Taking advantage of an offer to improve Setting Results

[Fakum]

Thanks for that VID Hiob,,, that was very informative, and thanks for the tip on the Google Doc

[Ozmandi]

@Fakum Stop chasing the FPS* holy grail  

*(I'm including the stuttering with FPS)

I read many post about FPS\Perfomance\VideoQuality\GraphicsCardEnhancments\WindowsYESNOsettings\etc., on this site as well as ALL of the accompanying youTube videos that come with those post.  I made so many changes, based on the last 'best' thread or video, that my FPS and video quality was degrading and I couldn't remember what the game looked like before those changes.

So I started testing my own system to see if I could notice what each setting did or did not do.  I have probably spent over 30 hours benchmarking my system and I'm only half way through the total combination of settings between NCP and DCS. But I have evidence of how each setting affects my FPS( I have screenshots and videos of each test).

I'm only saying this to warn you that chasing the FPS* holy grail will only lead to......well you get the point.

 

 

I have a 3080 and get great FPS and awesome video quality but I also have the stuttering you described.  As of right now, I'm okay with it because I don't spend too much time just looking sideways.

 

My advice is, listen to everybody and test every theory - in triplicate!  And take notes! 

 

 

p.s. on FPS counters:

Do you have GeForce Experience installed? It has an FPS counter that can be turned on.  I tested it against DCS and MSI afterburner and they all displayed the same FPS. I did this because it was hard for me to see the FPS text in DCS.  (by the way - the beta version of GeForce Experience can show more than FPS if wanted)

 

 

Edited February 26, 2021 by Ozmandi
mistakes

[Canada_Moose]

19 hours ago, Hiob said:

urns out, that the statement from someone earlier, that they don't go together (or it doesn't make sense) was also false.

 

Its ED themselves that say that. Just do a search for yourself. Of course there is nothing stopping you from setting them how you wish. Personally with a Titan Xp, 4xMSAA on its own in DCS with even a light mission and assets is not very pleasing from a performance perspective. If you add 1.5x SSAA into that as well, then good luck. ED recommends one or the other. More than likely for perf reasons alone.

 

 

19 hours ago, Hiob said:

On the matter of FPS-perception:

 

https://www.testufo.com/

 

 

 

You don't understand your own sources...

 

No, he doesn't

Edited February 26, 2021 by Canada_Moose

[Glide]

23 hours ago, Fakum said:

I have to drop back and punt again

Try Vsync = FAST now in the NCP.  This should push your FPS as high as they can go with Vsync on.

[Fakum]

Ozmandi, As grateful as I would be to improve my FPS, I'm actually more interested in smoothing things out, decreasing the stuttering. Its not as horrible now with my ongoing tinkering. Flying Choppers, I find myself more often looing out the sides then forward. Tinkering with achieving smoother performance, Im hoping the biproduct of that may also improve DCD FPS. Regarding GeForce Experience, no, I dont have it installed. I am using DCS for that. I was struggling with seeing those numbers as well but yesterday I discovered you can increase the size of that in DCS Graphics options, So that was a BIG help! Thanks for your advise.

 

Glide, I am tinkering with Vsync/G Sync,, but so far only from running an instant action for a minute on the ramp, I have to test inflight still, I havent seen any differences as of yet, but expect to when fly like the Hornet fast over land mass

[Fakum]

Heres another question not yet discussed, where to test? For just testing some framerates, I would jump into 2 specific instant action scenarios so I could repeat tests accurately. So what about testing just to monitor for any form of stuttering? Since I am currently focusing on making changes to minimize any form of stuttering, should I also be doing this testing in Instant Action? Initially, it would seem that makes sense, but could there be any reasons why I would want to make sure I am in a multiplayer server instead? This question occurred to me as I have been flying in the server that I am typically in, and I was getting occasional split second freezing, and a quick sputter. Not consistent, and I couldn't help but wonder if the cause was from my PC or the actual server?

[Glide]

5 hours ago, Fakum said:

I havent seen any differences as of yet,

So, you are back to getting 120 fps?

 

Stuttering happens when Windows swaps Dedicated GPU memory with Shared GPU memory.  This happens when your card runs out of dedicated VRAM.  You can watch them happen on the Task Manager performance window for your GPU.  You will see copy activity when you experience a stutter. 

 

As for where to test.  Always test the same mission or set of missions.  They can be instant action missions depending on how large.  Syria map missions can easily use over 11GB of VRAM from the start.  I also have certain textures and views that I always check when I make a change.  For example, the seam on the nose of the TF-51D when you add Anisotropic filtering stands out much sharper.

 

Those split second stutters are scattered around the map.  There's one right as you pass the fork in the river at Poti on Caucasus. 

 

Once upon a time, there was another flight sim with a famous stutter.  I was working with the "devs" on the AI changes, and since we had the advanced AI implemented, I developed a technique to replay a mission exactly the same way each time.  That's when I discovered the stutters happen in the exact same place, but in a flight sim this was very hard to pin down.  The devs could now pinpoint a stutter and begin diagnostics.  No more stuttering.  Patch #3 , if I recall. 

[Fakum]

No on the 120 FPS,, as I stated above, that was an error on my part only using the monitors FPS feature, now I am only using DCS FPS meter. I am getting 70 - 80 FPS running on Multiplayer I frequent. I found a specific instant action on Syria map in the Hornet (Airfield Defense) that gives me 32 FPS on the ramp, and KA-50 on Syria Map gives me 60 ish on the ramp, so those 2 are the ones I have been using for testing.

 

I will run task manager to check my ram usage tonight, thanks.

[Fakum]

I see your point

[Glide]

2 hours ago, Fakum said:

I see your point

I'm afraid to run Syria map with High Textures and High Terrain Textures is too much for our 11GB VRAM cards.  Drop these down and you can just fit Syria into the dedicated gpu memory and have no stutters.  I have put Syria on the shelf until I can upgrade my card.

 

My card gets 70-80 fps as well.  Right now I am running NCP Vsync = FAST, and I'm trying SSAA 1.5 and MSAA x2.  My card will stay very close to 60 fps on Caucasus and PG with these settings.  For me, that's perfect. 

[Ozmandi]

8 hours ago, Fakum said:

. Flying Choppers, I find myself more often looing out the sides then forward.

Ahh, now I get it. 

 

 

@FakumCan you describe your micro-stutters?

Like I said I see micro-stutters on my computer but I don't have a G-SYNC monitor, so maybe I'm not experiencing what you are. 

 

 

[Lange_666]

G-Sync is not a solution for stutters, it's a solution for tearing which happens when frames are not in sync (hence the name) with the refreshrate.

[Glide]

On 2/25/2021 at 10:22 AM, Hiob said:

I highly recommend to everyone to try out for yourself.

Thank you for making me think. BTW, we didn't say they don't go together, we said they eat up FPS like candy when then are together. 

I had an interesting experience with your settings.  I had been getting 60ish FPS with FAST, and I flew your settings without the FPS counter on.  DCS never looked better.  It was smooth as silk.

Then I turned on the FPS counter and saw 40ish FPS.  My cognitive bias kicked in, and I switched back to my previous settings.  After all, 60 is better, right?  But, I've said 30 should be enough, so I decided to eat my own cooking.  I would fly your settings without the FPS counter to see if I could notice anything.  I tried vsync  ON first, and that synced at a solid 30fps.  However, there were moments when it felt like the world was slowing down.  Then I tried FAST again, and it was perfection.  I'm leaving the FPS counter off for a week with your settings to see if my card can sustain this through a couple dozen dogfights on PG and Caucasus.   Thanks for testing.

 

https://en.wikipedia.org/wiki/Cognitive_bias

[silverdevil]

so i am back. i did a bunch of testing and could not get stuttering to fully disappear until i read this little tidbit posted in another topic by @Ozmandi. the post was noted for the ALT-ENTER fullscreen fix. i read into the blog a little more and found another NCP setting. i believe it made a big difference.

 

How To Reduce Lag - A Guide To Better System Latency

 

particularly this. seeing high fps and little pauses if any 

 

[Fakum]

LOL,,, this all has been very welcome advise, and i have tried to digest it all, but will admit, Im ready to check myself in somewhere LOL. Kidding aside, I have read thru this entire thread daily, tried to make notes on the keypoints and have experimented. Unfortunately, I dont know which way is down anymore. I am NOT really focusing so much on FPS, Its the micro stuttering. Right now, I seem to only be getting the quick hiccup evry couple minutes or so,, the side views seem fairly smooth. I belive glides response to those stutters make sense with me maxing out my Vram. My confusion at this point is all the chatter about Vsync & Gsync, Full Screen. Should I use any of them, and in what combinations? I mean, you can turn Vsync on in NCP, leave it off in DCS and vice versa, you can turn Vsync to fast in NCP, do you run Gsync as well, do you run full screen with or without and on and on and on,,, then it comes to the actual testing. what is the fastest method? Testing FPS is fast, but trying a new setting to check for stuttering is lengthy,, I mean you get in the air and look around for awhile and ask your self if you see a noticeable difference, or anything at all? This has been the most challenging and time consuming, and if you really dont know what you really should be changing to what, its makes it more difficult. Again, Im focusing on microstutters and settings impacting that,, vsync, gsync, fullscreen and possible capping of fps to 60?

[silverdevil]

6 minutes ago, Fakum said:

LOL,,, this all has been very welcome advise, and i have tried to digest it all, but will admit, Im ready to check myself in somewhere LOL. Kidding aside, I have read thru this entire thread daily, tried to make notes on the keypoints and have experimented. Unfortunately, I dont know which way is down anymore. I am NOT really focusing so much on FPS, Its the micro stuttering. Right now, I seem to only be getting the quick hiccup evry couple minutes or so,, the side views seem fairly smooth. I belive glides response to those stutters make sense with me maxing out my Vram. My confusion at this point is all the chatter about Vsync & Gsync, Full Screen. Should I use any of them, and in what combinations? I mean, you can turn Vsync on in NCP, leave it off in DCS and vice versa, you can turn Vsync to fast in NCP, do you run Gsync as well, do you run full screen with or without and on and on and on,,, then it comes to the actual testing. what is the fastest method? Testing FPS is fast, but trying a new setting to check for stuttering is lengthy,, I mean you get in the air and look around for awhile and ask your self if you see a noticeable difference, or anything at all? This has been the most challenging and time consuming, and if you really dont know what you really should be changing to what, its makes it more difficult. Again, Im focusing on microstutters and settings impacting that,, vsync, gsync, fullscreen and possible capping of fps to 60?

 

you have a better card than i do so my settings should give you at least the performance i have. i have vsync off in NCP and in game.

 

note your current settings and try mine. i was too interested in microstutters.

[Fakum]

Thanks for both tips this silverdevil

6 hours ago, Lange_666 said:

G-Sync is not a solution for stutters, it's a solution for tearing which happens when frames are not in sync (hence the name) with the refreshrate.

Thanks for elaborating on that, I have not experienced any tearing with any settings, so would i benefit more with it on or off I wonder,,

[silverdevil]

 

you can do full screen only or windowed and full screen mode or off

 

so if you are not in full screen mode (hence windowed), and this setting is set to full screen mode, then there is no effect. it makes sense to do windowed / full screen then it would always be enabled.

[Fakum]

ok, so if my monitor is set to run at 120, and im getting 70 fps average in DCS, it would then make sense to have this on. and I do understand what you mean about :

 

so if you are not in full screen mode (hence windowed), and this setting is set to full screen mode, then there is no effect. it makes sense to do windowed / full screen then it would always be enabled.

 

thanks

 

[Lange_666]

59 minutes ago, Fakum said:

Thanks for elaborating on that, I have not experienced any tearing with any settings, so would i benefit more with it on or off I wonder,,

 

Really? There is always tearing when you don't have a G-Sync/Freesync monitor and VSync is turned off and your frames are not capped.

And if at that point you add TrackIR into the equation... oh boy... horrible.

Edited February 27, 2021 by Lange_666

[Fakum]

That is interesting indeed. I can only say that the only thing I have seen is some micro stuttering and FPS changes with the changes i have been making. Andi have fisddled with VSYNC on and off along with Gsync etc

 

I have confirmed that setting the terrain and textures down a notch did in fact clear up some head room for the VRAM, and although i was not pleased with the visual impact (Im spoiled), it did in fact cut down on the stuttering.

 

Also, adding SSAA 1.5 to the equation cut my Frame rates in half

[Glide]

I think we know that DCS can't sustain 120 fps with any kind of quality on today's hardware.  @Fakum the only thing left to try is a different monitor. 

[Fakum]

LOL,, Im not trying to sustain 120. If I can keep steady at 60+, Im good,, and so far, I have achieved that. Im just trying to smooth out,, I was doing good with that it seemed until i went into multiplayer and see a bit to much stuttering,, and I dropped the terrain and textures again, expecting the VRAM drop would smooth it out, but it didnt LOL,,,,Im pretty sure my monitor is not a factor here.https://www.lg.com/us/monitors/lg-38GL950G-B-gaming-monitor

 

I might giv it all another test with FPS locked at 60 in NCP, if i do that should I be setting my monitor down from 120 to 60? Wouldnt that cut down a work load on the GPU?

[Glide]

11 minutes ago, Fakum said:

LOL,, Im not trying to sustain 120. If I can keep steady at 60+, Im good,, and so far, I have achieved that. Im just trying to smooth out,, I was doing good with that it seemed until i went into multiplayer and see a bit to much stuttering,, and I dropped the terrain and textures again, expecting the VRAM drop would smooth it out, but it didnt LOL,,,,Im pretty sure my monitor is not a factor here.https://www.lg.com/us/monitors/lg-38GL950G-B-gaming-monitor

 

I might giv it all another test with FPS locked at 60 in NCP, if i do that should I be setting my monitor down from 120 to 60? Wouldnt that cut down a work load on the GPU?

Correct.  Max Frame Rate = 60 in the NCP, but be sure to turn OFF vsync and gsync for that test.  One other long shot, turn off Shader Cache in NCP.  This will eliminate any disk issues.  However, your load times will increase the first time you fly a map.

 

Pro tip: if you want a quick way to tell if a setting has cleaned up the stuttering, use Ctrl+Z to speed up time.  This makes the stutters much easier to see.  Try this on Caucasus with vsync/gsync OFF, MSAA OFF, SSAA OFF, and Shader Cache OFF.   I think that is the smoothest I've ever got it.

Edited February 27, 2021 by glide

[Fakum]

Ok, thanks