May 17, 20206 yr 29 minutes ago, Noel said: I did an acid test of sorts last night w/ NGXu into KSFO-HD in ORBX NCA regional scenery and the outcome was essentially identical to not using FFTF. I could see the FFTF value in FFTF Dynamic was stuck right at where you would want it in this very hard to process scenario the closer I got to KSFO-HD, at 0.01. And here's the thing--it was no better than w/o FFTF. Frames dropped into the toilet in the worst spots just as it does w/ a fixed value of 0.33. This suggests to me the amount of work the main thread is tasked with for terrain loading is small relative to the total work involved the non-terrain loading domains for the main thread. The bulk of the work for terrain loading is being accomplished by all of those other logical processors in the 8 core CPU. So at an FFTF value of say 0.33, maybe 10% of the total capability of the LP assigned to the main thread is involved in preparing terrain to be sent to the other LPs which do the bigger part of the work of terrain loading. And at an FFTF of 0.01 maybe that 10% goes down to 0.1%. In either case there is simply nowhere near enough resources to borrow from the terrain loading prep work done by the main thread to overcome the non-terrain loading work the main thread is tasked with in these types of scenarios. So this suggests to me there will be some scenarios where it will matter, when that extra 10% more frame rate may keep me at >= 30fps, the critical value to maintain when you run unlimited/vsync to 30Hz. It is therefore a tool to use when you already have your sliders set up for a given flight such that very likely you're going to be able to maintain 30fps, but if you misjudged (by 10% or less!) FFTF Dynamic may help you maintain that 30. I learned quickly the best way to grapple w/ the above scenario is simply to deselect ORBX NCA for airplanes as hard on processing as the NGXu is, and it's totally fine that way. In fact, in may ways in the bay area it can look better w/ FTX Global only. My early conclusion therefore is: FFTF Dynamic adds just a little value when you have a higher end CPU in P3D4.x, and the real challenge is getting your sliders set up optimally for plane and scenery you're going to be in, as it's always been! Easy explanation : the main core divides its time per frame between performance and terrain assignments. The additional terrain cores get their assignments from the main core. So a lower FFTF will result in better performance but less assignments for the additional terrain cores. A higher FFTF will result in a lower performance and more assignments for the terrain cores. FIBRE FRAME TIME FRACTION is a direct command that helps tune the amount of time per frame that the CPU runs fibre jobs. Its default value is 0.33, meaning that the CPU will spend 1 millisecond on scenery loading in every 3 milliseconds of rendering. So, the lower the value, the better the frame rate, but there is also an increase in the blurries. Performance Tuning Tip: This setting, which defaults to a value of .33 (33%), defines the percentage of each frame that is devoted to loading scenery. Increasing this number can reduce "the blurries", but it can cause stutters, and can also reduce the overall frame rate. Try reducing the number to the lowest level where you still get smooth scenery paging. This will vary depending on disk speed as well as the type of flying, as well as the selected scenery settings. Edited May 17, 20206 yr by GSalden 5950x3d 5.4-5.7 GHz - Asus ROG 870 Crosshair Apex - GSkill Neo 2x 24 Gb 6000 mhz / cas 26 - MSI RTX 5090 Gaming Trio OC - 1x SSD M2 6000 2TB - 1x SSD M2 2800/1800 1Tb - Corsair 5400 case - Corsair 360 liquid cooling set - 3x 75’ TCL tv. 13600 6 cores @ 5.1 GHz / 8 cores @ 4.0 GHz (hypterthreading on) - Asus ROG Strix Gaming D - GSkill Trident 4x Gb 3200 MHz cas 15 - Asus TUF RTX 4080 16 Gb - 1x SSD M2 2800/1800 2TB - 2x Sata 600 SSD 500 Mb - Corsair D4000 Airflow case - NXT Krajen Z63 AIO liquide cooling - FOV : 200 degrees My flightsim vids : https://www.youtube.com/user/fswidesim/videos?shelf_id=0&sort=dd&view=0
May 17, 20206 yr That is a clear technical explanation thanks and it appears w/o the technical details I had the concept correct. That being said I think my conclusion still stands--it's not doing much of anything useful in a practical sense at least when frames are set unlimited in-sim. In fact no matter where I am in flight I see the realtime FFTF displayed is more or less stuck on 0.01 with an occasional blip to 0.33 including up in the air away from complex terrain loading demand and where there is plenty of main thread headroom and I would have assumed the realtime FFTF would have been more stuck on 0.33, but nope 0.01. Perhaps I don't understand what I'm looking at which is a distinct possibility! 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.
May 17, 20206 yr Here I use AGL mode and can clearly see the FFTF value going up when climbing till my max set altitude. In Framerate mode the FFTF only goes up when there is headroom between min and max. And goes down when the franerate is dropping. If you only see 0.01 then probably your system has difficulty to keep up performance , that is with the settings in FFTF Dynamic you are using, so no headroom left ... Edited May 17, 20206 yr by GSalden 5950x3d 5.4-5.7 GHz - Asus ROG 870 Crosshair Apex - GSkill Neo 2x 24 Gb 6000 mhz / cas 26 - MSI RTX 5090 Gaming Trio OC - 1x SSD M2 6000 2TB - 1x SSD M2 2800/1800 1Tb - Corsair 5400 case - Corsair 360 liquid cooling set - 3x 75’ TCL tv. 13600 6 cores @ 5.1 GHz / 8 cores @ 4.0 GHz (hypterthreading on) - Asus ROG Strix Gaming D - GSkill Trident 4x Gb 3200 MHz cas 15 - Asus TUF RTX 4080 16 Gb - 1x SSD M2 2800/1800 2TB - 2x Sata 600 SSD 500 Mb - Corsair D4000 Airflow case - NXT Krajen Z63 AIO liquide cooling - FOV : 200 degrees My flightsim vids : https://www.youtube.com/user/fswidesim/videos?shelf_id=0&sort=dd&view=0
May 17, 20206 yr 7 minutes ago, GSalden said: If you only see 0.01 then probably your system has difficulty to keep up performance , that is with the settings in FFTF Dynamic you are using, so no headroom left ... Then why is it that while I'm up above the Nevada dessert at FL370 in the NGXu and showing a main thread CPU load at say ~40%, and the terrain LPs coasting along at very low percentages, and yet still the realtime FFTF is pretty much stuck at 0.01? I have the Frame Range set at 31-33, because once again, all's super smooth until we hit <30 FPS, so the goal is to always keep frame rate at 30 or greater by modulating FFTF. When then doesn't FFTF displayed jump back up and stick at 0.33? As I say I might not understand how to set this up. Maybe something w/ the AGL setting is making this happen. Now if I lower the Frame Range to say 20-24, then yes the realtime FFTF will go up to 0.33. But again, why the above described behavior when in a very low demand scenario? This flight was from KRNO to KSAN-HD in Orbx So Cal. Frames stayed at 30 the entire way into 09 w/ one short stammer just before TD when frames dropped to 29 briefly. Do I need to also set something with the AGL slider as well perhaps? 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.
May 17, 20206 yr Only a guess: the update cycle is rather rude and slow, at least to my experience. And FPS fluctuate a lot in P3D. Means: if you are only a part of a second below 30FPS, you might even not notice but the tool thinks you are not above 30 and keeps low FFTF? I always use AGL, served me best so far. Greetings, Chris AMD Ryzen 7 9800X3D, 2x32GB DDR5 6000MT/s RAM, MSI RTX 4090 Ventus 3X, Windows 11 Home, MSFS2024
May 17, 20206 yr This whole topic has become so confusing it is impossible to determine what settings to use. In P3Dv4.5 I use Unlimited + VSYNC OFF + TB OFF. I use RTSS and limit frames to Scanline sync x/2 (-1). I have always set FFTF on AGL Mode with AGL Range 0 - 4000 and FFTF Range of 0.01 - 0.4. I fly GA aircraft on relatively short flights and have some intense scenery. But I have been pretty happy with performance (see specs below), but I don't know why anymore... I just tried a variety of settings P3D limited to 60 then 30, then back to unlimited, with AGL mode and then FPS Mode set 20 - 30, and both with and without RTSS scanline. Just one data point, but it seemed like P3D unlimited + VSYNC OFF and AGL Mode with RTSS at Scanline x/2 was the overall best for FPS / terrain loading (stutter & smoothness) / CPU utilization and heat. Does any of this make sense? Not sure it does to me anymore... @Achilles Philippopoulos? @GSalden? [CPL] : I9-9900K @5.0GHz HT ON, Maximus XI Hero, ASUS TUF RTX4080 OC, 32GB DDR4 3200 14, 1TB NVMe SSD, 500GB SSD, 1TB HDD, 40" Samsung 4K TV, Honeycomb Alpha & Bravo, Logitech Rudder Pedals, WIN11
May 17, 20206 yr 19 hours ago, lownslo said: Thanks Edward, nice to know I'm not the only one. I tried your recomendations and still no joy. I've deleted and then re-added my address for both a Billing and Delivery Address. It insists that I have a company name and VAT number... I'm not buying this product for my company, and I'm from the U.S. I appreciate your time and effort but I'll be updating to P3Dv5 without this app. Simply not worth the grief! Greg Sorry it did not work out. I was asked for VAT number only when I selected the option for invoice. Deselecting Invoice option allowed me to proceed to checkout. Wen you add new address it should show as a blue on top of the address with small checkmark on the right corner. It's all you need. Do not select anything else just terms of agreement. I do not have v4.5. I have only v5 now and I still decided to get it because I was offered a discount from the previous version. Not a huge amount though. I will play with it when PMDG 737 and 777 come out and then I will make a decision if I am gong to use it or not. I9-13900K | ASUS ROG Strix Z790-E Gaming LGA 1700 | MSI Gaming GeForce RTX 4090 24GB | CORSAIR iCUE H150i ELITE LCD Liquid Cooler | CORSAIR DOMINATOR PLATINUM 64GB (2X36) 5200MHx DDR5 | Thermaltake GF3 1650W 80+ Gold PSU | Samsung QN90C Neo QLED TV 50”
May 17, 20206 yr Commercial Member 2 hours ago, Noel said: Then why is it that while I'm up above the Nevada dessert at FL370 in the NGXu and showing a main thread CPU load at say ~40%, and the terrain LPs coasting along at very low percentages, and yet still the realtime FFTF is pretty much stuck at 0.01? I have the Frame Range set at 31-33, because once again, all's super smooth until we hit <30 FPS, so the goal is to always keep frame rate at 30 or greater by modulating FFTF. When then doesn't FFTF displayed jump back up and stick at 0.33? As I say I might not understand how to set this up. Maybe something w/ the AGL setting is making this happen. Now if I lower the Frame Range to say 20-24, then yes the realtime FFTF will go up to 0.33. But again, why the above described behavior when in a very low demand scenario? This flight was from KRNO to KSAN-HD in Orbx So Cal. Frames stayed at 30 the entire way into 09 w/ one short stammer just before TD when frames dropped to 29 briefly. Do I need to also set something with the AGL slider as well perhaps? Hi, When your frames jumps up and down simply means that your pc can not handle your p3d settings. Also setting frames to min 30 and max 31 with low fftf 0.01 and max 0.33 is not recommented. At 30.4 frames you will have 0.01 fftf and at 30.6 frames you will have fftf 0.33 Just go to an heavy frames airport, set agl (0...10000 ft) mode, set low fftf to 0.01 and high fftf to 0.25 or above. Lock your frames to whatever you like (for example 120 or 30). Press control+Z to see frames. Set your p3d settings according to your pc specs. Try to have as stable frames you can around 24. Thats your pc is capable for with fftf 0.01. Now, you can use the agl mode, or you can set the frame rate scenario with min 24 and max 30 frames. With this example you will have at 24 frames fftf 0.01 and at 30+ frames the max fftf setting. Between 24 and 30+ fftf value will change. This will offer you the best flight experience you can have with your pc. Covering the worst case is the best solution. Will make your worst case smooth and flyable. So, you will never have to stop a flight, change settings etc. Edited May 17, 20206 yr by Achilles Philippopoulos Achilles Flight Simulator Plaform Solutions
May 17, 20206 yr 44 minutes ago, Achilles Philippopoulos said: Hi, When your frames jumps up and down simply means that your pc can not handle your p3d settings. Also setting frames to min 30 and max 31 with low fftf 0.01 and max 0.33 is not recommented. At 30.4 frames you will have 0.01 fftf and at 30.6 frames you will have fftf 0.33 Thanks, that explains why I see fftf at 0.01 almost always. I will try your method and see how it's "smooth" and "flyable" compares with what I experience now--which is perfect, liquid smooth animation UNTIL I happen to get too close to <30 frames. Having used this method for 5y now and found it superior to all others and I almost never get to <30 because I understand how to set it up to avoid that. On occasion I misestimated the setup and have had to stop and readjust, but after 5y one figures it out pretty well. I will be surprised if 'smooth and flyable' at 24 frames compares w/ perfectly smooth and flyable at 30 that I had even w/ a 6.5y/o PC. If it's comparable, then we win. If not, we simply get rid of FFTF Dynamic and call it a charitable contribution to the cause. FFTF Dynamic could have been much more robust if it could also assess GPU demand and manipulate the lighting sliders in the background as well when GPU load approaches 100%. Now our friend Gerard maintains you can have the internal limiter set an UNLIMITED, yet still find some value in FFTF Dynamic, at least I think that is what he is hinting. You seem to be contradicting this here. 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.
May 17, 20206 yr Achilles, I can tell you one thing only...First time I am in Greece I owe you bottle of Metaxa.....! This morning I start reading this forum about FFTF and it was enough for me that some very reputable Avsim members gave Thumbs up. I bought it, made few recommended adjustments and difference.....HUGE....FSDT KLAX and 747 - I have gain between 7-10 FPS?! and smoothness is way better....minimal stutters. Completely different sim....and if I gain even 2 FPS at FTCYYZ I'll buy a dinner too....! I am recommending to everyone that has issues at heavy airports, FFTF works as advertised with my moderate hardware....We all know that sceneries could be issue and heavy on frames and we will never have perfect view, but for me this is huge step forward.... Alex
May 17, 20206 yr Commercial Member 33 minutes ago, cyyzrwy24 said: Achilles, I can tell you one thing only...First time I am in Greece I owe you bottle of Metaxa.....! This morning I start reading this forum about FFTF and it was enough for me that some very reputable Avsim members gave Thumbs up. I bought it, made few recommended adjustments and difference.....HUGE....FSDT KLAX and 747 - I have gain between 7-10 FPS?! and smoothness is way better....minimal stutters. Completely different sim....and if I gain even 2 FPS at FTCYYZ I'll buy a dinner too....! I am recommending to everyone that has issues at heavy airports, FFTF works as advertised with my moderate hardware....We all know that sceneries could be issue and heavy on frames and we will never have perfect view, but for me this is huge step forward.... minimal stutters can be fixed from fine tuning of the settings. It up to you and your pc. Thanks for the bottle 🙂 Achilles Flight Simulator Plaform Solutions
May 17, 20206 yr 12 minutes ago, Achilles Philippopoulos said: minimal stutters can be fixed from fine tuning of the settings In FFTF? I locked sim to 30, NCP is locked at 59 and, refresh rate is 59Hz as well (ASUS VA326, not even sure should I keep it there since is 144 Hz). What I have in in FFTF is 4 times per sec. Even that I don't see my FPS are not crossing 30 FPS I still see .33. Perhaps that's normal to jump sometimes over that even not displayed on FPS monitor... Alex
May 17, 20206 yr 23 minutes ago, cyyzrwy24 said: In FFTF? I locked sim to 30, NCP is locked at 59 and, refresh rate is 59Hz as well (ASUS VA326, not even sure should I keep it there since is 144 Hz). What I have in in FFTF is 4 times per sec. Even that I don't see my FPS are not crossing 30 FPS I still see .33. Perhaps that's normal to jump sometimes over that even not displayed on FPS monitor... The lower your monitor can go on refreshrate the more breathing space for your graphics card... Here I have 2x 4K in NVSURROUND. If I try to run that at 60 hertz the whole image is purple and with a false looking image ( 2080 Ti ).Therefore I use 25 or 30. You can try it yourself : look at the GPU and memory load of your graphics card at 60 or 30 hertz. In your case 144 / 72 / 36. 5950x3d 5.4-5.7 GHz - Asus ROG 870 Crosshair Apex - GSkill Neo 2x 24 Gb 6000 mhz / cas 26 - MSI RTX 5090 Gaming Trio OC - 1x SSD M2 6000 2TB - 1x SSD M2 2800/1800 1Tb - Corsair 5400 case - Corsair 360 liquid cooling set - 3x 75’ TCL tv. 13600 6 cores @ 5.1 GHz / 8 cores @ 4.0 GHz (hypterthreading on) - Asus ROG Strix Gaming D - GSkill Trident 4x Gb 3200 MHz cas 15 - Asus TUF RTX 4080 16 Gb - 1x SSD M2 2800/1800 2TB - 2x Sata 600 SSD 500 Mb - Corsair D4000 Airflow case - NXT Krajen Z63 AIO liquide cooling - FOV : 200 degrees My flightsim vids : https://www.youtube.com/user/fswidesim/videos?shelf_id=0&sort=dd&view=0
May 17, 20206 yr 3 minutes ago, GSalden said: You can try it yourself : look at the GPU and memory load of your graphics card at 60 or 30 hertz. In your case 144 / 72 / 36. This is what I have as a current GPU usage.... in the air currently...https://imgur.com/a/aUpfp7A Does that mean that I should use 72 or 36 instead of 59? Alex
May 17, 20206 yr 5 hours ago, bbuckley said: This whole topic has become so confusing it is impossible to determine what settings to use. In P3Dv4.5 I use Unlimited + VSYNC OFF + TB OFF. I use RTSS and limit frames to Scanline sync x/2 (-1). I have always set FFTF on AGL Mode with AGL Range 0 - 4000 and FFTF Range of 0.01 - 0.4. I fly GA aircraft on relatively short flights and have some intense scenery. But I have been pretty happy with performance (see specs below), but I don't know why anymore... I just tried a variety of settings P3D limited to 60 then 30, then back to unlimited, with AGL mode and then FPS Mode set 20 - 30, and both with and without RTSS scanline. Just one data point, but it seemed like P3D unlimited + VSYNC OFF and AGL Mode with RTSS at Scanline x/2 was the overall best for FPS / terrain loading (stutter & smoothness) / CPU utilization and heat. Does any of this make sense? Not sure it does to me anymore... @Achilles Philippopoulos? @GSalden? If I use Unlimited + VSYNC OFF in P3D and limit frames in RTSS to Scanline x/2 does FTFF have any effect in AGL mode or just Frames mode? [CPL] : I9-9900K @5.0GHz HT ON, Maximus XI Hero, ASUS TUF RTX4080 OC, 32GB DDR4 3200 14, 1TB NVMe SSD, 500GB SSD, 1TB HDD, 40" Samsung 4K TV, Honeycomb Alpha & Bravo, Logitech Rudder Pedals, WIN11
Archived
This topic is now archived and is closed to further replies.