Framerate Tanks When Any FF11 Instance Is In Focus

@Dodik I can't win here, Some of the people, yourself included make it hard to convey anything here, all you want to do is give advice under a veil of negativity.

"A completely fresh install of windows, with only the graphics driver, FF11, Windower, and Dgvoodoo installed"

Yes? that confirms I did the test with Dgvoodoo.

"Initial tests were without Dgvoodoo, used a completely clean FF11 install."

Yes? this confirms before I Installed Dgvoodoo and tested with it, I did a full clean test.

The "Initial" bit is the important bit there.

Is there a reason you only want to offer advice whilst being negative about it?

I do thank the people who have given advice without the animosity, I will not name them, but there have been a few people here who genuinely have tried to help and managed to do it in a respectful manner.

I do not want to get drawn into these negative comments, but I also have been walking on eggshells because of responses like yours, in an effort not to upset the very people who are being negative.

So please, it's not difficult to communicate your advice/thoughts in a polite manner.

So to clarify - with what you said above - this issue doesn't persist across the board? and you have to go to specific zones or areas of the game for it to manifest completely?

For example you seem to indicate Sheol C / Sortie are big culprits?
Also - are we only talking about duo boxing here - or are more than two chars involved? How many total chars do you have at once?

"it is still confirmable by checking CPU cores/threads and unlocking FPS though, "
I would argue that that is not a valid test or method to reproducing this issue. Unlocking frame rates is introducing something completely different into this equation.

And yes, with my mule just reactivated and logged in, zero issues hitting 58fps on Win 11 with an intel cpu for both chars.

Thanks for your testing.

Yeah, the issue is even worse tri and quad box.

My unlocking framerate suggestion was only if you could not find somewhere to push you below 57, totally agree with you about it not being the best way to test it.

I only suggest it if you have no way to get the fps below 57.

Over 90% of game areas I am at 57fps, so only have a small amount of areas outside of instanced zones to run tests inside a reasonable timeframe unfortunately.

But again, I can confirm the issue is present in zones like sortie, Odyssey, etc.

I don't see where I was negative, that seems to be coming from you.

The reason you are getting push back and constant reiteration to do one thing at a time is because your tests and what you say you have changed do not add up.

It is next to impossible to help someone remotely that cannot be consistent in what they say they changed and what they actually did or says they did one thing while they actually meant something else. How can anyone know you tested without installing dgvoodoo when you specifically say you installed it.

If you can actually replicate with a clean install then that's a lot easier, start by minimising changes even further and work backwards. Only change video drivers. Still issue? Use different driver version. Keep going until one version works or you discount drivers from being the culprit. None of this will work if any changes other than drivers are made.

I am sorry you feel this is negative towards you and good luck with your issue.

What is your core usage on the cores your FF11 instances are on if you don't mid me asking Yottaxa>

But not anywhere else?

Is it possible you are in some way just hitting the limits of your cpus? Could the frame rates on an unfocused tab just be phantom or separate issue?
For example I would not expect multi boxing performance on the surface pro to be amazing - so what were the specs again on the other two intel cpus?

Ill check cpu usage on core later.

Thank you for this, and it is a great suggestion, and one I have not tested alongside the fresh install.

I went to bed after some testing on the fresh install, so I have yet to try other things yet.Those and a handful of non instanced zones, Western Adoulin, Nashmau, Whitegate, Batallia Downs (S).

The instanced zones were only during alot of enemies, characters and effects on screen, the non instanced zones were just aiming my camera in static spots.

In all the zones where the framerates take those big hits, they all run at 57 capped if the windows are out of focus still, if that last bit of info can help in any way.

This is a common issue with those zones and I get the same even without multi box. It comes from the smoke effects that those zones have.

Which is why I suggested using the mog house and not any outside zone.

If the issue only happens in those zones it is not anything you can fix, other than dat modding to remove the smoke effects that is available somewhere.

And you seem to be conflating two issues, if there are any issues at all. The window being unfocused may make the game stop displaying effects, idk never noticed.

Those zones having FPS drops is "normal" for those zones. If that is the only issue you have you can stop trying to fix it - other than dat modding nothing will.

the 2 PC's are running

PC1: I9-9900k
RTX 3090
32gb DDR4 ram
FF11 installed on an NVME
Windows 10

PC2: Intel 11400F
RTX 4060ti
32gb DDR4 ram
FF11 installed on an NVME
Windows 10

Just let me know if you want more specs.

There's no way for me to get low FPS inside a moghouse, but if my CPU cores are still behaving in the same way inside a moghouse will that test suffice?

Again, I know people do not like testing with uncapped framerates, but if I uncap them inside the mog house in a following test, I will be able to see if the Cores maxing out behaviour is still having an effect on overall framerate.

Just to confirm, I am going to test exactly what you just told me first, once I confirm the fps stay at 57 constant I will check the usage on the cores the FF11 instances are bound to whilst inside the mog house.

The other test with framerates uncapped I will only do after doing your test, I hope that is acceptable?

...

Not getting it. Uncapping framerates means the game is running at 100% of what a single POL process can do without limits (no fps cap). Naturally, that means the cpu core running that process will be maxed out at 100% utilisation.

What does this prove other than the fact that uncapping framerate will max out cpu usage.

If you cannot replicate the issue inside the mog house you do not have an issue.

Core usage while frame capped is pretty much irrelevant, there are a lot of variables there. What you see as 100% usage on one core could be all processes being scheduled on that core while the rest are idle.

The issue claimed is FPS drops while in focus. That is what should be measured.

I can confirm a constant 57 FPS inside moghouse.

No real noticeable jumps in Core usages, all staying around the 20% utilisation range.

I did try it in a following test with uncapped FPS.

Still no noticeable change in core usages, still in the 20% range, but here's the strange thing now.

The problem has inverted inside mog houses, I'm at over 400fps on both instances, and losing roughly 30-60 fps on both instances when I do not have an FF11 instance in focus.


I do understand that, but how am I meant to test if framerate is being affected if I'm only testing areas where the framerate stays capped.

I do still have an issue when I go into zones where the framerate drastically drops, and I would 100% agree with you that it's a non issue, if all my FF11 instances didn't all max out their framerates in those zones the moment they are background windows.

I also thought the same as you, uncapping my framerate in the moghouse was just going to force my FF11 instances to max out their CPU cores, but it has me stumped, uncapped they stayed in the 20% utilisation range, so finding out that alone, made the test useful in 1 way.

I still don't really think OP is engaging fairly with this issue, and am pretty certain it's gsync/vsync related. But, I'll give them the benefit of the doubt here and point out that:

-Vsync/Gsync change the timings of when the game will be allowed to draw another frame. This means that you can be uncapped CPU and still not get the framerates you're intended to because the frame scheduling is preventing the next frame rather than a hardware limitation.

-Vsync/Gsync issues will likely not present when maintaining the target level. So, theoretically, frame limiter may be targeting 120 frames per second to match their monitor. When in MH, they can keep up with the scheduler, and thus they sit at 60. When outside MH, their CPU can keep up with 60 FPS but it can't keep up with 120, so issues with frame scheduling cause it to present less frames when game is in focus and sync is active than when it is not and scheduling is only being controlled by FFXI.

His issue is perfectly explainable by this, but since he insists he's correctly tested every possible variation.. I guess there's just no answer. Alternative explanations would almost all still effect you within the mog house. Anything input or windower related should still matter there. So, if you're not seeing a substantial drop inside the mog house, you can reasonably rule those out.

This is the first time that has been fully explained, the only thing, 1 of the 3 PC's tested was not Gsync, and I have disabled it in my testing of the other PC's.

With you being so confident the issue is this though, I know you are very knowledgeable and have not, and will not ever discount the information you provide.

I am going to try and source another monitor today if I can, as one last test.

Another one without Gsync, and if possible only 60hz.

You don't need to get a new monitor. You can limit the framerate of the monitor via software. I would recommend:

-Close all instances of FFXI.

-Right click desktop, choose display settings.
-Scroll down to 'advanced display settings'.
-Select your monitor, Set the refresh rate to 60(it may offer 59.9 or something instead, that's fine).

-Right click nvidia status bar icon -> NVIDIA control panel.
-Select 'Set up G-SYNC'.
-Disable G-Sync by unchecking 'Enable G-SYNC, G-SYNC Compatible' for each display, and clicking apply.
-Select 'Manage 3D settings', go to 'Global Settings', set the following:
Background Application Max Frame Rate: Off
Max Frame Rate: Off
Preferred refresh rate: Application-controlled
Vertical sync: Use the 3D application setting.
-Hit apply.
-Go to program settings. If pol.exe is present, or windower.exe is present in the dropdown, apply the same settings here.

-Open a new single instance of FFXI, with just one character. Stand in a busy area. Uncap FPS. Note the FPS, CPU usage, and GPU usage with and without the instance focused.

I know it might be frustrating, but given that this issue isn't constant or persistent, I am more inclined to suggest it is actually a non-issue in the context of it being a limitation of cpu/multiboxing and known laggy zones. To that point - does this issue ever manifest single box? or is it solid 58 full time when single.

He insists that under the same circumstances, he can maintain 60 FPS while out of focus and not while in focus. If you take him at face value, this shouldn't be the cause.

Stays solid 57/58 single box, yeah.

Managed to find an old monitor in my attic, 1080p 60z screen, just done a test on it, issue still persists.

Thank you so much for all your help Yottaxa, and others who have helped.

When the thread descends into insinuations of lying, think that's a good time to call it.

here is the proof FYI: your insinuation doesn't deserve validation but I gave a response anyway.

Thorny and Dodik, please would you mind not commenting in this thread? Your attitudes have not been helping.

edit. It's recorded at 3840x2160, and the FPS is in the top right corner of both windows, you may have to maximise to be able to read it.
Wanted to keep the tests as clean as possible, so only used windowers built in FPS overlay, which isn't the easiest to see unfortunately, but is readable maximised on youtube @4k.

Please everyone this is only a video to prove Thorny's insinuation wrong, No things like, try windowed fullscreen etc. I have tried those things, this is a single video in this single setup just to show the issue as it's happening, please be assured I don't only test in this configuration.

Computers aren't mystery devices. Everything operates in a preset way, and diagnostic data can be used to tell which causes are more or less likely. I didn't say not to take you at face value, I simply said that if your claims are taken at face value multiboxing cannot be the cause.. which is the same thing you believe.

I put considerable thought and effort into analyzing how your claims could all be true at once, and put out a potential explanation with proper diagnostic to begin evaluating it. Yet, you completely ignored my instructions and decided to nitpick a seperate comment instead.

You have 4 pages of telling people you've tried everything under the sun, but you're unwilling to do diagnostic work in controlled conditions to narrow down the cause. It's absolutely asinine, and this time I am intending insult.

Good luck.

You really don't know how that insinuates I'm lying?

You're an intelligent person, you know what you meant when you wrote that, you have played dumb and not owned up to anything you have done in this thread, at least in your last message your insult wasn't passive aggressive and you owned it.

You keep putting your head in the clouds, and keep repeating that I have not done the testing that people have recommended, even after each test that someone recommends, I have responded with the results.

The only one today I have not done yet was your list of Gsync instructions, and only because I was in my attic looking for a monitor, so I must have missed it when I finished the testing and returned to check FFXIAH.
And like every other bit of advice I have been given so far, I will do those instructions exactly as you wrote.

But know this, I did it in that exact way yesterday already, but have no issue repeating it.

As for you saying I am not using controlled conditions, I don't know how a full windows re-install with the absolute minimum required things installed isn't a controlled condition.

You seem to be a person incapable of owning up to your shortcomings, I hope you're nicer to people in real life and this is just one of those cases of someone behind a keyboard.

I have requested this thread be locked, thank you to the people who helped.

Has all the testing been on the LG TV? You could try making sure it's set up for PC rather than TV in the nvidia control panel and set up PC/Game mode on the TV. Maybe try with an actual monitor using display port when you can.

Thanks for the suggestion.
That video is a bit out of context, was just to show the issue happening.

My tests have been across multiple PC's and monitors and this LG TV, the TV PC already had screen recording software installed, so just used the LG for the video.

Used both displayport and hdmi.

My tests since last night are now on an MSI monitor through displayport on a completely barebones fresh re-install of windows and FF11, except for the little bit of testing I did on the 1080p 60hz non gsync monitor I got out of my atttic today.
I'd switched full time to using the MSI monitor and PC for testing around 48 hours ago, and re-installed windows last night.