March 25, 20224 yr 1 hour ago, psolk said: The first with a 49 ultrawide gsync premium with an onboard processor can have my money 🙂 Honestly, it's really close to achievable but sadly KEWR is my home airport so I have to dial my settings down to the performance there with DD KEWR as my least common denominator 😞 I doubt we will ever see it. But man do we need an updated ewr and dca. 5800X3D, 4090FE, 64GB DDR4 3600C16, Gigabyte X570S MB, EVO 970 M.2's, Alienware 3821DW and 2 22" monitors, Corsair RM1000x PSU, 360MM MSI MEG, MFG Crosswind, T16000M Stick, Boeing TCA Yoke/Throttle, Skalarki MCDU and FCU, Logitech Radio Panel/Switch Panel, Spad.Next
March 25, 20224 yr 9 minutes ago, micstatic said: I doubt we will ever see it. But man do we need an updated ewr and dca. The whole NYC area could do with a refresh although the default city itself looks worlds better in V5 than V4... EWR is so neglected, especially with all the recent changes. It is amazing how even DD KEWR looks and performs much better in V5 than V4 though. Have a Wonderful Day -Paul Solk
March 25, 20224 yr I go with dd ewr, dd for the Manhattan skyline, fsdt jfk and imaginsim lga 5800X3D, 4090FE, 64GB DDR4 3600C16, Gigabyte X570S MB, EVO 970 M.2's, Alienware 3821DW and 2 22" monitors, Corsair RM1000x PSU, 360MM MSI MEG, MFG Crosswind, T16000M Stick, Boeing TCA Yoke/Throttle, Skalarki MCDU and FCU, Logitech Radio Panel/Switch Panel, Spad.Next
March 25, 20224 yr 9 hours ago, SteveW said: That's a 4 core / 8 LP in HT mode. Assuming you don't have an extreme overclock, with HT enabled it will load files faster but it's a close call between two variations to try which depend on the system. I would first try: HT Enabled 4 core 11,11,11,01=AffinityMask = 253 11,11,11,01=P3DCoreAffinityMask = 253 00,00,00,01=MainThreadScheduler = 0 00,00,01,00=RenderThreadScheduler = 1 00,00,10,00=FrameWorkerThreadScheduler = 2 [JobScheduler] AffinityMask=253 P3DCoreAffinityMask=253 MainThreadScheduler=0 RenderThreadScheduler=1 FrameWorkerThreadScheduler=2 That will load scenery faster but RenderThreadScheduler and FrameWorkerThreadScheduler will occupy the same core. They usually don't consume the entire core bandwidth together, depends on scenery mainly. If the core is hitting 100% (second core which is the 3rd and 4th logical processor combined) then try: HT Enabled 4 core 11,01,01,01=AffinityMask = 213 11,01,01,01=P3DCoreAffinityMask = 213 00,00,00,01=MainThreadScheduler = 0 00,00,01,00=RenderThreadScheduler = 1 00,01,00,00=FrameWorkerThreadScheduler = 2 [JobScheduler] AffinityMask=213 P3DCoreAffinityMask=213 MainThreadScheduler=0 RenderThreadScheduler=1 FrameWorkerThreadScheduler=2 That will be a little slower at loading files but still faster than HT disabled at the same clock rate. Thank you very much for your recommendation Steve! I`ll try it out ASAP, and let you know the outcome! Outcome: Apparently the first recommendation worked like a charm so far! Thank you @SteveW Thank you! Edited March 26, 20224 yr by rafikihd SIM: P3DV5.3 CPU: Intel Core i7-4790k @ 4.0GHz GPU: GTX1060 6GB OS: Windows 10 64Bits RAM: XPG DDR3 1600 16GB (2x 8GB) Display: Samsung LC27JG50QQLXZX 27´
March 26, 20224 yr I did a few more tests on 10900k and 12900K and used the affinity mask as suggested by Mr. Steve W. Even on a 42" monitor 60 Hz (that will put lots of strain on GPU or CPU) on PMDG and FSLABS at JFK will do 30 FPS, don't go crazy about FPS if it will dip to 28, just to be noticed it is smooth and no stutters with the following settings (you should have no issues with a smaller monitor). In my P3D config (after backup) I changed: [JobScheduler] As recommended by Steve [TERRAIN] LOD_RADIUS=7.0 UseGlobalTerrainView=True And this if you wanna try: Edited March 26, 20224 yr by LRBS 747 Captain for the last 39 years, and still learning.
March 26, 20224 yr Author thanks for all the replys. As far as add ons, I have most ORBX products, PMDG, QW787, major airports are also 3rd party... FSDT, Imaginesim, flytampa, etc. I think easily have over 100 add ons in all. Monitor is a 49" CRG9 samsung.
March 26, 20224 yr @LRBS monitor size means nothing, it is monitor resolution that matters. A 49' CRG9 is 5120x1440 same as my 49 ultrawide. HD, 4K, 5K, is what determines performance not monitor size. Also, with all due respect but setting Cloud coverage to high from Max will prevent true Overcast, you've turned down water to medium, set Aircraft AI to 47% but road, ship and leisure boats to 40% then cranked shadow settings, textures to 15cm and mesh to 2M... Just personally but flying an add-on like FSL, QW, PMDG I turn down ships, leisure boats, Cloud coverage to Max, drop shadows down a notch, don't sacrifice AA and get a spectacular result. You are sacrificing some major settings for some eye candy that to me personally is not nearly as important with a good add-on aircraft that is resource intensive.. All personal preference of course though. Also, what AI are you using? Have a Wonderful Day -Paul Solk
March 26, 20224 yr Paul, I agree with you about the resolution, thou based on my experience and what I noticed on my PCs, monitor size also plays a role in performance. I think that both have some degree of affecting the performance. Regarding the settings I listed, it's just a matter of preference or a start point to play with, everybody's PC is different and we have our own priorities. And for the a/c ai traffic, in my case I use the default, for some reason when I go above 50% I notice intermittent stutters (even I dial down other settings). I didn't try any other software, to be honest. I just tried to be helpful and share as much as I can with my limited knowledge. 747 Captain for the last 39 years, and still learning.
March 26, 20224 yr 19 hours ago, Rob_Ainscough said: Not to worry, next gen AMD CPU and GPU will have no problem achieving 60 FPS everywhere with everything installed. The LG C9 gsync range is 40-120 ... as you know stutters are possible (regardless of hardware) at various locations across the globe, hopefully LM address this. LG is releasing the C2 in May that's OLED 42" with Gsync, but don't know it's Gsync range ... it's on my to buy list for my dev PC. Cheers, Rob. I've been pushing and pushing this morning since I've been up since 1am unable to sleep anyway... With a bit of tweaking you are right, I can get pretty closed to maxed out settings and sustain 60FPS the whole time. I found myself just dropping a few things I may even be able to raise back up again like terrain textures but I also had heavy cloud cover. Thanks to @micstatic I even installed DD NYC and it looks brilliant. I never knew it was updated! The big difference was the Job Scheduler. Huge thanks to you, @SteveW and @MammyJammy for the affinity mask thread which i studied intently overnight. With my 12700 I went with the settings from that thread. I'm using a similar system to the OP and similar add-ons. Our AM will be different as I am a 12700 not a 12900 [JobScheduler]AffinityMask=1048575P3DCoreAffinityMask=1015284MainThreadScheduler=0RenderThreadScheduler=5FrameWorkerThreadScheduler=10 So I've cranked my AA to 4xSS, textures to 2048, had to drop autogen and scenery draw distance down a notch to prevent items popping in the far distance and I don't run reflections or dynamic reelections mainly due to preference. I also drop Mesh down by a notch, don't run the Use High Resolution Textures as I use Orbx Global so wasn't sure if there was any point. The other thing I am very conservative on is shadows so there are things I can play with. With the settings below I can pretty much sustain 60FPS even at KEWR with AIG and the Global Ships package. Next thing to bump up is my Texture Resolution beyond 1M but I'm very happy with almost everything right now. I would just like the entire image in the sim to be a bit crisper, it's a bit washed out at the further LOD levels whereas if I bank the airplane and look down it's crystal clear. @LRBS resolution is resolution, size shouldn't impact performance but will impact your ability to perceive things so stutters are far more noticeable on a 50 inch screen than a 30 inch screen. Unfortunately, very few people use default AI, you can see my AI levels with AIG below but using default AI definitely plays a factor in your settings. Also, I am shooting for consistent 60FPS not 30. Appreciate you trying to assist and sharing your settings! Edited March 26, 20224 yr by psolk Have a Wonderful Day -Paul Solk
March 30, 20224 yr @SteveW, appreciate a recommendation for my setup. At the moment, I have the following: [JobScheduler] AffinityMask=4095 P3DCoreAffinityMask=4095 MainThreadScheduler=0 RenderThreadScheduler=2 FrameWorkerThreadScheduler=4 I have an i7-8700K Coffee Lake (at 4.7 GHz) with 6 cores and 12 threads (HT is on). With the above configuration I noticed core 0 is always at 100% while running P3Dv5. Edited March 30, 20224 yr by David Vega dv Win 10 Pro || i7-8700K || 32GB || ASUS Z370-P MB || NVIDIA GeForce GTX 1080 Ti 11Gb || 2 960 PRO 1TB, 840 EVO My Files in the AVSIM Library
March 30, 20224 yr Commercial Member 1 hour ago, David Vega said: i7-8700K Coffee Lake (at 4.7 GHz) with 6 cores and 12 threads (HT is on). With the above configuration I noticed core 0 is always at 100% while running P3Dv5. With P3D Display Settings fps locked the sim will be making look-ahead frames so you will see the main Logical Processor Zero running at or near 100%. Using the fps Unlocked on the slider and setting VSync=On will limit frame making to the monitor refresh frequency and if that is low enough the LP will show less than 100%. The monitor refresh might be adjustable to a lower frequency if too high for the desired display settings, and the Nvidia Control Panel, Manage 3D Settings, Program Settings tab, Prepar3D.exe profile can be be set up with Vertical sync=Adaptive (half refresh rate) to reduce the vsync frequency to half.. For 6 core 12 LP I would use this JobScheduler setting: HT Enabled 6 core 05,04,03,02,01,00=core number 11,11,11,01,01,01=AffinityMask = 4053 11,11,11,01,01,01=P3DCoreAffinityMask = 4053 00,00,00,00,00,01=MainThreadScheduler = 0 00,00,00,00,01,00=RenderThreadScheduler = 1 00,00,00,01,00,00=FrameWorkerThreadScheduler = 2 [JobScheduler] AffinityMask=4053 P3DCoreAffinityMask=4053 MainThreadScheduler=0 RenderThreadScheduler=1 FrameWorkerThreadScheduler=2 That will free up the three main cores 0 1 and 2 from sharing with background threads which will now run doubled on cores 3 4 and 5. Steve Waite: Engineer at codelegend.com
March 30, 20224 yr Thank you @SteveW. I have an ROG PG279Q G-Sync monitor at 144 Hz. FPS was unlocked internally. On NCP, Monitor Technology is set to G-SYNC, Max Frame Rate: 36 fps, Vertical sync: use the 3D app... . Is this better than selecting "Fixed Refresh" in Monitor Technology and then Adaptive (half refresh rate) in Vertical sync? dv Win 10 Pro || i7-8700K || 32GB || ASUS Z370-P MB || NVIDIA GeForce GTX 1080 Ti 11Gb || 2 960 PRO 1TB, 840 EVO My Files in the AVSIM Library
March 30, 20224 yr Commercial Member 52 minutes ago, David Vega said: Is this better than selecting "Fixed Refresh" in Monitor Technology and then Adaptive (half refresh rate) in Vertical sync? It might well be. Maybe if you can set the monitor to 72Hz and set vsync=half refresh in NCP, and in P3D Display vsync=on tp=off unlocked fps should give a stable 36fps, compare that. An alternative which I think will be less stable but worth checking is to leave the monitor at 144Hz and set the fps limit to 36fps in NCP Prepar3D.exe profile. Steve Waite: Engineer at codelegend.com
March 30, 20224 yr Again, thank you @SteveW. I have it set with "... monitor at 144Hz and set the fps limit to 36fps in NCP Prepar3D.exe profile" and it's buttery smooth. Not complaining, but I still see core 0 at 100%. Maybe that's just the way it has always been. I'm OK with these settings. Thank you again. Edited March 30, 20224 yr by David Vega dv Win 10 Pro || i7-8700K || 32GB || ASUS Z370-P MB || NVIDIA GeForce GTX 1080 Ti 11Gb || 2 960 PRO 1TB, 840 EVO My Files in the AVSIM Library
March 30, 20224 yr Commercial Member 47 minutes ago, David Vega said: I still see core 0 at 100% Yes, you will if the display demand is greater than the CPU throughput then the core will reach 100%. Steve Waite: Engineer at codelegend.com
Archived
This topic is now archived and is closed to further replies.