: F-14 Tomcat / F/A-18C Hornet / AV-8B NA / F-5E Tiger II / A-10C Warthog / F-16C Viper / UH-1 Huey / LOFC3 / Spitfire LF Mk. IX / P-51D Mustang / WWII Assets Pack / Normandy 1944 / NTTR Map / Persian Gulf Map / Syria Map

Link to comment
Share on other sites

I was also fairly unimpressed with the results of FSR until I bumped the Steam SS up from 90 to 130 on my Reverb. Running FSR and Simplex 2.1 combined is giving some really nice results for me now in clarity and performance.  

Asus Z790 PLUS WIFI D4, 13700K RTX 4090 FE, M2. HP Reverb G2.

Link to comment
Share on other sites

2 hours ago, oz555 said:

I was also fairly unimpressed with the results of FSR until I bumped the Steam SS up from 90 to 130 on my Reverb. Running FSR and Simplex 2.1 combined is giving some really nice results for me now in clarity and performance.  

Right, this is different than ReShade.  It's not sharpen filter.  It does come with its own sharpen but it's very rough sharpen.  I turn its sharpen way down and use ReShade sharpen instead.

If your SteamVR vertical resolution is 3000 and set render scale to 0.67, your frames will be rendered at 2010 vertical resolution and then upsample to 3000.  But even with upsampling algorithm, it will look worse if you don't compensate.  You need to compensate by increasing SteamVR resolution.  How much you increase will depend on whether you want quality or performance.  In this example, if you increase SteamVR resolution to 4477 vertical resolution, you're basically back to where you were.  FSR will render at 3000.  There will be performance hit from upsampling algorithm however so I wouldn't recommend going that high.  And you probably don't need to.

Link to comment
Share on other sites

has anyone tried up sampling directly in the MOD rather than in SS e.g. render scale of 1.2 or whatever 

SYSTEM SPECS: Hardware Intel Corei7-12700KF @ 5.1/5.3p & 3.8e GHz, 64Gb RAM, 4090 FE, Dell S2716DG, Virpil T50CM3 Throttle, WinWIng Orion 2 & F-16EX + MFG Crosswinds V2, Varjo Aero
SOFTWARE: Microsoft Windows 11, VoiceAttack & VAICOM PRO

1569924735_WildcardsBadgerFAASig.jpg.dbb8c2a337e37c2bfb12855f86d70fd5.jpg

Link to comment
Share on other sites

7 hours ago, speed-of-heat said:

has anyone tried up sampling directly in the MOD rather than in SS e.g. render scale of 1.2 or whatever 

7 hours ago, speed-of-heat said:

has anyone tried up sampling directly in the MOD rather than in SS e.g. render scale of 1.2 or whatever 

Where would you start with SS 50%?


Edited by Sr.
ok...I can't type a reply for some reason

Ryzen 7 5800X3D | 64GB DDR4 3600| MSI RTX 4080 16GB Ventus 3X OC  | Samsung 970 Evo 2TB NVME | HP Reverb G2 | DIY Head Tracker Cap | Logitech X-56 throttle | VKB NXT Premium |  Win 11

"Any sufficiently advanced technology is indistinguishable from magic."

--Arthur C Clark

Link to comment
Share on other sites

8 hours ago, Taz1004 said:

Right, this is different than ReShade.  It's not sharpen filter.  It does come with its own sharpen but it's very rough sharpen.  I turn its sharpen way down and use ReShade sharpen instead.

If your SteamVR vertical resolution is 3000 and set render scale to 0.67, your frames will be rendered at 2010 vertical resolution and then upsample to 3000.  But even with upsampling algorithm, it will look worse if you don't compensate.  You need to compensate by increasing SteamVR resolution.  How much you increase will depend on whether you want quality or performance.  In this example, if you increase SteamVR resolution to 4477 vertical resolution, you're basically back to where you were.  FSR will render at 3000.  There will be performance hit from upsampling algorithm however so I wouldn't recommend going that high.  And you probably don't need to.

My render scale is 0.77. Just tried Steam SS at 100, 130 and 150 and yes, the image seems to be the same, so the question is what is the max Steam SS we should use before it becomes  wasteful?

Have just tried adding the ReShade also. Agreed, the sharpening is a much higher quality.  


Edited by oz555

Asus Z790 PLUS WIFI D4, 13700K RTX 4090 FE, M2. HP Reverb G2.

Link to comment
Share on other sites

8 hours ago, speed-of-heat said:

has anyone tried up sampling directly in the MOD rather than in SS e.g. render scale of 1.2 or whatever 

using 1.2 in FSR and 65 or 75 in SteamVR SS - you get problems with the clouds...but some kind of antialiasing.
With higher SteamVR SS the performance hit is too high for me.

FSR@0.83, SteamVR SS 100% and for game 84% with ReShade gives me good results for multiplayer.

FC3 | F-14A/B | F-16C | SA342  R5 5600X | 32GB | RX 6800XT | TM Cougar | T-Rudder mk.IV | HP Reverb G2

Thrustmaster Cougar Throttle USB standalone MOD

Link to comment
Share on other sites

I’m still testing various combinations and the general idea is to use Reshade for sharpening and FSR purely for supersampling.  Still trying to figure out the difference between per app setting at 100% Steam SS with low FSR number vs lower Steam SS with high FSR number—70% Steam SS on my setup is still over 3K x 3K per screen.  The idea of having to run two graphics mods at the same time doesn’t sit too well with me though.

PC: 5800X3D/4090, 11700K/3090, 9900K/2080Ti.

Joystick bases: TMW, VPC WarBRD, MT50CM2, VKB GFII, FSSB R3L

Joystick grips: TM (Warthog, F/A-18C), Realsimulator (F-16SGRH, F-18CGRH), VKB (Kosmosima LH, MCG, MCG Pro), VPC MongoosT50-CM2

Throttles: TMW, Winwing Super Taurus, Logitech Throttle Quadrant, Realsimulator Throttle (soon)

VR: HTC Vive/Pro, Oculus Rift/Quest 2, Valve Index, Varjo Aero, https://forum.dcs.world/topic/300065-varjo-aero-general-guide-for-new-owners/

Link to comment
Share on other sites

5 hours ago, oz555 said:

My render scale is 0.77. Just tried Steam SS at 100, 130 and 150 and yes, the image seems to be the same, so the question is what is the max Steam SS we should use before it becomes  wasteful?

Have just tried adding the ReShade also. Agreed, the sharpening is a much higher quality.  

Same?  Between 100, 130 and 150?  Sure it's working?  FSR or NIS is not meant to be supersampling.  You can try higher renderscale than 1 but really it's designed to be lower.  It's algorithm is designed to clean up artifacts when images are upsampled from lower resolution images.  There's nothing to clean up if it's downsampled.  And SteamVR doesn't do any algorithm for upsampling.  So in a way, you're basically not using FSR if you use higher than 1 render scale.

As for the math, take your Per App SteamVR vertical resolution and multiply by Render Scale.  That's what FSR is rendering at.  Vertical resolution divided by Render Scale should be max.  With some performance hit from FSR of course.  Somewhere between the max and original resolution should be good balance of performance and quality.


Edited by Taz1004
  • Like 2
Link to comment
Share on other sites

OK FSR actually increases my frametime by around 1 ms.  With FSR my frametime even in the main menu is >11 ms which causes my headset to drop the framerate to 45 fps, without it my frametime is 10 which gives 90 fps.  During actual gameplay the difference is between 0.5-1ms, with exact same SteamVR settings.  Image quality is actually better without it, less muddy in the distant.

 

it also seems to alter pixel rendering numbers in per app SS, Steam SS at 100% shows 3800x3800 pixels per screen which is crazy. 


Edited by Supmua

PC: 5800X3D/4090, 11700K/3090, 9900K/2080Ti.

Joystick bases: TMW, VPC WarBRD, MT50CM2, VKB GFII, FSSB R3L

Joystick grips: TM (Warthog, F/A-18C), Realsimulator (F-16SGRH, F-18CGRH), VKB (Kosmosima LH, MCG, MCG Pro), VPC MongoosT50-CM2

Throttles: TMW, Winwing Super Taurus, Logitech Throttle Quadrant, Realsimulator Throttle (soon)

VR: HTC Vive/Pro, Oculus Rift/Quest 2, Valve Index, Varjo Aero, https://forum.dcs.world/topic/300065-varjo-aero-general-guide-for-new-owners/

Link to comment
Share on other sites

20 hours ago, Supmua said:

OK FSR actually increases my frametime by around 1 ms.  With FSR my frametime even in the main menu is >11 ms which causes my headset to drop the framerate to 45 fps, without it my frametime is 10 which gives 90 fps.  During actual gameplay the difference is between 0.5-1ms, with exact same SteamVR settings.  Image quality is actually better without it, less muddy in the distant.

 

it also seems to alter pixel rendering numbers in per app SS, Steam SS at 100% shows 3800x3800 pixels per screen which is crazy. 

 

Thats weird.  What does the openvr_mod.log say its doing with the textures?    A lot of times using ReShade will cause the OpenFSR to not work due to using some of the same .dll's.

MSI X790 Carbon Wifi, 13900K, MSI 4090 Suprim X, 64gb DDR5 6400CL30, HP Reverb G2, VKB NXT Evo, VKB STECS Max, Thrustmaster rudder pedals

Link to comment
Share on other sites

17 hours ago, Taz1004 said:

Something's not right.  It has no control over SteamVR setting.

It seems that per app SS is now the percentage of what you set globally.  Could've sworn it was 1:1 in the past, so 100% per app means 140% if that's what the global SS is.

PC: 5800X3D/4090, 11700K/3090, 9900K/2080Ti.

Joystick bases: TMW, VPC WarBRD, MT50CM2, VKB GFII, FSSB R3L

Joystick grips: TM (Warthog, F/A-18C), Realsimulator (F-16SGRH, F-18CGRH), VKB (Kosmosima LH, MCG, MCG Pro), VPC MongoosT50-CM2

Throttles: TMW, Winwing Super Taurus, Logitech Throttle Quadrant, Realsimulator Throttle (soon)

VR: HTC Vive/Pro, Oculus Rift/Quest 2, Valve Index, Varjo Aero, https://forum.dcs.world/topic/300065-varjo-aero-general-guide-for-new-owners/

Link to comment
Share on other sites

1 hour ago, Taterbootz said:

Thats weird.  What does the openvr_mod.log say its doing with the textures?    A lot of times using ReShade will cause the OpenFSR to not work due to using some of the same .dll's.

no idea, since I have trashed the whole thing since.  The bottomline is I got 45 fps in the main menu instead of 90 fps due to 1 ms diff in frametime with the openvr_api.dll from FSR compared to the default one, tried swapping back and forth 3 times and results were the same so I decided to stop since there are better things to do.

PC: 5800X3D/4090, 11700K/3090, 9900K/2080Ti.

Joystick bases: TMW, VPC WarBRD, MT50CM2, VKB GFII, FSSB R3L

Joystick grips: TM (Warthog, F/A-18C), Realsimulator (F-16SGRH, F-18CGRH), VKB (Kosmosima LH, MCG, MCG Pro), VPC MongoosT50-CM2

Throttles: TMW, Winwing Super Taurus, Logitech Throttle Quadrant, Realsimulator Throttle (soon)

VR: HTC Vive/Pro, Oculus Rift/Quest 2, Valve Index, Varjo Aero, https://forum.dcs.world/topic/300065-varjo-aero-general-guide-for-new-owners/

Link to comment
Share on other sites

12 hours ago, Supmua said:

It seems that per app SS is now the percentage of what you set globally.  Could've sworn it was 1:1 in the past, so 100% per app means 140% if that's what the global SS is.

It always was cumulative.  Why I always recommend using the vertical resolution number in Per App.  Percent value is different per headset and has no meaning other than to you.


Edited by Taz1004
Link to comment
Share on other sites

25 minutes ago, Supmua said:

It seems that per app SS is now the percentage of what you set globally.  Could've sworn it was 1:1 in the past, so 100% per app means 140% if that's what the global SS is.

It will do that if you set global to Auto instead of Custom


Edited by Sr.

Ryzen 7 5800X3D | 64GB DDR4 3600| MSI RTX 4080 16GB Ventus 3X OC  | Samsung 970 Evo 2TB NVME | HP Reverb G2 | DIY Head Tracker Cap | Logitech X-56 throttle | VKB NXT Premium |  Win 11

"Any sufficiently advanced technology is indistinguishable from magic."

--Arthur C Clark

Link to comment
Share on other sites

1 hour ago, Supmua said:

no idea, since I have trashed the whole thing since.  The bottomline is I got 45 fps in the main menu instead of 90 fps due to 1 ms diff in frametime with the openvr_api.dll from FSR compared to the default one, tried swapping back and forth 3 times and results were the same so I decided to stop since there are better things to do.

That sucks.  It was more than likely a conflict with reshade and FSR.  I get a massive performance increase running FSR with minimal visual issues. 

MSI X790 Carbon Wifi, 13900K, MSI 4090 Suprim X, 64gb DDR5 6400CL30, HP Reverb G2, VKB NXT Evo, VKB STECS Max, Thrustmaster rudder pedals

Link to comment
Share on other sites

On 11/19/2021 at 3:46 AM, speed-of-heat said:

has anyone tried up sampling directly in the MOD rather than in SS e.g. render scale of 1.2 or whatever 

Just tried this with 1.3 and 1.5 and it wouldnt upscale past some god awful resolution (like 1544x1530 or something) no matter how I set the PD and SS.

MSI X790 Carbon Wifi, 13900K, MSI 4090 Suprim X, 64gb DDR5 6400CL30, HP Reverb G2, VKB NXT Evo, VKB STECS Max, Thrustmaster rudder pedals

Link to comment
Share on other sites

OK did some more testing, I guess persistency paid off.

 

It turns out there was a problem with my NVIDIA driver and Image Scaling which didn't show on the control panel, so I guess it never worked properly.  Had to download the Hotfix version, turned off HDR, and restored to default settings, before image scaling option would show up in the control panel.

 

Next I tried the system wide NIS setting, but it crashed SteamVR right away, so had to disable this.  Then I tried the FSR 2.0 with DCS, and found that it reduced my frametime to 8ms in the main menu which is pretty good (normally 10-11ms).  Image quality is relatively meh even at 0.8 scaling, but tolerable at 0.9.  FSR Sharpening is bad so I'm sticking with Reshade VR for this.  With Reshade VR sharpening (CAS 1.0, Intensity 1.2), the quality is very close to my usual settings except the MFD texts aren't as crisp but still completely legible on the Hornet.  This way I shaved off about 3ms in frametime during gameplay, which can be huge in VR depending on the situations.  IQ still not as good as no-scaling + Rehade VR (might not even be noticeable on lower-res headsets compared to the VP2), but not bad either.


Edited by Supmua

PC: 5800X3D/4090, 11700K/3090, 9900K/2080Ti.

Joystick bases: TMW, VPC WarBRD, MT50CM2, VKB GFII, FSSB R3L

Joystick grips: TM (Warthog, F/A-18C), Realsimulator (F-16SGRH, F-18CGRH), VKB (Kosmosima LH, MCG, MCG Pro), VPC MongoosT50-CM2

Throttles: TMW, Winwing Super Taurus, Logitech Throttle Quadrant, Realsimulator Throttle (soon)

VR: HTC Vive/Pro, Oculus Rift/Quest 2, Valve Index, Varjo Aero, https://forum.dcs.world/topic/300065-varjo-aero-general-guide-for-new-owners/

Link to comment
Share on other sites

1 hour ago, Taz1004 said:

System wide NIS?  You mean from Nvidia control panel?  That needs to be disabled and enable NIS from FSR 2.0 config if you want sharper MFD.

Quality loss is expected if you don't compensate by raising SteamVR SS.  Goal is to go as low as possible with FSR (or NIS) and raise SteamVR SS to get back to previous quality.  Or performance depening on what's more important to you.  To benefit most from NIS algorithm.

Yep control panel, and it crashed my Vive console and SteamVR and had to be disabled.  When it's on if you have PC icon on the desktop it will display NIS on top of the icon, which is neat.  The on-screen fonts became jagged right away at default sharpening on 4K TV, had to reduce sharpening to 0.25 in the control panel for texts to look near normal.  Of course it doesn't work with my VR, but in 2D games I did gain some FPS while retaining mostly the same picture quality. 

 

VR is another story because the screen is right next to my eyes, any type of artifacts are easily seen.

 

I'm already rendering around 6K resolution (3000x3000 pixels per panel) on my headset can't go any higher than this because framerate suffers.  The thing about the Vive Pro 2 is that it produces great images at 2:1 reprojection but at 3:1 there are more image retention artifacts so I'm trying to avoid this at all cost.

PC: 5800X3D/4090, 11700K/3090, 9900K/2080Ti.

Joystick bases: TMW, VPC WarBRD, MT50CM2, VKB GFII, FSSB R3L

Joystick grips: TM (Warthog, F/A-18C), Realsimulator (F-16SGRH, F-18CGRH), VKB (Kosmosima LH, MCG, MCG Pro), VPC MongoosT50-CM2

Throttles: TMW, Winwing Super Taurus, Logitech Throttle Quadrant, Realsimulator Throttle (soon)

VR: HTC Vive/Pro, Oculus Rift/Quest 2, Valve Index, Varjo Aero, https://forum.dcs.world/topic/300065-varjo-aero-general-guide-for-new-owners/

Link to comment
Share on other sites

We don't have the exact same setup so please don't assume that your settings will also apply to me.  As I posted before typical high setting such as 0.77 doesn't look good on the Vive Pro 2, from various trial runs I need around 0.87-0.9 to get the quality close to what I have with just ReshadeVR (cutoff criteria is clarity of Hornet's AMPCD screen upon glancing down, if texts are fuzzy then it's not good enough).

 

It might work wonderfully in 2D, but on a high-res VR headset right next to the face it's pretty easy to spot artifacts.  So anyway, my exact Steam SS right now is 3280x3280 (for 100% SS) so the rendering is still high at 0.87 scale (2854x2854) to get the image quality that I want.  I don't see the point of raising SS further since the frametime difference is already small (2-3ms) and it will get even smaller to the point of why even do this if it's not going to make a big difference in gameplay.  Anyway 3ms improvement is fine, can't complain.

 

EDIT: Ran several MP missions with my bud and it's all good, it's pretty solid actually no crash or stutters.

 

EDIT: After several more ground missions, I'm back to just Reshade only for now as the clarity and sharpness are still the best and this is the main reason that I bought a hi-res VR headset in the first place (not for fps).  Image quality just can't compare with the FSR option, very evident in ground spotting and low level bombing.


Edited by Supmua

PC: 5800X3D/4090, 11700K/3090, 9900K/2080Ti.

Joystick bases: TMW, VPC WarBRD, MT50CM2, VKB GFII, FSSB R3L

Joystick grips: TM (Warthog, F/A-18C), Realsimulator (F-16SGRH, F-18CGRH), VKB (Kosmosima LH, MCG, MCG Pro), VPC MongoosT50-CM2

Throttles: TMW, Winwing Super Taurus, Logitech Throttle Quadrant, Realsimulator Throttle (soon)

VR: HTC Vive/Pro, Oculus Rift/Quest 2, Valve Index, Varjo Aero, https://forum.dcs.world/topic/300065-varjo-aero-general-guide-for-new-owners/

Link to comment
Share on other sites

1 hour ago, Roger01 said:

Is this mod works with Airlink?
I already test that, but I see no difference (fps and image quality)

Thx!

Is this mod works with Airlink?
I already test that, but I see no difference (fps and image quality)

Thx!

The mod uses openvr_api.dll, so you need to force the Quest 2 to use OpenVR with DCS (DCS default mode for Oculus/Meta's headsets is Oculus runtime).  I haven't done this in ages but I think OVGME can do this.  I think OVR Toolkit can do it also but that's payware.  There will be a performance hit though, as Oculus runtime is better optimized for their headsets.


Edited by Supmua
  • Thanks 1

PC: 5800X3D/4090, 11700K/3090, 9900K/2080Ti.

Joystick bases: TMW, VPC WarBRD, MT50CM2, VKB GFII, FSSB R3L

Joystick grips: TM (Warthog, F/A-18C), Realsimulator (F-16SGRH, F-18CGRH), VKB (Kosmosima LH, MCG, MCG Pro), VPC MongoosT50-CM2

Throttles: TMW, Winwing Super Taurus, Logitech Throttle Quadrant, Realsimulator Throttle (soon)

VR: HTC Vive/Pro, Oculus Rift/Quest 2, Valve Index, Varjo Aero, https://forum.dcs.world/topic/300065-varjo-aero-general-guide-for-new-owners/

Link to comment
Share on other sites

  • Recently Browsing   0 members

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