Windows 8.1 Mouse Lag Negatively Affecting Some Games?

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
Are any of you experiencing this mouse pointer issue with Windows 8.1?

It's only video games that don't use Raw Input for reading mouse movements that are in trouble. Something has changed in the way Windows updates pointer positions and that's what is apparently causing the problems.
 
It happens only in few games which don't use raw input, but ask the OS for the mouse location, in multi-monitor configurations, for users who use both high DPI and high poll rate mouse. That minimizes the number of affected users to a very small number.
 
Since upgrading I've had 8.1 lockup several times. It seems like chrome + youtube has a tendency to lock my system up but I've also had games grind everything to a halt as well.
 
Holy shit this is bad. Thanks for linking the article Steve. I can't navigate the menus in Deus Ex it's so bad.
 
I haven't noticed it at all, with more than 20 games installed, ranging from 20 year-old games to current. But I don't own either of the 2(?) games mentioned. 8.1 for me is rock solid. Like faugusztin mentioned, the planets practically have to align before this becomes an issue. Every single one of the old unreal-engine games runs fine for me under 8.1, including Deus Ex.
 
Must not be an issue with Logitech MX518. I've at least played DX:HR in Windows 8.1 and it was fine. I played Metro 2033 recently but I don't think I was on 8.1 yet.
 
I also have two monitors, one which I game on and the other for temps and performance stuff.
 
So basically, it affects poorly coded games.
Using DirectInput doesn't mean the game is "poorly-coded". Microsoft has yet to deprecate the API, and using it for mouse input is perfectly within the realm of correctness. This appears to be Microsoft's issue.
 
Using DirectInput doesn't mean the game is "poorly-coded". Microsoft has yet to deprecate the API, and using it for mouse input is perfectly within the realm of correctness. This appears to be Microsoft's issue.

Correct me if I'm wrong, using raw input or directinput are not the issue. It is the games which do not use these but simply query mouse position and "translate" them to mouse movement are having issues.
 
Has anyone had this issue not in games? Like in regular desktop navigation?

I feel like since installing 8 or 8.1 (not sure as I installed them in rapid succession) my bluetooth combo keyboard/trackpad has a very laggy mouse pointer, especially - I think - during heavy CPU use.
 
Zarathustra[H];1040305531 said:
Has anyone had this issue not in games? Like in regular desktop navigation?

I feel like since installing 8 or 8.1 (not sure as I installed them in rapid succession) my bluetooth combo keyboard/trackpad has a very laggy mouse pointer, especially - I think - during heavy CPU use.

On my HTPC that is.

But that being said, it is an older 1.8Ghz Dual Core Athlon in a little mini Dell Zino HD box. Really time to build myself a new HTPC...
 
Correct me if I'm wrong, using raw input or directinput are not the issue. It is the games which do not use these but simply query mouse position and "translate" them to mouse movement are having issues.
Yeah, that seems to be the case: the article was sparse on details. I'm guessing the WINAPI functions for reading the cursor don't offer any guarantees as to their timeliness, but they should be pretty reliable. Whether there are any correctness issues in these implementations I can't say.
 
So basically, it affects poorly coded games.
Raw input code doesn't make it good code. It's like saying anyone who uses OpenGL is a bad coder, but Direct 3D makes you a good coder.

On the other hand Microsoft did some really bad coding here.
 
Zarathustra[H];1040305531 said:
Has anyone had this issue not in games? Like in regular desktop navigation?

I feel like since installing 8 or 8.1 (not sure as I installed them in rapid succession) my bluetooth combo keyboard/trackpad has a very laggy mouse pointer, especially - I think - during heavy CPU use.

I thought it was a bad mouse pad. I find myself missing click items far more often now. And sometimes I swear I click something, and nothing.
 
Raw input code doesn't make it good code. It's like saying anyone who uses OpenGL is a bad coder, but Direct 3D makes you a good coder.

On the other hand Microsoft did some really bad coding here.

I never said that. On the other hand, I am guessing it is much more efficient to use the RAW input then it is to run it through an API.

If it is indeed something that they changed or messed up with WINAPI calls, then sure, it is MS fault. On the other hand, it could be that those games affected are not using the WINAPI calls liek they should be used.

For all we know, MS fixed a bug and it messed up a few games that are using the code in a very specific way.

I have actually had that happen multiple times in some of my code, of which I will not discuss here since they were apparently security bugs that MS fixed. And no, it had nothing to do with hacking, but on non-updated machines, those bugs could very well lead to some pretty bad security issues.

And I was using the WINAPI as it was documented. I ended up having to update it either 2 or 3 times when different security patches were released. The WINAPI is pretty poorly documented from what I have seen. Even excamples from MS themselves tend to be incomplete and/or totally broken.

No, it is a behavior that broke between updates which caused the problem. It makes a lot more sense to programmers than lay people. :p

See above. It could be that MS fixed a bug and the games are not using the code properly and they need to have a simple fix done.

Guess we will have to wait and see what comes of this.
 
Joga had something to say about this in another Win8.1 thread, and has an interesting workaround. I'll quote him here:

"Mouse lag/negative acceleration in games without Raw Input - This is my biggest issue. Any game that doesn't have a "raw input" option (which is quite a lot of games) has pretty terrible mouse lag in Windows 8.1. It's apparently caused by the Multi-DPI monitor improvements in 8.1. There's a workaround by putting each game in Windows 8 compatibility mode and checking "disable scaling on high DPI settings" and running as administrator, but I really don't feel like doing that for every one of my games. I tried it out in CS:GO - enabling and disabling raw input (m_rawinput 1 or 0 respectively) and could immediately feel the difference - it's especially noticeable on higher DPI/sensitivities. In Windows 8.0, switching raw input on or off doesn't perceptibly change the sensitivity/acceleration."
 
So basically, it affects poorly coded games.

That would be basically the majority of games.

Competent programming is not highly valued at most large game companies; all that matters is that it "basically works" even if it is poorly unoptimized, unmaintainable code.
 
I have Metro 2033 & Deus Ex: Human Revolution plus a slew of other games installed on my 8.1 pro rig and osaw the issues with Dues Ex when I was running the Razer software. I ripped out the software + drivers and just let 8.1 install the mouse and haven't had any major issues ... I never saw any issues when using my MS blue tooth mouse. It wouldn't surprise me if more issues start to pop up related to hardware drivers and software programs that tie into drivers and configuration software related to the hardware, it's a new OS with a major reboot / OS upgrade tied into it... a lot of games and software are bound to have issues until the software makers and hardware companies update their software for their devices or other bla bla bla crappppp.
 
This bug really sucks because it's in wow. Not only due you get the new lag you get your character camera angle randomly spun which makes raiding and serious pvp a no go. But it sounds like it might(?) be a fixable solution in updated mouse drivers. I guess we will wait and see.
 
See above. It could be that MS fixed a bug and the games are not using the code properly and they need to have a simple fix done.

Guess we will have to wait and see what comes of this.
No, MS changed a behavior in Windows 8.1, which had been consistent since Win32 was released. See this for one of the first reports of this bug back in July: http://social.msdn.microsoft.com/Fo...s-8-pro?forum=windowsgeneraldevelopmentissues and this for an explanation of the change in behavior: http://msdn.microsoft.com/en-us/library/windows/desktop/dd464660(v=vs.85).aspx

There certainly is a poor programming practice, but it was MS who did it in this case. Completely changing the behavior of an API instead of creating a new function is level 1 stupidity.
 
fake edit: lol @ people calling most games "poorly coded" when they probably wear slip on or velcro shoes. You're not in the brain trust to evaluate who doesn't know what they're doing.
 
Don't know about the input lag because I can't even start the only game I have, BF3, after installing the update.
 
If it is indeed something that they changed or messed up with WINAPI calls, then sure, it is MS fault. On the other hand, it could be that those games affected are not using the WINAPI calls liek they should be used.
So for years games have worked fine, but 8.1 is released and there's an issue that somehow relates to bad coding by developers? Yea, cause and effect isn't something you're good at.
For all we know, MS fixed a bug and it messed up a few games that are using the code in a very specific way.
How does this make Microsoft not look like they're at fault? If there's a bug in WINAPI for years, and just now fixed it, then that means developers had to do a workaround in the code. Now that Microsoft fixed it, the workaround just bugs out and it's bad coding?

People would be surprised how often developers have to purposely make bugs in their code to get things working, cause of another bug which they can't do nothing about.
 
And this is why I wait to do such a big update.

Glad I chose to avoid until MS gets their shit together.
 
fake edit: lol @ people calling most games "poorly coded" when they probably wear slip on or velcro shoes. You're not in the brain trust to evaluate who doesn't know what they're doing.

When you have games being released this century that are still using x87 instructions instead of SSE, yeah, I'm going to call it incompetent.

When you have games that use undocumented APIs and then break when those undocumented APIs change, yeah, I'm going to call it incompetent.

When you design a modal interface with ten different modes, simply because you are too lazy to do a proper PC interface, yeah, I'm going to call it incompetent.

When you have to disable the LAA flag because you've decided to use the LSB of the memory pointer for data storage, yeah, I'm going to call it incompetent.

I've hacked enough on WINE, getting games to work, to know the kinds of things that are going on under the hood. I don't necessarily blame the programmers because it is the management that are idiots who set unrealistic deadlines and who don't understand that you can't make software by snapping your fingers.

I am not saying this mouse issue has anything to do with that. I am just replying to someone who mentioned poorly coded games. In this case, it looks like Microsoft has joined the ranks of the incompetent (or rather, increased their stature) by basically putting the programming equivalent of Word auto-complete into the Win32 APIs (and programming in Win32 is already like trying to masturbate with a brillo pad; I will take Java/Swing over Win32 any day). They have violated both the principle of least astonishment by changing the output from what it is expected and by ASSuming that it should automatically try to scale things up in High DPI mode without being asked. If a function is named GetCursorPos, it should return the cursor position, not some scaled up value that isn't reflective of the real position. Brain dead.
 
When you have games that use undocumented APIs and then break when those undocumented APIs change, yeah, I'm going to call it incompetent.
GetCursorPos() exists in both Win16 and Win32 APIs, and it's always been a documented API.

The problem is that MS broke an existing API, changing it's expected behavior "by design" (see MS employee's reply in the first link) in certain cases. The bug can, and probably will, be fixed by MS.

I'm not going to bother with more technical discussion with you because you have your mind worked up over things which don't apply. lol n00bs.
 
In the article, an Italian website was linked that explained simply starting the game in windows 7 or 8 compatibility fixed the mouse issue. Not sure if that changes the DirectX path as well...but there you go.
 
GetCursorPos() exists in both Win16 and Win32 APIs, and it's always been a documented API.

The problem is that MS broke an existing API, changing it's expected behavior "by design" (see MS employee's reply in the first link) in certain cases. The bug can, and probably will, be fixed by MS.

I'm not going to bother with more technical discussion with you because you have your mind worked up over things which don't apply. lol n00bs.

Did you read the part where I said "I am not saying this mouse issue has anything to do with that"?
 
So for years games have worked fine, but 8.1 is released and there's an issue that somehow relates to bad coding by developers? Yea, cause and effect isn't something you're good at.

How does this make Microsoft not look like they're at fault? If there's a bug in WINAPI for years, and just now fixed it, then that means developers had to do a workaround in the code. Now that Microsoft fixed it, the workaround just bugs out and it's bad coding?

People would be surprised how often developers have to purposely make bugs in their code to get things working, cause of another bug which they can't do nothing about.

Workarounds are NOT bugs.

That being said, if you have to make a workaround to get around a bug (not including specific driver issues that you have no control over), then you are doing it wrong.

The way it should be done.. if you find a bug in the code of the program you are working on, you first fix that bug before you do anything else.

I've looked at and worked on enough code to know how poorly a lot of code is written.

For that matter, I have even taken a program and basically re-written it from the ground up because it was so bad.
 
The problem is that MS broke an existing API, changing it's expected behavior "by design" (see MS employee's reply in the first link) in certain cases. The bug can, and probably will, be fixed by MS.

Side effect (support for higher than standard USB poll rate) is not an "expected behaviour". It is just a side effect of previous input/display design, and because they reworked the DPI thing, the original side effect is not present anymore.
 
I haven't noticed any issues in any of the games I've played since I upgraded to 8.1 last week. Could just be what I'm playing or my hardware.

I'm a programmer and I have to live with stuff like this all the time. Microsoft sometimes fixes bugs or introduces new ones that cause me to have to change my code to compensate. Sometimes we have to force our customers to not update their copy of Windows until we can roll out a fix. Oh fun! But I expect they'll probably fix this one pretty fast.
 
But I expect they'll probably fix this one pretty fast.

Fix what ? The fact, that the new multi-monitor multi-DPI aware code can't handle 200Hz+ mouse poll rate, because the new code for correct identification of the current mouse location is more complex and simply can't finish sooner than in 5ms (200Hz) ? Official USB poll rate is 125Hz, so MS implementation is well in the limits of what is required, it is only the high poll rate mouses which have problems.
 
They should also try and deal with the ridiculous amount of people having wifi issues. I upgraded and can't install my wifi drivers on a laptop I have. Went online to find a fix and apparently it's a big problem and has been going on for a year now.
 
Back
Top