Windows 11 24h2 has worse latency than Windows 10 22h2!

[ShrimpBrime]

Frankly I've never used it and similarly frowned at all the CS(2) ultra pro's chasing their 300 FPS pipe dream that still relies mostly on the server and the internet connection But yeah... measuring latency. I just feel latency. Much more reliable - its either responsive or too slow.

Feel micro seconds latency. Placebo.

Audio guy. Uses latencymon program, fakes fix. Replaced bad rca cable instead....

It's like fps. You can see the difference with a number. If you didn't see a number and its smooth, none the wiser. See console gaming.

Humans have infinite resolution, but all scientifically test show the brain is easily fooled. See latencymon program.

Even aida memory latency test. Like we can feel 10ns. Human reaction time is 250ms.

We are looking at (measuring) ms and us.
OK, whats the softwares poll rate set at? 1000ms? It is never a constant poll rate.

It's just a useless program everyone just installed. Fathom what a millisecond in time is.... haha.

 

[Vayra86]

Feel micro seconds latency. Placebo.

Audio guy. Uses latencymon program, fakes fix. Replaced bad rca cable instead....

It's like fps. You can see the difference with a number. If you didn't see a number and its smooth, none the wiser. See console gaming.

Humans have infinite resolution, but all scientifically test show the brain is easily fooled. See latencymon program.

Even aida memory latency test. Like we can feel 10ns. Human reaction time is 250ms.

We are looking at (measuring) ms and us.
OK, whats the softwares poll rate set at? 1000ms? It is never a constant poll rate.

It's just a useless program everyone just installed. Fathom what a millisecond in time is.... haha.

Of course there is placebo in play. But if its fine, its fine. Smooth is smooth. I've often 'double checked' my assumptions about a certain FPS number for example, and through that, I can reliably tell at what FPS I just stop caring. That's really it, speed beyond perception is useless to me.

 

[ShrimpBrime]

Of course there is placebo in play. But if its fine, its fine. Smooth is smooth. I've often 'double checked' my assumptions about a certain FPS number for example, and through that, I can reliably tell at what FPS I just stop caring. That's really it, speed beyond perception is useless to me.

A cat is 20 to 50ms reaction time. Imagine that. Being able to avoid a snake bite. (Guess I'm reaching for perspectives).

 

[Ruru]

Will try but I use my PC for gaming, so not using XMP results in lower performance, although I have the 5800x3d CPU which is not really dependent on the RAM speeds...

I know, but I would be more interesting that would that help with the latency issues?

A cat is 20 to 50ms reaction time. Imagine that. Being able to avoid a snake bite. (Guess I'm reaching for perspectives).

Cats are dopest animals ever. That's the reason why they can avoid a snakebite.

 

[grvn]

> We have a person here that just wants to tell us W11 24H2

I think is the worst Windows version to be released in years.

So many known bugs, crashes in certain games, again known facts. Alt-tabbing issues, lower performance in games etc.

Windows 11, version 24H2 known issues and notifications

View announcements and review known issues and fixes for Windows 11, version 24H2

learn.microsoft.com

Yes, Windows 10 22h2 is more stable / reliable.

 

[Bill_Bright]

If this was a useful tool for windows diagnostics and had good reports, guys like Bill_Bright would be up my ass making such claims.

I would be up your ass for being an idiot and for making claims about others when you CLEARLY have no clue, or right to even pretend you would know what they would do.

For the record, Latency mon is a useful tool. But is it a tool to CONCLUSIVELY diagnose a problem? Of course not! Just like a good multimeter is a useful tool for troubleshooting power supplies but is NOT useful to CONCLUSIVELY diagnose a faulty supply. Just like benchmark and stress programs are useful for testing cooling in extreme conditions but often are NOT very useful at determining if case cooling is adequate during real-world use.

So probably best for you, ShrimpBrime, to keep your off-tone comments about others to yourself, or risk demonstrating idiocy - again.

 

[ShrimpBrime]

I would be up your ass for being an idiot and for making claims about others when you CLEARLY have no clue, or right to even pretend you would know what they would do.

For the record, Latency mon is a useful tool. But is it a tool to CONCLUSIVELY diagnose a problem? Of course not! Just like a good multimeter is a useful tool for troubleshooting power supplies but is NOT useful to CONCLUSIVELY diagnose a faulty supply. Just like benchmark and stress programs are useful for testing cooling in extreme conditions but often are NOT very useful at determining if case cooling is adequate during real-world use.

So probably best for you, ShrimpBrime, to keep your off-tone comments about others to yourself, or risk demonstrating idiocy - again.

Latencymon is NOT a useful tool.

But you are!! Thanks Bill!!

 

[Bill_Bright]

Yeah right.

So here's an idea. Instead of just constantly whining and crying about Latencymon and criticizing others every time it is mentioned, why don't you demonstrate your superior knowledge on the subject and propose a solution? Helping others - what a concept, huh? Use your vast knowledge to demonstrate your expertise and be useful!

So instead of constantly complaining, show us how to determine if we even have latency issues, show us the best tool(s), how bad our latency issues may be, how to determine the source of the latency issues, and most importantly, how to fix them. I mean since clearly, you know all about them, share your vast knowledge with the rest of us inferior ignoramuses. Okay?

 

[ShrimpBrime]

Yeah right.

So here's an idea. Instead of just constantly whining and crying about Latencymon and criticizing others every time it is mentioned, why don't you demonstrate your superior knowledge on the subject and propose a solution? What a concept, huh? Use your vast knowledge to demonstrate your expertise and be useful!

So instead of constantly complaining, show us how to determine if we even have latency issues, show us the best tool(s), how bad our latency issues may be, how to determine the source of the latency issues, and most importantly, how to fix them. I mean since clearly, you know all about them, share your vast knowledge with the rest of us inferior ignoramuses. Okay?

So set up this on my system. The one that doesn't seem to have latency issues. With system interrupts or audio.... then show you a screen shot of some random latency nobody gives a fuck about and of all searches and forums, nobody has come to a conclusion about anything before during or after using that particular program??

Why on earth would I do that??

How about you take the time and prove me wrong. We know that's not gonna happen. Next its going to be a third post about you trying to belittle me in your own cute old guy way. Well please, if it makes you feel better about yourself, proceed!!

 

[95Viper]

Stop the insulting/bickering/belittling of members!
Stick on the topic And, that is not each other!

 

[Mr_Engineer]

I haven't tried the Latencymoon app, but just a suggestion to OP, maybe you can try disabling any Windows 11 features/services that are not being used and test again to check if latency gets lowered.

 

[ShrimpBrime]

A thing I noticed on these Forums, people become very defensive when it comes to OS discussions, sometimes I laugh and I feel like I'm talking to MS representatives, not actual users...

Has nothing to do with the OS.

All apps that alter windows settings or shows latency values and system interrupts can be viewed from windows it's self. 3rd party apps make it easier to read some and adjust some of these settings. Nothing more.

I'll use this wonderful thread as an example.

revisiting hpet bcdedit tweaks: what are your timer bench results and settings?

only for advanced users: be aware that once settings are defined by bcdedit these can cause irreversible changes and damage to your system such as low fps and stuttering until windows is reinstalled! currently im in the process of revisiting some supposed tweaks for hpet and their implications...

www.techpowerup.com

Perhaps that thread can help you? Everything you see in that thread can be done through windows without 3rd party software.

Latencymon, for reasons already mentioned is snake oil garbage.

 

[Chomiq]

My point is, I don't think it totally fair to compare W10 and W11 results when running both OSs on hardware designed for W10. Same applies if running both on W11 hardware.

What? Since when is hardware designed for specific OS?

 

[grvn]

What? Since when is hardware designed for specific OS?

It's not, they are just saying random things.

One thing I know for sure is that the Windows 11 scheduler should be in theory optimized for Intel Big-Little like arhitecture and in theory should make better use of the little cores.

But in real life, and from what I've seen on YouTube videos, Windows 10 22h2 still has on-par / better performance in gaming vs Windows 11 24h2. So even this scheduler is either busted or poorly optimized.

 

[ShrimpBrime]

It's not, they are just saying random things.

One thing I know for sure is that the Windows 11 scheduler should be in theory optimized for Intel Big-Little like arhitecture and in theory should make better use of the little cores.

But in real life, and from what I've seen on YouTube videos, Windows 10 22h2 still has on-par / better performance in gaming vs Windows 11 24h2. So even this scheduler is either busted or poorly optimized.

Why not try the testing with Tiny11 or some lite version of W11??

 

[Vayra86]

A thing I noticed on these Forums, people become very defensive when it comes to OS discussions, sometimes I laugh and I feel like I'm talking to MS representatives, not actual users...

I hope you can see thats not where Im coming from. It would serve your own use case best if you stick to commenting on what YOU experience. Generalization doesnt help.

The topic at hand is a good, but I hope you realize by now, flawed example. Ive shown you much lower latencies in the same version you struggle with. Games never hang or stutter, everything is buttery.

It might very well be true that your use case does not match well with W11 and does match better with W10. And it leads to just one logical outcome; you will run W10. You might even be very well served with an LTSC build of it. Can recommend that anyway. New OS features arent going to be helpful to latencies.

 

[grvn]

Why not try the testing with Tiny11 or some lite version of W11??

Thanks I will try that. What I'm afraid is that they "cut" so many things that are tied to hard with Windows that it might break general usage?

 

[Bill_Bright]

What? Since when is hardware designed for specific OS?

Please. Think about this for a second.

The answer is, since day one.

Note I did NOT say that the HW was designed exclusively for one specific OS only. That said, I probably should have said "HW is designed to run "optimally" with a specific OS". My bad and my apologies for not stating it more clearly.

It's not, they are just saying random things.

Really?

Then why doesn't XP run on today's hardware right out of the box? Or W11 run on XP era hardware right out of the box? Why, is there even a W11 compatibility checker to see if our W7 and W10 era hardware supports W11? Why were millions and millions of W10 users complaining they could not upgrade to W11 because the CPU or motherboard would not support it?

Why won't W11 run with many legacy CPUs or motherboards?

Why did entire industries complain about the massive expenditures they had to make in hardware when upgrading from CP/M to DOS? Then again from DOS to Windows? Then again from W95/98 to XP. Why did big corporates force Microsoft to code in legacy HW support in XP? It wasn't just for running their legacy proprietary software!

Why are there emulation programs if you can just run any OS on any hardware?

Why can't we run AndroidOS on Apple iPhones, or vice versa?

***

Getting back to LatencyMon, IMO, one of its problems is its name. Those less familiar with the program see the name and think the program is intended to monitor ALL potential latency issues. They then tend to use it incorrectly, don't understand or misunderstand what it does, then incorrectly think it is some sort of "quack remedy" for something its not. That is NOT true.

If you look at the logo for LatencyMon, it clearly says it is a "Real-time audio stability checker". Real-time "audio stability" checker hardly suggests latency issues to most users familiar with the traditional definition of the word, latency. To most, latency refers to "time", as in the time it takes for data to get from point A to point B. Some of us grew up with Ping - a tool (command) to check if a distant end is out there, and the "time" it takes for data to travel between here and there and back again (latency).

So how "audio stability" fits in that classic definition of "latency" is confusing at best, and misleading too. We've seen some users with fiber optic bandwidth internet connections panic, thinking their ISP and network are having major issues, when really it has nothing to do with their network. Some panic even when they are experiencing no audio issues at all!

Resplendence has a fairly decent tutorial for LatencyMon here. Another is this one: Using LatencyMon – A Quick Step-By-Step Guide. Note the Preparation section.

And BTW, a simple check shows LatencyMon was last updated just 18 months ago in July 2023. Not "for at least 3 years" as previously claimed.

***

Another thing you will notice over here is people blowing things out of proportions based on what is essentially a sample size of 1. Happens a lot.

Agreed. And not just "out of proportion" but they totally believe their anecdotal sample size of one renders an entire point moot - as though their exception makes the rule when the vast majority (100s of millions of users) have not experienced that problem at all.

 

[ShrimpBrime]

Thanks I will try that. What I'm afraid is that they "cut" so many things that are tied to hard with Windows that it might break general usage?

It depends on who did the lite version what they cut out. Generally it's stuff like MSPaint 3D for example. Others very heavily stripped. But most copies should explain what's been removed if reputable at all.

If it says you have high latency or that your PC is not capable of rendering smooth audio, it's probably false reading.

Remember this tool hasn't been updated over 3 years and not likely to be accurate for W11, let alone W10.

And if its hardware bound as suggested by the electronics engineer, it would be specifically stated on their very ad bloated web site with very short and vague description that hasn't changed in well over a decade and a half now. As far as I'm aware anyways.

 

[Bill_Bright]

Remember this tool hasn't been updated over 3 years

Again, not true. As just indicated above, it has been updated a couple times since then with the last update in July 2023. Please stop spreading misinformation.

But frankly, so what if it hasn't been updated? Nothing says a program has to be updated every X number of months to stay viable. Plus it's a bit hypocritical to suggest LatencyMon is no good because it has not been updated recently, then suggest the use of TimerBench 1.5 that was last updated 2/25/2020! In fact, that is just the exe timestamp. A quick look at the TimerBench 1.5 About screen and note it clearly states,

Coded by Matthias "mat" Zronek, overclockers.at, 2018-2019

 

[ShrimpBrime]

Again, not true. As just indicated above, it has been updated a couple times since then with the last update in July 2023. Please stop spreading misinformation.

But frankly, so what if it hasn't been updated? Nothing says a program has to be updated every X number of months to stay viable. Plus it's a bit hypocritical to suggest LatencyMon is no good because it has not been updated recently, then suggest the use of TimerBench 1.5 that was last updated 2/25/2020! In fact, that is just the exe timestamp. A quick look at the TimerBench 1.5 About screen and note it clearly states,

Like I said Bill.

You go to that site and DL it. I'm not going to. I've used it before. Said I had a bunch of audio latency. There was none. No indication I had audio problems. Program says there's a lot of interrupts. Well yeah, thousands every second. As far as I'm aware, system interrupts are pretty normal operation of the operating system.

Again, as I stated, you don't need this program to view system interrupts. You know this better than anyone I would think.

 

[Bill_Bright]

That has absolutely nothing to do with what you quoted but, okay. Moving on.

 

[Vayra86]

Then why doesn't XP run on today's hardware right out of the box? Or W11 run on XP era hardware right out of the box? Why, is there even a W11 compatibility checker to see if our W7 and W10 era hardware supports W11? Why were millions and millions of W10 users complaining they could not upgrade to W11 because the CPU or motherboard would not support it?

Why won't W11 run with many legacy CPUs or motherboards?

Why did entire industries complain about the massive expenditures they had to make in hardware when upgrading from CP/M to DOS? Then again from DOS to Windows? Then again from W95/98 to XP. Why did big corporates force Microsoft to code in legacy HW support in XP? It wasn't just for running their legacy proprietary software!

Why are there emulation programs if you can just run any OS on any hardware?

Why can't we run AndroidOS on Apple iPhones, or vice versa?

***

This can all be answered with a very simple one letter word.

Commerce.

These are not technical issues. They are artificial issues to create unique selling points and sell product. New OS versions these days are most definitely not created for ANY technical necessity, or new hardware reality, they are created to present you a new EULA so they can get a bit more data than last time. MS is practicing that M.O. as we speak. The TPM 2.0 requirement is the latest most obnoxious one I can mention that is supposedly 'hardware related' ... except its not. Its just a barrier to sell more hardware to people who can't circumvent said barrier.

Now of course, I know you'll be thinking and probably already typing 'WRONG because 32 bit and 64 bit and and and.' YES, we are aware, there ARE versions that were necessary from a technical point of view. But even then, we would probably need like 3 or 4 OS versions, not 11+, and certainly not a single new version since Windows 7.

 

[Bill_Bright]

Ummm, not really. That is only partially correct. Sure, money is always a factor but it is not in the way you imply.

It is important to note that contrary to what many seem to think, advances in hardware technologies are NOT controlled by Microsoft. NVIDIA, Intel, AMD and others have their own paths that are not dictated by Microsoft. Also, and sadly, the bad guys have significant influences here too by forcing MS and HW makers to implement security features - features not supported in legacy systems.

It is easy to blame Microsoft for everything people don't like. And that is unfortunate because much is partially and often totally unwarranted. Remember, except for some games, almost every computing task done with Windows can be done with Linux. Microsoft knows this well.

Also worth understanding and remembering is that Windows actually makes up just a small part of Microsoft's business. So too often Microsoft is given (accused of) too much credit for what is happening these days. The truth is, only about 12% of Microsoft's revenue comes from Windows.

Remember too, Microsoft gave away 100s of millions W10 and W11 licenses - for free!

Yeah, TPM 2.0 was a hassle but don't forget, Thunderbird email clients use TPM to handle encrypted or key-signed messages. The Firefox and Chrome web browsers also employ the TPM for certain advanced functions, such as maintaining SSL certificates for websites. Plenty of consumer tech besides PCs uses TPMs, as well, from printers to connected-home accessories.

Yeah, MS requires it but can you really blame them? The security mess we are in was caused by bad guys, not Microsoft. And the bad guys were allowed to proliferate because Norton, McAfee and the others failed to stop them. And computers got infected because users failed to keep their computers updated and they clicked on every unsolicited link they found.

But who got blamed? Did Norton and McAfee and the others get blamed even though they whined and cried to Congress and the EU that it was their job to block malware? Did the users (and IT/security people) get blamed for letting the bad guys in? Did the bad guys get blamed? No, no and NO!! Microsoft got blamed.

Microsoft learned quickly they were damned if they do and damned if they don't. So Microsoft decided they would rather get blamed for enforcing restrictive security measures that help keep users safe than get blamed for the actions of the bad guys, the failures of the security software industry to do their jobs, and the lack of safe computing practices by users.

And while I don't like some of those restrictive measures either, I have to applaud MS for that decision. The proof is in the statistics. By far, the greatest security threat today is through hacks and breaches at the corporate/organizational levels, not our home computers. And I note in most of those breaches, it was due to the failure of those network administrators to do their jobs properly, and the failure of C-Level execs to be held accountable.

 

[Vayra86]

Ummm, not really. That is only partially correct. Sure, money is always a factor but it is not in the way you imply.

It is important to note that contrary to what many seem to think, advances in hardware technologies are NOT controlled by Microsoft. NVIDIA, Intel, AMD and others have their own paths that are not dictated by Microsoft. Also, and sadly, the bad guys have significant influences here too by forcing MS and HW makers to implement security features - features not supported in legacy systems.

It is easy to blame Microsoft for everything people don't like. And that is unfortunate because much is partially and often totally unwarranted. Remember, except for some games, almost every computing task done with Windows can be done with Linux. Microsoft knows this well.

Also worth understanding and remembering is that Windows actually makes up just a small part of Microsoft's business. So too often Microsoft is given (accused of) too much credit for what is happening these days. The truth is, only about 12% of Microsoft's revenue comes from Windows.

Remember too, Microsoft gave away 100s of millions W10 and W11 licenses - for free!

Yeah, TPM 2.0 was a hassle but don't forget, Thunderbird email clients use TPM to handle encrypted or key-signed messages. The Firefox and Chrome web browsers also employ the TPM for certain advanced functions, such as maintaining SSL certificates for websites. Plenty of consumer tech besides PCs uses TPMs, as well, from printers to connected-home accessories.

Yeah, MS requires it but can you really blame them? The security mess we are in was caused by bad guys, not Microsoft. And the bad guys were allowed to proliferate because Norton, McAfee and the others failed to stop them. And computers got infected because users failed to keep their computers updated and they clicked on every unsolicited link they found.

But who got blamed? Did Norton and McAfee and the others get blamed even though they whined and cried to Congress and the EU that it was their job to block malware? Did the users (and IT/security people) get blamed for letting the bad guys in? Did the bad guys get blamed? No, no and NO!! Microsoft got blamed.

Microsoft learned quickly they were damned if they do and damned if they don't. So Microsoft decided they would rather get blamed for enforcing restrictive security measures that help keep users safe than get blamed for the actions of the bad guys, the failures of the security software industry to do their jobs, and the lack of safe computing practices by users.

And while I don't like some of those restrictive measures either, I have to applaud MS for that decision. The proof is in the statistics. By far, the greatest security threat today is through hacks and breaches at the corporate/organizational levels, not our home computers. And I note in most of those breaches, it was due to the failure of those network administrators to do their jobs properly, and the failure of C-Level execs to be held accountable.

Like most things its a double edged sword, it is partially correct, I agree. But even so, Microsoft could have easily not released numerous versions of this OS and instead made the fewer ones that do need to exist, better. Surely you're not in disagreement on that one... heck, even Microsoft was telling us that with Windows 10. 'This is the one!'

Ever since Windows 7 the main driver for a new version was commercial of nature. With 8 and the Metro UI and now 11 with its AI bullshit as the most glaring examples. They could have easily added TPM in a module available only to new systems, for example, or entirely optional.