ankh21

No, running on the 3D cores was fine, AMD driver controlled or via BIOS settings, both gave priority to the 3D cores. I don't know, if is was due to a BIOS / MOBA problem or something else. I made two BIOS upgrades after that but was not in the mood to enable the virtualization back. I cannot remember exactly...i think, it was in the FsHUD Discord channel, where another user mentioned a similar performance problem, and disabling SVM mode in the BIOS did the trick for him, too. Might be a very specific problem of my MSI motherboard. Just an idea, as i said.

As someone said before, it's not very helpful to fiddle around the settings without a plan. But just as an idea: when i upgraded my system to a 7950x3D, i was shocked by the bad performance. The culprit was the motherboard's virtualization setting. After deactivating this in the BIOS, the performance was the expected one. I have to confess, that i am still on FS2020 with the 7950 CPU and a 4080 Super, so i cannot comment on the FS24 settings, but perhaps a closer look in to the BIOS might be a good idea. Wish u luck, Matthias

As far as i read the latest message from Reset XPDR, you have to enable the experimental settings manually, so if you don't want to use it or test it, just leave the app as it is. To use the inbuild 2024 "dynamic" setting together with the app's "auto" setting was a special wish of another user. If there is no experimental work or developement, we would't be at the point AutoFPS has brought us during the last month...so, i trust in the carefull adoption the author is working on...and don't activate the new, experimental feature as i am on FS2020 🙂

Oh, tell an Italian that cooking pasta is easy 🫣 Beside of jokes, I think, that asking a question is never wrong! And the author shows on so many pages, that is is always willing to give assistance and explanations! And if you need to translate the Readme to another language, there is double of work for you, so anybody will call you lazy. So my apologies if my answer before sounded like an accusation, that was not my intention!

As they mentioned, they were using the DynamicLOD before, where there is just a adjustment depending on given steps. But the BIG advantage of AutoFPS is the „auto“ 😂 And I can strongly recommend to read the readme. It’s long, but written very systematic. So really most of the questions here are answered there already! At the beginning I was experimenting a lot, but settings fitting to one situation were not ideal in another situation. After really working with the readme, I now have a setting perfect for my system and needs. And I am just doing boring IFR flights. The potentials of this marvelous peace of software are much bigger, seeing the latest version e.g.

Isn’t that’s the way it works? You are in FPS Cap mode with IFR. So, if your system easily achieves the set TLOD of 200, it increases the TLOD up to 2x of your limit, so you end up with 400 or 500 when your limit is set to 250 before. The upper limit should be the TLOD setting your system can run on in nearly every situation. If you want to stop at 200, you have to use another mode i guess. The readme is very detailed, so perhaps you will find another mode stopping at your given limit. Or you set your upper TLOD limit to 100 or 125, so you will see 200 or 250 with cap and IFR.

Oh,wow, thank you so much for this excellent explanation! That’s I like most with your support, you always explain what the software is doing and why. But I just forgot that there is a config file to look at. I will try to adjust the mentioned line a little bit. Running the system close to its limits is all we want since the first day of Flight Simulator 😂

May i ask a question: I use the Cap option, with FG on limited to 60/sec. Works like charm. But sometimes, AutoFPS tries to increase the TLOD, framerate drops a little bit, but TLOD is lowered more than it has been before. For example, i am at FL400 over Germany, which seems to be very demanding on TLOD. So I am very happy seeing a TLOD of 275 or even 300. Then, AutoFPS starts a new seeking phase, but after increasing just a little bit, 280 or 305, it decreases instantly to 225 or 250. So I end up with less TLOD than before. It makes sense to lower TLOD significantly to stabilize frames in demanding situations, but is there a way to recover to the level before? Or to decrease in smaller steps? Thank you so much! Matthias

Interesting... But one stupid question i didn't find the answer to...back to TAA in the AA settings? At the moment, i am on FS2020, Ryzen 7950x3D, 4080 Super, DLSS Quality (Render x 0.8) and DSR 2,25, which is exactly the opposite of Image Scaling. During daytime, ghosting is not as obvious, but at night it's very annoying. And DSR is perhaps more demanding on the CPU. Will give it a try.

Thanks for your explanations. Yes, i already tried to switch back to DX11, but due to a known bug, FS won't start in DX11, although i never installed and do not use the NVIDIA App, just NV Inspector. And i was too lazy to do a reinstall, neither of FS nor of the NVIDIA driver. So i decided to turn in DX12 again, and untick the VRAM + seeting, as the results with your app are wonderful even without this option turned on. Greetings, Matthias

The latest test version 2 with changed FPS Cap algorithm gives be the best results I ever had with your software heaven sent 😇 Thank you so much! Rock solid 60 fps (FG on) in nearly every situation. Interestingly, I had to disable the VRAM+ option (FS 2020), as it lowered the LODs as expected, but much harsher than the bare FPS cap alone. Sitting in my PMDG 737, EDDF, with traffic around, GPU-Z says 100% VRAM on my 4080 Ti, but TLOD is 100 or even 150 without any stutters (ok, panning in external view does stutter a little bit). With VRAM + ticked on top, TLOD drops to 50 or less cause of the max filled VRAM. Even lowering texture resolution from ultra to high has only little effect in this scenario. I suspect, DX12 fills the VRAM with nearly everything it can get, so it triggers the VRAM + reduction unnecessarily. Very strange. So, again, thanks a lot, your marvelous piece of software gives me the best after-work-relaxing short hops across Europe I ever had 🤩

Yes, as it is stated, that F is for Ultra PERFORMANCE now, not ultra quality as with the version before. So, at the moment, there seems to be no way to have ultra quality in the sim.

Hm, so no Ultra-Quality at all? That’s a pitty. I tested now with Preset E, which is marked as „unused“ in my NVIDIA inspector’s settings, and cannot tell any difference with Preset F and the new, max available „quality“ setting in MSFS. Compared with the ultra-quality setting before, it might be a little bit more blurred, but it’s hard to tell, cause I cannot do a real side by side pllit monitor comparison. I am very curious about your findings and hints in that. Matthias

I will have a close look to both products, but watching the REX video i really think, the turbulent shaking is only a camera effect. The wings of the airplanes are rock steady. Perhaps i am wrong, but beeing skeptical in this point, the whole promotion gets the smell of some hot air. I will wait and see. Matthias

Thanks for explanation :smile: At the end i only wanted to assist soulflight to choose the correct settings for a beautiful DX10 experience. But as we just talk about these details... why is the unconverted effect less demanding on the CPU than the converted? Where is the difference in terms of CPU workload between a DX9 / unconverted file and the converted DX10 effect? Matthias