Jump to content
Sign in to follow this  
brucewtb

Marked FPS decrease after 5.3 update

Recommended Posts

23 hours ago, brucewtb said:

currently have HT on but sceptical about any benefit from this.  As for EA with my system there is certainly at least a 5fps penalty for having it off if 2D clouds are used but no difference if the clear weather preset is used (ie no clouds).  I am curious as to whether others see this penalty for using 2D clouds instead of volumetric clouds as it is quite counter intuitive.. I use the Job Scheduler settings suggested by Steve earlie

I can clearly confirm what you are stating here. This is also the case at my end. 

 

2d clouds are more fps heavy then volumetric Clouds (set to medium or even high). 

Clear weather in ea mode has also a fps hit. And I'm not sure why! My system runs a lot smoother with ea on and vol clouds on  even at 1\8 coverage. So visually spoken almost no difference to clear skies. 

I, as a LM beta tester, have reported this and would like to ask you all to report this too. I assume a shader or VRAM issue. 

Marcus

Edited by mpo910

Regards,

Marcus P.

xaP1VAU.png

Share this post


Link to post
6 hours ago, brucewtb said:

Well in these tests I have been relying on what the FPS counter was telling me which I would have thought was quit objective rather than a subjective what looks best (but I agree at the end of the day "what looks best" is the test that matters). The frames in the mid to high 30s refers to the test situation which I run with vsync off - my usual practice is to have vsync on so I get fps of 30 (hopefully) which is the refresh my Sony TV is set to.

The Shift+Z counter is not that easy to assess. It changes so rapidly.

But turning off VSync pushes up demand on the GPU and if it hits 100% that’s when the stutters start. Better to leave it on for a smoother experience.

With engines running at YSSY just go into slew mode and go up to 500ft. Then rotate through 360° using the 1 key on the keypad. Don’t watch the frame rate. Watch the outside view for smoothness. The human eye is very sensitive to any variations. If it’s without any stutters that’s perfect. If you get stutters reduce settings until they stop.


Ray (Cheshire, England).
System: P3D v5.3HF2, Intel i9-13900K, MSI 4090 GAMING X TRIO 24G, Crucial T700 4Tb M.2 SSD, Asus ROG Maximus Z790 Hero, 32Gb Corsair Vengeance DDR5 6000Mhz RAM, Win 11 Pro 64-bit, BenQ PD3200U 32” UHD monitor, Fulcrum One yoke.
Cheadle Hulme Weather

Share this post


Link to post

Marcus for me EA and VOL clouds on is the opposite of you my GPU goes to almost  max  so until I can resolve that I am using ASCA with AS

  • Like 1

MSI X670E Carbon Wifi, 7800X3D, TUF 4090 , X4 NVME's. OS on a 2TB 980 Pro , MSFS on " 2TB WD Black , G Skill F5 32 GB RAM ( 6000 allegedly) Corsair RM1000 PSU, Artic Freezer 280 AIO Cooler . Phanteks P600s Case ,TCL QM8B 50" 120/144 Hz TV, Warthog Stick and TCA Captains throttle ( full pack) Logitech / Saitek Rudder Pedals , Streamdeck XL / Streamdeck +

Share this post


Link to post
4 hours ago, Ray Proudfoot said:

The Shift+Z counter is not that easy to assess. It changes so rapidly.

But turning off VSync pushes up demand on the GPU and if it hits 100% that’s when the stutters start. Better to leave it on for a smoother experience.

With engines running at YSSY just go into slew mode and go up to 500ft. Then rotate through 360° using the 1 key on the keypad. Don’t watch the frame rate. Watch the outside view for smoothness. The human eye is very sensitive to any variations. If it’s without any stutters that’s perfect. If you get stutters reduce settings until they stop.

Ray maybe just maybe our once mighty 1080ti's are not cutting it anymore ?


MSI X670E Carbon Wifi, 7800X3D, TUF 4090 , X4 NVME's. OS on a 2TB 980 Pro , MSFS on " 2TB WD Black , G Skill F5 32 GB RAM ( 6000 allegedly) Corsair RM1000 PSU, Artic Freezer 280 AIO Cooler . Phanteks P600s Case ,TCL QM8B 50" 120/144 Hz TV, Warthog Stick and TCA Captains throttle ( full pack) Logitech / Saitek Rudder Pedals , Streamdeck XL / Streamdeck +

Share this post


Link to post
1 minute ago, sonny147 said:

Ray maybe just maybe our once mighty 1080ti's are not cutting it anymore ?

Whilst clearly not as powerful as the 2 or 3 series it’s still a very capable card. Just set your sliders appropriately.

I don’t recall Bruce mentioning Ai traffic but I can get smooth results at YSSY and other places with plenty of Ai.

I cannot overemphasis how important it is to disable 3D people in FlyTampa products. They have a significant hit on performance.

  • Like 1

Ray (Cheshire, England).
System: P3D v5.3HF2, Intel i9-13900K, MSI 4090 GAMING X TRIO 24G, Crucial T700 4Tb M.2 SSD, Asus ROG Maximus Z790 Hero, 32Gb Corsair Vengeance DDR5 6000Mhz RAM, Win 11 Pro 64-bit, BenQ PD3200U 32” UHD monitor, Fulcrum One yoke.
Cheadle Hulme Weather

Share this post


Link to post
1 hour ago, Ray Proudfoot said:

Whilst clearly not as powerful as the 2 or 3 series it’s still a very capable card. Just set your sliders appropriately.

I don’t recall Bruce mentioning Ai traffic but I can get smooth results at YSSY and other places with plenty of Ai.

I cannot overemphasis how important it is to disable 3D people in FlyTampa products. They have a significant hit on performance.

roger ! 


MSI X670E Carbon Wifi, 7800X3D, TUF 4090 , X4 NVME's. OS on a 2TB 980 Pro , MSFS on " 2TB WD Black , G Skill F5 32 GB RAM ( 6000 allegedly) Corsair RM1000 PSU, Artic Freezer 280 AIO Cooler . Phanteks P600s Case ,TCL QM8B 50" 120/144 Hz TV, Warthog Stick and TCA Captains throttle ( full pack) Logitech / Saitek Rudder Pedals , Streamdeck XL / Streamdeck +

Share this post


Link to post
On 2/10/2022 at 3:56 PM, mpo910 said:

I can clearly confirm what you are stating here. This is also the case at my end. 

 

2d clouds are more fps heavy then volumetric Clouds (set to medium or even high). 

Clear weather in ea mode has also a fps hit. And I'm not sure why! My system runs a lot smoother with ea on and vol clouds on  even at 1\8 coverage. So visually spoken almost no difference to clear skies. 

I, as a LM beta tester, have reported this and would like to ask you all to report this too. I assume a shader or VRAM issue. 

Marcus

Thanks Marcus for confirming this observation.  It is probably why I first noticed an apparent loss of FPS with the latest updates as the saved test scenarios used 2D clouds with no EA.  Something has changed with the way the sim gets displayed with EA off.  While I have to allow for the way I have tweaked the black settings on my TV to get better cloud contrast and darker nights in MS2020 with EA off it now looks pretty ordinary mainly because there is no haze - is there a setting somewhere I have missed?  And of course with EA on you get that excessive haze - if they could reduce the haze and tweak the clouds so they are at least up to (albeit imperfect) MSFS 2020 standards then the look would be quite good.  As it is while the 2D clouds look much better than the current volumetric offering (IMHO) that is no longer an option due to the reduced FPS and the way the sim gets displayed without EA.

Bruce


Bruce Bartlett

 

Frodo: "I wish none of this had happened." Gandalf: "So do all who live to see such times, but that is not for them to decide. All we have to decide is what to do with the time that is given to us."

Share this post


Link to post

Weird.  I only have a 1080 and volumetric clouds are fine for me (I don't use them though).  I use EA with 2D clouds.

Also - after much experimenting with several different affinity masks - I found the smoothest option for me and my i7 6700K:

Hyperthreading off, default AM.

No matter what I did with an AM - and a couple were good, hyperthreading off works best for my system in terms of keeping frames up, and keeping it smooth.

Share this post


Link to post

I feel you need to be an MS2020 user to assess the quality of P3D5 volumetric clouds.  There are issues with MS2020 clouds for sure but they are much better than those we get in P3D5 and just shows what can be achieved in this 3D clouds space.  Still not sure about hyperthreading -  with P3D5 all I seem to get is increased temps but in MSFS2020 I don't get excessive temps with it on but then CPU use is much less like only about 30% with the GPU doing most of the heavy lifting at 90%.  It also seems to me that in P3D there is a lot of computing power being left on the table with the CPU only running at 50% overall (although usually 100% on at least one core) and only 30% GPU usage.  I guess it is a bit similar with MSFS2020 with only 30% CPU usage but at least the GPU gets a workout.

Bruce


Bruce Bartlett

 

Frodo: "I wish none of this had happened." Gandalf: "So do all who live to see such times, but that is not for them to decide. All we have to decide is what to do with the time that is given to us."

Share this post


Link to post

I performed a test yesterday of the impact of EA=On (which also forces HDR=On) at Schiphol slewed up to a couple of thousand feed in a C&D PMDG737-800.

I usually run ASP3D / ASCA but for this test I wanted to control cloud layers manually. I added one low-level Cumulus layer 3/8; one medium Stratus layer 6/8 and one Cirrus layer 2/8.

I rotated the aircraft through 360 degrees so all visible cloud was brought into view. VSync was Off for the test but normally it's on. I use a very old but useful utility that draws a continuous line of fps so my readings are taken from that and not the Shift+Z reading which I find impossible to use accurately.

  • With EA=Off and HDR=Off it varied from 38-47.
  • With EA=On; HDR=On and VC Off it varied from 27-35
  • With EA=On; HDR=On and VC at Low it varied from 26-32
  • With EA=On; HDR=On and VC at Medium it varied from 25-32
  • With EA=On; HDR=On and VC at High it varied from 22-26
  • With EA=On; HDR=On and VC at Ultra it varied from 15-18.

For me the best results are obtained with EA=Off and HDR=Off.

 

  • Like 1

Ray (Cheshire, England).
System: P3D v5.3HF2, Intel i9-13900K, MSI 4090 GAMING X TRIO 24G, Crucial T700 4Tb M.2 SSD, Asus ROG Maximus Z790 Hero, 32Gb Corsair Vengeance DDR5 6000Mhz RAM, Win 11 Pro 64-bit, BenQ PD3200U 32” UHD monitor, Fulcrum One yoke.
Cheadle Hulme Weather

Share this post


Link to post

Ray -- it would be interesting to use EA=On and HDR=Off. You can turn off HDR in P3D.cfg by changing HDR=1 TO HDR=0.

Edited by pgde

Gigabyte x670 Aorus Elite AX MB; AMD 7800X3D CPU; Deepcool LT520 AIO Cooler; 64 Gb G.Skill Trident Z5 NEO DDR5 6000; Win11 Pro; P3D V5.4; 1 Samsung 990 2Tb NVMe SSD: 1 Crucial 4Tb MX500 SATA SSD; 1 Samsung 860 1Tb SSD; Gigabyte Aorus Extreme 1080ti 11Gb VRAM; Toshiba 43" LED TV @ 4k; Honeycomb Bravo.

 

Share this post


Link to post
28 minutes ago, pgde said:

Ray -- it would be interesting to use EA=On and HDR=Off. You can turn off HDR in P3D.cfg by changing HDR=1 TO HDR=0.

Okay, one for tomorrow. But it’s unlikely I’ll stay with it as I have tried EA a couple of times and can’t live with it.

But having just agreed to that it’s not possible. Every time you make a change in the settings if EA is ON and HDR is OFF you can’t confirm the change.

It would have to be done externally via the cfg and I’m not sure I have the enthusiasm to be honest.


Ray (Cheshire, England).
System: P3D v5.3HF2, Intel i9-13900K, MSI 4090 GAMING X TRIO 24G, Crucial T700 4Tb M.2 SSD, Asus ROG Maximus Z790 Hero, 32Gb Corsair Vengeance DDR5 6000Mhz RAM, Win 11 Pro 64-bit, BenQ PD3200U 32” UHD monitor, Fulcrum One yoke.
Cheadle Hulme Weather

Share this post


Link to post
On 2/13/2022 at 10:30 PM, Ray Proudfoot said:

I performed a test yesterday of the impact of EA=On (which also forces HDR=On) at Schiphol slewed up to a couple of thousand feed in a C&D PMDG737-800.

I usually run ASP3D / ASCA but for this test I wanted to control cloud layers manually. I added one low-level Cumulus layer 3/8; one medium Stratus layer 6/8 and one Cirrus layer 2/8.

I rotated the aircraft through 360 degrees so all visible cloud was brought into view. VSync was Off for the test but normally it's on. I use a very old but useful utility that draws a continuous line of fps so my readings are taken from that and not the Shift+Z reading which I find impossible to use accurately.

  • With EA=Off and HDR=Off it varied from 38-47.
  • With EA=On; HDR=On and VC Off it varied from 27-35
  • With EA=On; HDR=On and VC at Low it varied from 26-32
  • With EA=On; HDR=On and VC at Medium it varied from 25-32
  • With EA=On; HDR=On and VC at High it varied from 22-26
  • With EA=On; HDR=On and VC at Ultra it varied from 15-18.

For me the best results are obtained with EA=Off and HDR=Off.

 

Interesting I will have to give this a try but I wonder if the results could different with a 3000 series GPU as I have?

Bruce


Bruce Bartlett

 

Frodo: "I wish none of this had happened." Gandalf: "So do all who live to see such times, but that is not for them to decide. All we have to decide is what to do with the time that is given to us."

Share this post


Link to post
7 hours ago, brucewtb said:

Interesting I will have to give this a try but I wonder if the results could different with a 3000 series GPU as I have?

Bruce

Higher fps I would imagine. Remember to set clouds manually. Don’t use ASP3D/ASCA for this specific test.


Ray (Cheshire, England).
System: P3D v5.3HF2, Intel i9-13900K, MSI 4090 GAMING X TRIO 24G, Crucial T700 4Tb M.2 SSD, Asus ROG Maximus Z790 Hero, 32Gb Corsair Vengeance DDR5 6000Mhz RAM, Win 11 Pro 64-bit, BenQ PD3200U 32” UHD monitor, Fulcrum One yoke.
Cheadle Hulme Weather

Share this post


Link to post

I am more interested in relative rather than absolute difference as in absolute terms you would expect higher fps from a 3000 series card than a 1000 series card - ie would you get better performance with VC than 2D clouds as some have found or, like you, better with VC?

Bruce


Bruce Bartlett

 

Frodo: "I wish none of this had happened." Gandalf: "So do all who live to see such times, but that is not for them to decide. All we have to decide is what to do with the time that is given to us."

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  
  • Tom Allensworth,
    Founder of AVSIM Online


  • Flight Simulation's Premier Resource!

    AVSIM is a free service to the flight simulation community. AVSIM is staffed completely by volunteers and all funds donated to AVSIM go directly back to supporting the community. Your donation here helps to pay our bandwidth costs, emergency funding, and other general costs that crop up from time to time. Thank you for your support!

    Click here for more information and to see all donations year to date.
×
×
  • Create New...