HUGE Fix at Input Lag!!!!

[Anonymous768119]

Well i have to admit after playing few hours, those modifications in nvidiainspector made a significant difference !!

Could you describe what difference?

[diakou]

I mean, if those settings are pure inputlag improvement settings, I don't mind running an inputlag test with my reflex monitor later.

[drjackmann]

curious if anyone but the original poster of this nvidia profile inspector tweak has been able to edit those last variables (MAX_PENDING_CMD_BUFFERS and MCFORCEHOSTSTAGINGBUFFERSIZE). Not seeing them anywhere.

[deama]

https://imgur.com/a/MLU6y17 ¯\_(ツ)_/¯

Are you sure that you do use proper version of nvidiaProfileInspector? https://github.com/Orbmu2k/nvidiaProfil ... g/2.3.0.12

Oh ok, I was using an old version of inspector.

[deama]

curious if anyone but the original poster of this nvidia profile inspector tweak has been able to edit those last variables (MAX_PENDING_CMD_BUFFERS and MCFORCEHOSTSTAGINGBUFFERSIZE). Not seeing them anywhere.

You have to use the newer nvidia inspector version, try this one:
https://github.com/Orbmu2k/nvidiaProfil ... g/2.3.0.12

I'm curious what these settings actually do or if it's just placebo? Haven't noticed much of a difference, but I can measure it later myself using a high speed camera; I reckon there's gonna be 0 difference honestly lol.

[drjackmann]

yeah post back if you end up running any tests. I tried and honestly felt zero difference.

[howiec]

curious if anyone but the original poster of this nvidia profile inspector tweak has been able to edit those last variables (MAX_PENDING_CMD_BUFFERS and MCFORCEHOSTSTAGINGBUFFERSIZE). Not seeing them anywhere.

I'm curious what these settings actually do or if it's just placebo? Haven't noticed much of a difference, but I can measure it later myself using a high speed camera; I reckon there's gonna be 0 difference honestly lol.

The impact of those 2 settings are 100% not placebo with Apex Legends and other games on my system.
Different values change the "mouse feel" and "overshoot/undershoot" as well. Similar to how ASYNC10 or ASYNC9 stuff is 100% not placebo and certainly can affect mouse feel and FPS smoothness.

Ultimately, the best settings depend on your preference, system HW & SW specs, various system/peripheral settings, game played, etc.

[deama]

God I hate how the only evidence we have here with these tweaks --not just the ones in his thread, but for most of these in general-- is "feeling".

Could you guys do a mouse graph test while apex is running in the background with the settings above set and unset?

[howiec]

God I hate how the only evidence we have here with these tweaks --not just the ones in his thread, but for most of these in general-- is "feeling".

Could you guys do a mouse graph test while apex is running in the background with the settings above set and unset?

I know what you mean because I hate not being able to describe things like this at least somewhat quantitatively.

But this is something that you can't unless you want to create an electro-mechanical test fixture that coordinates precise, user-defined, or at the very least, repeatable physical mouse trajectories and then directly compare in-game reticle movement speeds, distances, and "curves" over time via recorded high FPS video (which inherently takes into account total input-lag, frame variance, etc.) for the various settings configs.
So right now the best I can do is describe the changes that are 100% reproducible in a qualitative manner...

FYI, mouse graph tests using MouseTester are relatively crude, and a MouseTester frequency/interval plot doesn't show what's going on in the game/CPU/GPU "chain"... All you accomplish by running a game during MouseTester is adding extra load or "noise" to the system during recording of those plots.

You may be able to use GPU View to illustrate some of the aforementioned settings' effects but that still doesn't exactly describe the "mouse feel" changes.

Just try the different values for the various settings and see for yourself. I'd rather not waste time proving that stuff like command buffers and asynchronous compute are real because... they just are and there's plenty of evidence already out there. That's not debatable. What really matters is what values of these settings work best for you and your use-case.

[ffs_]

God I hate how the only evidence we have here with these tweaks --not just the ones in his thread, but for most of these in general-- is "feeling".

I kinda agree with you. In my opinion besides proper basic setup of BIOS & OS there are not too much useful tweaks which make real difference.