October 30, 20214 yr Just spent a couple of hours researching MSFS settings. Part of that was the YouTube video below showing side by side comparison videos of each MSFS graphics setting. I had been running everything on Ultra, but after watching the video I reduced many setting down to medium or high as higher settings made little visual difference. The FPS saved by the changes have enabled me to run a render resolution of 130 while keeping Terrain Level of Detail at 400 and Objects Level of Detail at 200. Two settings made a big difference to texture sharpness out to the far distance. Texture Resolution HIGH instead of ULTRA (the video shows this well). Texture Super Sampling at 8x8 That combo together with a Render Scale of 130 gives amazingly sharp textures as far as the eye can see. VSYNC is on with FPS locked to 30. Developer mode shows FPS pretty steady at 40 fps, but locking at 30 seems best. My sim looks better, and is smoother. CPU Ryzen 7800X 3D RAM 32GB Corsair VENGEANCE DDR5 6000MHz GPU GEFORCE RTX 4090 Monitor AOC AGON AG352UCG UltraWide G-Sync @ 3440x1440 Internal Storage 1TB NVMe PCIe SSD External Storage Three 4Tb HDs
October 31, 20214 yr Author Some shots showing no blurry textures at all in the distance with ... Render Scale=130 (up from 100) Texture resolution=high (down from Ultra)Texture Super Sampling at 8x8 (up from 4x4) These settings eliminate distant blurry ground textures on my PC while keeping smooth frames - not possible with all settings at Ultra, but reducing most settings back to medium or high (no loss of quality as seen in the video) gives the fps headroom. Edited October 31, 20214 yr by MrBitstFlyer CPU Ryzen 7800X 3D RAM 32GB Corsair VENGEANCE DDR5 6000MHz GPU GEFORCE RTX 4090 Monitor AOC AGON AG352UCG UltraWide G-Sync @ 3440x1440 Internal Storage 1TB NVMe PCIe SSD External Storage Three 4Tb HDs
October 31, 20214 yr That video is so old... many things have changed in graphics and CPU usage since September 2020. Also, are we supposed to judge the quality of the textures and notice any difference in resolution / supersampling from a 1080p Youtube compressed video? With your system and fps limiter at 30 fps you can run everything on Ultra... 7800X3D | 2x32 GB DDR5-6000 CL32 | RTX 5080 | Alienware OLED 34" | 1 Gbps fiber
October 31, 20214 yr The old video shows how each option will affect performance. Since my system is different than the OP's my trials did not match his suggestions. I did like the OP suggested but could not see any great difference in the distance. I'm running higher options than the OP because I have a 2K monitor. A 4K movie or 1440 looks better than the same movie on a 2K television set too. So despite my higher settings the resolution makes the overall effect sub par to the videos I see on Youtube. But I have the fun of seeing the sliders all the way to the right when I escape the flight 😁 5800X3D, RTX4070, 600 Watt, one or two 1440p 32" screens, 64 GB RAM, 4 TB PCle 3 NVMe, Warthog throttle, VKB NXT EVO stick, Honeycomb Alpha yoke, CH quad, 3 Logitech panels, 2 StreamDecks, Desktop Aviator Trim Panel. Crystal Light VR.
October 31, 20214 yr Author 4 hours ago, MrFuzzy said: That video is so old... many things have changed in graphics and CPU usage since September 2020. Also, are we supposed to judge the quality of the textures and notice any difference in resolution / supersampling from a 1080p Youtube compressed video? With your system and fps limiter at 30 fps you can run everything on Ultra... My system will not run everything at ultra with a render resolution of 130. The video may not be current, or be at an ideal resolution, but it does show raising settings gives diminishing returns. Like many, I saw blurry ground textures at a certain distance from the aircraft. Raising the Render scale does help reduce these blurry textures. However, I didn’t want to reduce graphic quality. The video clearly shows very little difference in quality between medium/high/ultra for many settings. By reducing these settings, especially the ones I mentioned, a Render Scale of 130 was achieved. I realise these exact settings will not give the same results on every system, but It may help some achieve better graphics. CPU Ryzen 7800X 3D RAM 32GB Corsair VENGEANCE DDR5 6000MHz GPU GEFORCE RTX 4090 Monitor AOC AGON AG352UCG UltraWide G-Sync @ 3440x1440 Internal Storage 1TB NVMe PCIe SSD External Storage Three 4Tb HDs
October 31, 20214 yr I may have missed it but don't see where you stated your original resolution... 1080/2k/4k/5k? Where did you start before going to render resolution of 130? Apologies if I missed it. Have a Wonderful Day -Paul Solk
October 31, 20214 yr Author 5 hours ago, psolk said: I may have missed it but don't see where you stated your original resolution... 1080/2k/4k/5k? Where did you start before going to render resolution of 130? Apologies if I missed it. Agon 35inch Ultrawide G-SYNC Monitor 3440x1440. I had everything on Ultra and Render scale at 100. The video shows huge increase in quality in graphics quality between Low/Medium, but less of a jump from Medium upwards. Some settings like Ambient Occlusion have very little difference from Medium upwards. I reduced most settings downwards but noticed very little drop in quality, if any. However, the saving in FPS resulted in the option to raise Render Scale. A Render Scale of 130 AND increasing Texture Super Sampling to 8x8 gave me the results in the screenshots above. CPU Ryzen 7800X 3D RAM 32GB Corsair VENGEANCE DDR5 6000MHz GPU GEFORCE RTX 4090 Monitor AOC AGON AG352UCG UltraWide G-Sync @ 3440x1440 Internal Storage 1TB NVMe PCIe SSD External Storage Three 4Tb HDs
October 31, 20214 yr 2 minutes ago, MrBitstFlyer said: Agon 35inch Ultrawide G-SYNC Monitor 3440x1440. I had everything on Ultra and Render scale at 100. The video shows huge increase in quality in graphics quality between Low/Medium, but less of a jump from Medium upwards. Some settings like Ambient Occlusion have very little difference from Medium upwards. I reduced most settings downwards but noticed very little drop in quality, if any. However, the saving in FPS resulted in the option to raise Render Scale. A Render Scale of 130 AND increasing Texture Super Sampling to 8x8 gave me the results in the screenshots above. Ah sorry, didn't see your sig... I am going to try 8x8 first. I'm at 5120x1440 so 130 may be a bit aggressive but I was at 400/400 LOD now I'm at 250/200 and can't notice a difference. Also lowered pre-cache from ultra to low and that seemed to help tremendously for some reason. Have a Wonderful Day -Paul Solk
October 31, 20214 yr 12 hours ago, MrFuzzy said: With your system and fps limiter at 30 fps you can run everything on Ultra... Just did a test, limited the FPS to 30, from the MSFS menu and from nVidia CP, still have more than 30 FPS. I9 12900k@ 5 GHz | 64 GB DDR5 | Asus Z790 | RTX 5090 | 8K Samsung Q900C 75" | VKB MK III PRO L | Virpil Throttle MONGOOST-50 | MFG Crosswind | TrackIR5
October 31, 20214 yr I'm running RS 130, all Ultra, Vsync to 30, LOD-O 200, Texture SS 8x, and finally am able to consider only LOD-T prior to flight launch: if I'm in the 787 HD mod, flying into the worst of the worst terminals as long as I keep LOD-T at 220 or less I get 100 percent stutter-free performance from gate to gate. Landing at KSFO the other day with LOD-T at 300 killed frames down to 25-26: simply going to 220 fixed that completely. I can increase LOD-T much higher (300-400) depending on plane and airports involved and retain perfect performance. LOD-T seems to have the most predictable impact on the Main Thread, making it easy to setup a flight so that one never needs to stop and adjust, gate to gate. Noel System: 9900X3D Noctua NH-D15 G2, MSI Pro 650-P WiFi, G.SKILL 64GB (2 x 32GB) 288-Pin PC RAM DDR5 6000, WD NVMe 2Tb x 1, Sabrent NVMe 2Tb x 1, RTX 4090 FE, Corsair RM1000W PSU, Win11 Home, LG Ultra Curved Gsync Ultimate 3440x1440, Phanteks Enthoo Pro Case, TCA Boeing Edition Yoke & TQ, Cessna Trim Wheel, RTSS Framerate Limiter w/ Front Edge Sync. Aircraft used in MSFS 2024: Fenix A320, Aerosoft CRJ, FBW, WT 787X, I-Fly 737 MAX 8, Citation Longitude.
October 31, 20214 yr 31 minutes ago, Noel said: I'm running RS 130, all Ultra, Vsync to 30, LOD-O 200, Texture SS 8x, and finally am able to consider only LOD-T prior to flight launch: if I'm in the 787 HD mod, flying into the worst of the worst terminals as long as I keep LOD-T at 220 or less I get 100 percent stutter-free performance from gate to gate. Landing at KSFO the other day with LOD-T at 300 killed frames down to 25-26: simply going to 220 fixed that completely. I can increase LOD-T much higher (300-400) depending on plane and airports involved and retain perfect performance. LOD-T seems to have the most predictable impact on the Main Thread, making it easy to setup a flight so that one never needs to stop and adjust, gate to gate. Maybe Asobo should look into creating a "Dynamic LOD". If you're on the ground and up close to things you'll see them anyway. The higher LOD (if the scenery is made with proper LOD levels) will only be useful at higher altitudes where you can see further. So low altitude = lower LOD, high altitude = higher LOD. It could help with the framerates sinking so bad every time we land. But I think there's something else going on with that as well that needs to be fixed. Edit: Thinking about it some more, I don't want Asobo to lower the LOD at low altitudes, if I can see where the tree line stops in the horizon, it's not high enough. Edited October 31, 20214 yr by Republic3D AMD Ryzen R9 9950X3D | Asus Astral RTX 5080 OC | 32 GB DDR5 6000 CL30 | 3440x1440 G-Sync | Logitech Pro Throttles Rudder Yoke Panels | Thrustmaster T.16000M FCS | TrackIR 5 | Oculus Rift S
October 31, 20214 yr 3 hours ago, Republic3D said: Maybe Asobo should look into creating a "Dynamic LOD". If you're on the ground and up close to things you'll see them anyway. The higher LOD (if the scenery is made with proper LOD levels) will only be useful at higher altitudes where you can see further. So low altitude = lower LOD, high altitude = higher LOD. It could help with the framerates sinking so bad every time we land. But I think there's something else going on with that as well that needs to be fixed. Edit: Thinking about it some more, I don't want Asobo to lower the LOD at low altitudes, if I can see where the tree line stops in the horizon, it's not high enough. I brought that idea a little while ago for the same reasons. I think you could do an intelligent algorithm that looks at the elevation and distance from the observer and only cull detail when that detail is truly in line of sight and within the range of what your basic LOD setting was. So if the surrounding view is within Terrain LOD of 300, and you can't actually see detail at LOD 300 because it's flat, then reduce the Terrain LOD substantially. Whereas if you could do line of sight to some mountains or other that are within your LOD 300 setting then don't reduce LOD, or something along those line. Noel System: 9900X3D Noctua NH-D15 G2, MSI Pro 650-P WiFi, G.SKILL 64GB (2 x 32GB) 288-Pin PC RAM DDR5 6000, WD NVMe 2Tb x 1, Sabrent NVMe 2Tb x 1, RTX 4090 FE, Corsair RM1000W PSU, Win11 Home, LG Ultra Curved Gsync Ultimate 3440x1440, Phanteks Enthoo Pro Case, TCA Boeing Edition Yoke & TQ, Cessna Trim Wheel, RTSS Framerate Limiter w/ Front Edge Sync. Aircraft used in MSFS 2024: Fenix A320, Aerosoft CRJ, FBW, WT 787X, I-Fly 737 MAX 8, Citation Longitude.
November 1, 20214 yr 1 hour ago, Noel said: I brought that idea a little while ago for the same reasons. I think you could do an intelligent algorithm that looks at the elevation and distance from the observer and only cull detail when that detail is truly in line of sight and within the range of what your basic LOD setting was. So if the surrounding view is within Terrain LOD of 300, and you can't actually see detail at LOD 300 because it's flat, then reduce the Terrain LOD substantially. Whereas if you could do line of sight to some mountains or other that are within your LOD 300 setting then don't reduce LOD, or something along those line. Yep, some sort of dynamic Terrain LOD could be great. I don't need to render 10 miles in either direction when taxiing around at the airfield. They should experiment with that, as long as they make it optional. Kind of like with the Pre-Cache. AMD Ryzen R9 9950X3D | Asus Astral RTX 5080 OC | 32 GB DDR5 6000 CL30 | 3440x1440 G-Sync | Logitech Pro Throttles Rudder Yoke Panels | Thrustmaster T.16000M FCS | TrackIR 5 | Oculus Rift S
November 1, 20214 yr 9 hours ago, Sphynnx said: Just did a test, limited the FPS to 30, from the MSFS menu and from nVidia CP, still have more than 30 FPS. Maybe your monitor refresh rate is not 60 Hz? The "30 fps" setting in MSFS means 1/2 refresh rate, not actually 30 fps. Another thing they will probably never fix... 7800X3D | 2x32 GB DDR5-6000 CL32 | RTX 5080 | Alienware OLED 34" | 1 Gbps fiber
November 1, 20214 yr 38 minutes ago, MrFuzzy said: Maybe your monitor refresh rate is not 60 Hz? The "30 fps" setting in MSFS means 1/2 refresh rate, not actually 30 fps. Another thing they will probably never fix... It's vsync--what is there to fix? Noel System: 9900X3D Noctua NH-D15 G2, MSI Pro 650-P WiFi, G.SKILL 64GB (2 x 32GB) 288-Pin PC RAM DDR5 6000, WD NVMe 2Tb x 1, Sabrent NVMe 2Tb x 1, RTX 4090 FE, Corsair RM1000W PSU, Win11 Home, LG Ultra Curved Gsync Ultimate 3440x1440, Phanteks Enthoo Pro Case, TCA Boeing Edition Yoke & TQ, Cessna Trim Wheel, RTSS Framerate Limiter w/ Front Edge Sync. Aircraft used in MSFS 2024: Fenix A320, Aerosoft CRJ, FBW, WT 787X, I-Fly 737 MAX 8, Citation Longitude.
Archived
This topic is now archived and is closed to further replies.