July 24, 201312 yr Odd. i dont recall any suggestions higher than 6.5 at that was at your own risk of OOM's.
July 24, 201312 yr I run FSX at 1280x1024x32bit with a LOD Radius of 4.5, and I still get performance issues in very highly detailed areas with my i5 2500k @ 4.3Ghz/1GB GeForce GTX 560Ti/8GB DDR3-1600 RAM powered system. How does anyone keep FSX stable at very high detail settings with a LOD Radius of 8.5 and a huge widescreen monitor? Christopher Low AMD Ryzen 7 9800X3D CPU / 64GB DDR5-6000 RAM / 12GB Nvidia RTX 4070 Super GPU / Gigabyte X870E Aorus Elite Wifi 7 / 1+2TB Samsung Evo Plus M2 Nvme UK2000 Beta Tester
July 24, 201312 yr IIRC if you put a parameter in that is beyond range, FSX will revert to default, so by him putting in 32, it would not display the highest available LOD, it would display the default LOD, which is 4.5. The video is at dusk, and is a little blurry from the resolution, so you can't see that far out anyway. other oddities in his config:time fraction at .66 (most peeps try for as close to .1 as they can) and bandwidth multi at 400 which is an fs9 setting. 400 is overkill, but with the fiber frame time fraction that high, it would be able to keep up I guess oh, and texture max load=1024 is good. you can go 4096 but you're asking for OOM's
July 24, 201312 yr Unfortunately anything higher as 4.5 is asking for problems in regards too OOM and less FPS in combination with other higher slider settings ;-) André
July 24, 201312 yr I see he's got MipBias=6. Completely outdated tweak. Also, he seems to be under the impression that LOD radius affects autogen radius. It doesn't. He's overdoing that system pretty hard because you can look on his final approach and see that the textures are blurry and the autogen isn't even loading (usually happens when LOD is too high and FSX can't keep up with loading textures, FPS can remain stable but they simply not load). Only time I use 6.5 is outside Orbx regions. Their textures are crisp enough that at 4.5 it just looks fine to me as things blur outside that radius. With FTX Global I may not have to ever touch LOD again and avoid any issues that come with it if its ground textures look the same way. IIRC if you put a parameter in that is beyond range, FSX will revert to default, so by him putting in 32, it would not display the highest available LOD, it would display the default LOD, which is 4.5. The video is at dusk, and is a little blurry from the resolution, so you can't see that far out anyway. other oddities in his config:time fraction at .66 (most peeps try for as close to .1 as they can) and bandwidth multi at 400 which is an fs9 setting. 400 is overkill, but with the fiber frame time fraction that high, it would be able to keep up I guess oh, and texture max load=1024 is good. you can go 4096 but you're asking for OOM's His .cfg is a mess. I run FSX at 1280x1024x32bit with a LOD Radius of 4.5, and I still get performance issues in very highly detailed areas with my i5 2500k @ 4.3Ghz/1GB GeForce GTX 560Ti/8GB DDR3-1600 RAM powered system. How does anyone keep FSX stable at very high detail settings with a LOD Radius of 8.5 and a huge widescreen monitor? I've personally found that raising the LOD doesn't affect performance much. What it does to me (only in heavy city areas) is create blurries and autogen may stop loading. This is exactly what you see in his video.
July 24, 201312 yr Commercial Member With the desktop heap tweak I can run with LOD = 6.5 and textures at 2048 and do a flight into YVR in the NGX with no OOMs. If I were to bump the textures to 4096 I will get an OOM. LOD = 6.5 is only a killer if fail to make compromises else where. For me, I can't tell the difference between 4096 and 2048 textures so I keep the lower setting there and can achieve the higher setting else where, in this case LOD. As far as LOD=32 this sounds a bit outrageous. I'd be curious to see a side-by-side comparison of 4.5, 6.5, 9, and 32. I have a feeling 32 and 9 will be the same..... - Jordan Jafferjee - AMD Ryzen 9 7950X3D | Asus X670-E Pro Prime | Gigabyte RTX4080 Eagle | 64G G.Skill Trident Z.5 DDR5-6000 | Arctic Liquid Freezer II 360 | 2x2TB Samsung 990 Pro NVME | NZXT H7 | Win 11 24H2 | TM Warthog Flight Stick + Throttle | Honeycomb Alpha + Bravo | MFG Crosswind Rudder Pedals | Samsung 43" Odyssey Neo G7 | Dell U3415W
July 24, 201312 yr I run FSX at 1280x1024x32bit with a LOD Radius of 4.5, and I still get performance issues in very highly detailed areas with my i5 2500k @ 4.3Ghz/1GB GeForce GTX 560Ti/8GB DDR3-1600 RAM powered system. How does anyone keep FSX stable at very high detail settings with a LOD Radius of 8.5 and a huge widescreen monitor? By using moderate settings, NO BP=0, keep water Low2.0, basically use common sense, and not overdo anything.
July 24, 201312 yr I'm sorry but I see nothing exciting or out of the ordinary in that video, except unusual panning technique and 5.2ghz OC on nitrogen, supposedly. Intel i7 10700K | Asus Maximus XII Hero | Asus TUF RTX 3090 | 32GB HyperX Fury 3200 DDR4 | 1TB Samsung M.2 (W11) | 2TB Samsung M.2 (MSFS2020) | Arctic Liquid Freezer II 280mm AIO | 43" Samsung Q90B | 27" Asus Monitor
July 24, 201312 yr Whatever his settings, the image quality in his video is mediocre IMHO. [email protected] - ROG Strix Z790-E - 2X16Gb G.Skill Trident DDR5 6400 CL32 - MSI RTX 4090 Suprim X - WD SN850X 2 TB M.2 - XPG S70 Blade 2 TB M.2 - MSI A1000G PCIE5 1000 W 80+ Gold PSU - Liam Li 011 Dynamic Razer case - 58" Panasonic TC-58AX800U 4K - Pico 4 VR HMD - WinWing HOTAS Orion2 MAX - ProFlight Pedals - TrackIR 5 - W11 Pro (Passmark:12574, CPU:63110-Single:4785, GPU:50688)
July 24, 201312 yr No AI at all will make a significant difference to overall performance.... Christopher Low AMD Ryzen 7 9800X3D CPU / 64GB DDR5-6000 RAM / 12GB Nvidia RTX 4070 Super GPU / Gigabyte X870E Aorus Elite Wifi 7 / 1+2TB Samsung Evo Plus M2 Nvme UK2000 Beta Tester
July 24, 201312 yr No AI at all will make a significant difference to overall performance.... He's using UT2. You set the sliders at 0% in FSX with that program because it injects externally.
July 24, 201312 yr Commercial Member Doesn't matter. He did all that change, cranked all sliders out and maxed everything, went to His recording and rendered it frame by frame with FSRender. It probably look an hour to render but it made His slides got turn out smooth. That's how people do videos. His system isn't running the sim like that with those settings. Yeah I think this is what he did, there's no way one could be getting that kind of performance with those insane settings with any consumer PC. It would cause OOM in no time anyway...
July 24, 201312 yr Another weird conflict in the CFG file: the screen resolution for [DISPLAY.device] for DX9 is 1280x768. Most monitors with the 4:3 aspect ratio are 1280x1024. Wide-screen low-res maybe 1280x720. And at one time the rig was ran with DX10, but with [DISPLAY.device] set to 1842x1036. This is the same aspect ratio as 1280x720. I agree with William JSS - something is fishy here... John Howell Prepar3D V5, Windows 10 Pro, I7-9700K @ 4.6Ghz, EVGA GTX1080, 32GB Corsair Dominator 3200GHz, SanDisk Ultimate Pro 480GB SSD (OS), 2x Samsung 1TB 970 EVO M.2 (P3D), Corsair H80i V2 AIO Cooler, Fulcrum One Yoke, Samsung 34" 3440x1440 curved monitor, Honeycomb Bravo throttle quadrant, Thrustmaster TPR rudder pedals, Thrustmaster T1600M stick
July 24, 201312 yr Lots of good meat for speculation: William Sequeira suggested how he probably made the video - it's pretty much well known. Bonchie says his config is a mess - and he's dead right: some of those figures are from FS9 - (TBM @ 400), and many of his settings - especially when combined - are frame killers: The oddball 1842 x 1036 means he's just feeding a television, but, in answer to "what's fishy" - if you look down to the [Main] section, and find "ProcSpeed" you'll see just how well this particular setup is all working. One can delete this number, and FSX will recreate a new one. It was used as a yardstick by the Aces team, and left in the config file for some reason. As many of you know - I have been using DX10 exclusively over the last year, and have been helping many people setup their systems by using TeamSpeak and TeamViewer. I've also been collecting, (over 100) and editing a few fsx.cfg's over that same period, and one of the things I found was that the "strength", or "capability" of any particular pc to run FSX at smooth and constant high frames - is reflected in that number. My pc, along with several other high-end pc's which I've worked with over that period are all up between 19,000 and 21,000 - Hesse's being the highest, my own being at 20606. His number of :- ProcSpeed=10491, shows a pc which is nothing more than mediocre, with those quoted settings being sufficiently crippling as to bring that pc to it's knees. Bogus video. i7 [email protected] | 32GB RAM | EVGA RTX 3080Ti | Maximus Hero VII | 512GB 860 Pro | 512GB 850 Pro | 256GB 840 Pro | 2TB 860 QVO | 1TB 870 EVO | Seagate 3TB Cloud | EVGA 1000 GQ | Win10 Pro | EK Custom water cooling.
July 24, 201312 yr He's using UT2. You set the sliders at 0% in FSX with that program because it injects externally. I use UT2, so I know how it works. You can see from the empty stands at Manchester Airport that he is not running any AI planes in that clip. That will significantly improve the framerate, which means that his movie clip will appear smoother than it would if the skies and airports were populated with AI planes (which is presumably what we all do to make FSX more realistic)? Those scan lines make it look like he was filming a TV! Christopher Low AMD Ryzen 7 9800X3D CPU / 64GB DDR5-6000 RAM / 12GB Nvidia RTX 4070 Super GPU / Gigabyte X870E Aorus Elite Wifi 7 / 1+2TB Samsung Evo Plus M2 Nvme UK2000 Beta Tester
Create an account or sign in to comment