FPS comparision: DCS 2.0.5 vs. 2.1.0 over Vegas (1080 Ti)

Reading a lot about VR performance in 2.1.0, I wanted to quantify this a bit more.

Therefore, I monitored the FPS for a saved track in 2.0.5 and 2.1.0 and compared them.

This is a free flight over Las Vegas comparing mainly MSAA and deferred shading settings while leaving the others constant.

Pixel density is set to 1.8 and the remaining settings are in the spoiler. Note that shadows are completely off and visibility range is low.

MSAA settings with deferred shading OFF

(click on image to enlarge)

Using identical settings in 2.0.5 (black line) and 2.1.0 with MSAA=4x (red) leads to quite a bit FPS loss (~9 FPS). The average is going from 87 down to 78 FPS.

With MSAA=2x (green) the average frame rate is 85 and with MSAA=Off (blue) it goes back up to 87.

MSAA settings with deferred shading ON

(click on image to enlarge)

This graph shows the FPS when deferred shading is ON in DCS 2.1.0. Here, the MSAA setting hits quite hard. On average the FPS drop down to 37.

With MSAA=2x there is still a significant performance loss. Average FPS in this case are 48.

However, by setting MSAA to off, the deferred shading has nearly no impact on the FPS. The average value increases back to 86.

Bottom Line

So for the case I considered here, it seems that MSAA is hitting relatively hard on the frame rates. This is particularly true, if deferred shading is activated.

Deferred shading without MSAA, however, reduces the performance only slightly.

So mixing MSAA with deferred shading does not seem to be a good idea at the moment ;)

You can watch the track I used on YT, where I did a comparison between my old 980 Ti and my new 1080 Ti. The Vegas part starts at 3m41s.

Very good post. I came to the same conclusion after a couple hours of testing.

This is an excellent analysis, thanks!

What to use instead of msaa ? Because even it is on , it looks jaggied

Yeah but the fact is performance you`re measuring is in the best scenario. In reality where some units are, helis and planes, you are seeing spitfires jiddering pipper all the time :/

Yeah but the fact is performance you`re measuring is in the best scenario. In reality where some units are, helis and planes, you are seeing spitfires jiddering pipper all the time :/

Partly true. There are no units - that's right, but I'm flying over a heavily populated area. I would not call this the best scenario. But that's actually not the point.

The point was to compare the performance of 2.0.5 and 2.1.0 with identical settings to a see if there is a difference in performance due to the new lighting and shaders.

You can add AI units etc. and you will see less FPS, of course. But this FPS drop will be (roughly) the same in both versions. So all conclusions stay valid.

i tried msaa off and yes it provided more than x2 fps boost but visuality is worse than terrible without aa

i tried msaa off and yes it provided more than x2 fps boost but visuality is worse than terrible without aa

Yes, just noticed that too after some experiments.

I leave MSAA set to x2 now

Thanks

I think I'm noticing everyone who's claiming great framerates has a 1080ti. :D Great post.

Yes, just noticed that too after some experiments.

 

I leave MSAA set to x2 now

 

Thanks

Wags just posted that turning msaa off in DCS and enabling it through your graphics card may help.

Try it and see what happens!

TJ

Sent from my iPhone using Tapatalk

By what name does MSSA go under in the Nvidia control panel? I can find MFAA but no MSSA!

By what name does MSSA go under in the Nvidia control panel? I can find MFAA but no MSSA!

Multisample anti-aliasing (MSAA)

Just select 2X under the Antialiasing - Setting option.

nVidia Inspector is much clearer with heaps more options (Sparse Grid Supersampling is the hoily grail of jaggie-removal!)

Wags just posted that turning msaa off in DCS and enabling it through your graphics card may help.

 

Try it and see what happens!

 

TJ

 

 

Sent from my iPhone using Tapatalk

I thought Video Card settings were ignored by VR Headset?

You are totally right, disabling msaa in DCS got me like 50% boost...the problem is nvidia control panel AA settings doesnt seem to do anything in my case...any ideas??

You are totally right, disabling msaa in DCS got me like 50% boost...the problem is nvidia control panel AA settings doesnt seem to do anything in my case...any ideas??

Don't think NCP will have any effect on what is displayed in the VR headset. I tried testing both AA and AF different settings using NCP rather than setting in the game GUI, and they had no effect at all. I went back to application controlled for these and set them in the game.

Doing it that way though might help the monitor users.

Don't think NCP will have any effect on what is displayed in the VR headset. I tried testing both AA and AF different settings using NCP rather than setting in the game GUI, and they had no effect at all.

+1. Played around with the settings in NCP cranking them up to max but did not see any improvement in image quality at all.

Sorry, I was refering to non VR use...

 

EDIT: Also, dont forget to use anisotropic filtering from your graphics card control panel, not only for the performance but also because it seems that it doesnt work at all otherwise.

As we are in a VR forum, thought I would mention these do not have any effect on the display in the VR headset... certainly some nice tips for monitor users undoubtedly...

You are right mate, already posted in a better related thread! :thumbup:

I think I'm noticing everyone who's claiming great framerates has a 1080ti. :D Great post.

Lol I just said that earlier in another thread.

new rendering engine doing new gen rendering techniques requires high end video cards to get good frames, unheard of.

new rendering engine doing new gen rendering techniques requires high end video cards to get good frames, unheard of.

:music_whistling: ...lol