Unstable Pc performance and lack of visual smoothness.

[Inputdelay]

Silly question: is there a ham-radio operator nearby? Two options possible:
He himself disturbs with high power DX transmission.
He can assist you by scanning most of all bands from 100kcy to above 2GHz

Either I don't know about them, or they don't exist, because I had a radio frequency center, but they didn't stay for long, about 2-3 hours. And they should have noticed this when they were looking for the problem.

[Analog Kid]

What's a "radio frequency center"?

[Inputdelay]

What's a "radio frequency center"?

https://www.ucrf.gov.ua/en

And I had an radio-monitoring inspection at my place. And that's the response they sent me:

"Based on the results of the measures carried out by employees of the Central Branch of the UCRF in your presence on the premises and in the rooms of the building, in the public radio frequency bands, including the frequency bands for cellular radio communications, terrestrial television broadcasting, terrestrial analog audio broadcasting, and broadband wireless radio access, there were no violations of the radio emission parameters of the specified radio technologies and unacceptable radio interference in the public radio frequency bands.
It was recorded that the PC system unit itself emits broadband radiation in the approximate range of 308–730 MHz with levels up to ~-60 dBm.
In addition, we would like to inform you that the Central Branch of the Ukrainian State Centre for Radio Frequencies has not received any requests from operators providing telecommunications services in the village of Yours or from other residents of the specified area to identify and eliminate sources of radio interference that may cause reception problems in public radio frequency bands."

[johansen]

That particular freeze frame was made on a Xiaomi Mi 8 with custom software and custom camera software with manual settings, but now I have an iPhone XS.

this is starting to get even weirder.

what have you done to verify your software works and that phone is able to accurately record substantially higher than 360fps?

the 240 pixel wide image you posted appears to be the same image shifted once in the background and the character walking i presume, is shifted twice.

that alone is interesting but there may be different frame rates for the game characters vs the background.

[Inputdelay]

That particular freeze frame was made on a Xiaomi Mi 8 with custom software and custom camera software with manual settings, but now I have an iPhone XS.

this is starting to get even weirder.

what have you done to verify your software works and that phone is able to accurately record substantially higher than 360fps?

the 240 pixel wide image you posted appears to be the same image shifted once in the background and the character walking i presume, is shifted twice.

that alone is interesting but there may be different frame rates for the game characters vs the background.

First, the character did not move, meaning he was static like the background. I moved the mouse from left to right, thereby moving the in-game camera.

Second, I filmed a UFO test for frame skipping on a verified 360 Hz monitor. https://testufo.com/frameskipping

If the camera was shooting with frame skipping or interpolation, it would be visible in the test.

The camera module and processor datasheets confirm the possibility of high-speed shooting.

[Buriedcode]

You know that just because a monitor has a refresh rate of 144Hz doesn't mean to say the source updates the frame 144 times a second? Most online videos are 30fps, and using a phone camera to capture high-speed events doesn't really show anything, since it will capture a fraction of a frame update.

[Inputdelay]

You know that just because a monitor has a refresh rate of 144Hz doesn't mean to say the source updates the frame 144 times a second? Most online videos are 30fps, and using a phone camera to capture high-speed events doesn't really show anything, since it will capture a fraction of a frame update.

Firstly, this freeze frame was taken from a video on a phone that can record in this range.

Secondly, at that moment the freeze frame was taken, in the game the FPS was over 500.

Thirdly, the screen was recorded at 360 Hz, not 144 Hz.

And most importantly, it's not even about the photo itself, but the fact that no matter how many times I make similar recordings and no matter what device I use, the result will be the same-the image is blurred and doubled, even though it shouldn't be.

And it changes constantly; at some point, it may become brighter and clearer with less blurring, but this state is not stable and is constantly changing.

[Buriedcode]

It is going to be very difficult to obtain objective evidence of this - since you have to record what a monitor is displaying, whilst also taking into account its refresh rate as well as the recording devices frame rate, frame dwell time, and on a phone it is more difficult since they perform compression and filtering on the fly.  I realise you are trying to get a clear recording of the problem, because, well, we're engineers - we need to measure

Also, how do you know the FPS of the game you're playing? Often the FPS indicators in-game are wrong, and besides, your system has to drop a lot of frames because you can't display that frame rate.  I have had stuttering/freezing in a game whilst NVIDIA insists the FPS is >180 when it is clearly below 30 (got an old GX card).  Anything over 60 is a waste anyway.

Since you say this occurs on almost everything it displays, dont' bother with games - instead, download a video, with a fixed frame/bit rate - verified by checking the metadata - and play it in VLC.  Record with your phone at a reasonable framerate - 60fps 1080p is perfectly adequate to spot the issues you claim you're seeing, if you can tell the difference between 60fps and 120fps then you're better than most humans at spotting it.  Also, for the recording, mount the phone/camera at a fixed point , at a fixed distancxe from the screen, and try to keep the lighting the same.  Even then it won't be great since uploading it will compress it and youtube will automatically downscale it to 30fps, and compress it to oblivion.

I think everyone in this thread is now curious as to what exactly the problem is - you've provided lots of details and it seems to vague and that you say it affects so many different systems/displays makes it very very odd.

[Inputdelay]

It is going to be very difficult to obtain objective evidence of this - since you have to record what a monitor is displaying, whilst also taking into account its refresh rate as well as the recording devices frame rate, frame dwell time, and on a phone it is more difficult since they perform compression and filtering on the fly.  I realise you are trying to get a clear recording of the problem, because, well, we're engineers - we need to measure

Also, how do you know the FPS of the game you're playing? Often the FPS indicators in-game are wrong, and besides, your system has to drop a lot of frames because you can't display that frame rate.  I have had stuttering/freezing in a game whilst NVIDIA insists the FPS is >180 when it is clearly below 30 (got an old GX card).  Anything over 60 is a waste anyway.

Since you say this occurs on almost everything it displays, dont' bother with games - instead, download a video, with a fixed frame/bit rate - verified by checking the metadata - and play it in VLC.  Record with your phone at a reasonable framerate - 60fps 1080p is perfectly adequate to spot the issues you claim you're seeing, if you can tell the difference between 60fps and 120fps then you're better than most humans at spotting it.  Also, for the recording, mount the phone/camera at a fixed point , at a fixed distancxe from the screen, and try to keep the lighting the same.  Even then it won't be great since uploading it will compress it and youtube will automatically downscale it to 30fps, and compress it to oblivion.

I think everyone in this thread is now curious as to what exactly the problem is - you've provided lots of details and it seems to vague and that you say it affects so many different systems/displays makes it very very odd.

In order to find out the exact FPS value in games, I used all available types of measuring instruments, and the margin of error was no more than 5-7% (MSI Afterburner + RTSS, in-game FPS, in-game indicators, Intel Preset Mon, CapFrameX, Steam In-Game Overlay, FRAPS, and FPS Monitor). We also tracked not only the average FPS but also 0.1% and 1% FPS, as well as framerate. This was to ensure that there were no anomalies, and none were found.

This problem was observed on all graphics cards, all versions of NVIDIA from the GTX 1000 series to the RTX 5000 series inclusive (video cards from AMD and Intel vendors have exactly the same problems). I have a 360 Hz monitor, and I can clearly see the difference between 60, 144, 240, 300, and 360 Hz.

As for video, I have already done similar tests, and there were no differences from any video sequence, whether it was a video, a game, or even File Explorer on the desktop. I want to clarify that the problem is complex:

1) Reduction in overall image clarity.

2) Ghosting: the remaining trail of moving objects.

3) Image splitting during sudden mouse movements.

And they have some common root cause, because when the computer is working “normally,” the improvement occurs in all areas at once.

As for VLC, it is very buggy. I also noticed that media players smooth out the image, unlike the actual application that is running. This is because the frames have the same interval, and vertical synchronization is enabled by default in media players.

[dervu]

I experienced same symptoms for around 13 years in my apartment block.
I can confirm if you exclude all other possibilities, then some kind of interference from power grid is causing it. The thing is that it must be on some fundamental level, so that you will not see errors in most of widely available software benchamrks and tests, maybe except some ultra detailed comparison of video output in place where this issue is present vs place where it's not present or some kind of PCI-E analyzers that costs thousands to see packet timing.
It is clearly something that affects system as a whole, not something that would disrupt just one thing in the system, like memory showing errors memtest on unstable overclock.

Confirmed multiple times going back and forth between two locations using laptop and PC, during one year. Also gone through multiple laptops and PCs through those 13 years trying to figure out what it is and ruled out hardware completely.
Most obvious symptom I discovered only last year is that timer measurements inside CPU-Z were glitching in a way it would do like microstops every x seconds that you could visually notice. It was at affected place.
While in unaffected place it would go flawlessly smooth like butter. No matter version of software.

I was going through regulatory institution to help me with that. I had evidence of AM interference affected regulated radio program. It was hard to convince old guys that hardware does not work properly. Second approach with AM radio interference also failed, as it was also available on FM, so it was non issue for them.
For couple of years I was suspicious about elevators in my apartment block. It's 1990s building, old elevators,  TN-C electrical system. Then around the time those symptoms appeared it was at time they mounted VFDs to those elevators, while keeping old motors.
At first couple of years I wouldn't connect the dots, thinking something must have been wrong with my laptop, being too old or heating too much, so my journey through all tiers of hardware began.
Then sometime year ago housing association mentioned tender for replacing all elevators as they approached end of service life.
This year elevators were replaced to completely new ones, according to all new rules.

All symptoms are gone.

Now tell me how can a regular person prove that to anyone? Most people wouldn't event touch 1% of what I did to try to figure out what it is. From trying to measure it, to trying all possible filtering there is, to just being lucky in the end and thousands of euros less in wallet.

One thing is for sure. That not everyone might have the same thing, as people like to think they have what others do. You have to consider:
- if everything was ok before that
- when it started
- what has changed when it started (closest environment)
- rule out all software and hardware problems

Of course going with same hardware to other place and having no symptoms at all for prolonged period of time makes you suspicious, so it's also good approach if you have strong cues it can be such issues and save some time when you know your things with hardware good enough.

What doesn't help is lack of common way to measure it that in affordable way.
No one will want to spend thousands on hardware for measurement hardware, second apartment and get month of vacation just to test if monitor output is different x% on average on exact same automated scenario in same perfect lighting between two places.

It might be just like this thing with animation error posted above. People can tell, but there is not enough ways to measure some things yet, and if they are there, they are expensive.

For now it's a struggle for anyone who experiences it, as it's most probably something out of your reach to fix it on your own, unless you get lucky with source being too old and about to be replaced or it has other symptoms affecting critical functions, or some EMC people who will go above and beyond to find it (probably again about more money for their time).
I bet on supraharmonics, but again to prove it, it's another stack of money.

I'll also add that on my side trying out such things did not solve the issue:
- Medical grade isolation transformer
- Online UPS
- FN700Z filter

Only those stacked up made some singificant improvement but it was really unpleasant way to playing in a room that looked like hardware lab and noisy UPS, so it wasn't tested long enough.

[Inputdelay]

It is going to be very difficult to obtain objective evidence of this - since you have to record what a monitor is displaying, whilst also taking into account its refresh rate as well as the recording devices frame rate, frame dwell time, and on a phone it is more difficult since they perform compression and filtering on the fly.  I realise you are trying to get a clear recording of the problem, because, well, we're engineers - we need to measure

Also, how do you know the FPS of the game you're playing? Often the FPS indicators in-game are wrong, and besides, your system has to drop a lot of frames because you can't display that frame rate.  I have had stuttering/freezing in a game whilst NVIDIA insists the FPS is >180 when it is clearly below 30 (got an old GX card).  Anything over 60 is a waste anyway.

Since you say this occurs on almost everything it displays, dont' bother with games - instead, download a video, with a fixed frame/bit rate - verified by checking the metadata - and play it in VLC.  Record with your phone at a reasonable framerate - 60fps 1080p is perfectly adequate to spot the issues you claim you're seeing, if you can tell the difference between 60fps and 120fps then you're better than most humans at spotting it.  Also, for the recording, mount the phone/camera at a fixed point , at a fixed distancxe from the screen, and try to keep the lighting the same.  Even then it won't be great since uploading it will compress it and youtube will automatically downscale it to 30fps, and compress it to oblivion.

I think everyone in this thread is now curious as to what exactly the problem is - you've provided lots of details and it seems to vague and that you say it affects so many different systems/displays makes it very very odd.

https://youtu.be/KGq_4cLWrfQ

Here is a video example showing the presence and absence of the problem with absolutely identical system and software settings. Same settings of overdrive.

The only difference in the video is the location. If I take all the same devices and bring them to my friend's house, there would be no such problem.

The camera exposure was identical in both cases and was set manually. The only difference in the video is the color temperature, but it does not affect the problem.

The ghosting problem is visible to the naked eye. You can pause the video at any time while watching it and see that the blurring occurs constantly.

Also adding a collage of frames from the video to make it clearer.

[thm_w]

We had this thread before, may even be the same user: https://www.eevblog.com/forum/beginners/pc-performance-correlated-with-relative-humidity/

Basically nothing ever gets proven, because the problem they describe likely never existed.
Thats not to say there aren't 1000 possible annoyances that could exist in one room and not another (sound pollution, lights flickering, poor air quality, etc.).

Now tell me how can a regular person prove that to anyone? Most people wouldn't event touch 1% of what I did to try to figure out what it is. From trying to measure it, to trying all possible filtering there is, to just being lucky in the end and thousands of euros less in wallet.

You prove it with measurements. Any sort of interference within the processor/OS/PC itself should be measurable with software without too much difficulty, probably even the GPU as well. You don't need a PCIE analyzer for that.

[SteveThackery]

There is a problem with your video.

This video is emphatically NOT at a high frame rate! I can detect judder on the faster moving objects in the second half of the video (the part with a blue cast). Look closely at the balls at the point where they bounce (so moving fastest) - the judder is as plain as day. It looks like 60fps to me, or just possibly 120Hz.  Maybe you recorded it at some super high frame rate, but it isn't now. Perhaps YouTube has messed with it?

So we have no judder on the first sample, but some ghosting; and we have no ghosting but obvious judder on the second sample. I simply don't believe you have "absolutely identical system and software settings" for both samples. It can't be - that judder proves that the refresh rate, or frame rate, on the second sample is much slower.

If you don't believe me open up the video in a video editing suite and examine it frame by frame. If I get time I'll do that tomorrow or the day after and report back.

[SteveThackery]

We had this thread before, may even be the same user: https://www.eevblog.com/forum/beginners/pc-performance-correlated-with-relative-humidity/

Basically nothing ever gets proven, because the problem they describe likely never existed.

Ah, that's interesting. Anyway, a frame-by-frame inspection might be informative.

[thm_w]

Ah, that's interesting. Anyway, a frame-by-frame inspection might be informative.

Sure, a good modern smart phone might do 240fps which can probably give a lot of information.

The video posted seems a bit off as you say as well, the first segment is running at a very stable 720fps and language is english.
The second segment FPS is a lot lower (fps in top right?), and is also jumping all over the place, and language is no longer in english. Doesn't look like the same setup to me.

[Inputdelay]

We had this thread before, may even be the same user: https://www.eevblog.com/forum/beginners/pc-performance-correlated-with-relative-humidity/

Basically nothing ever gets proven, because the problem they describe likely never existed.
Thats not to say there aren't 1000 possible annoyances that could exist in one room and not another (sound pollution, lights flickering, poor air quality, etc.).

Now tell me how can a regular person prove that to anyone? Most people wouldn't event touch 1% of what I did to try to figure out what it is. From trying to measure it, to trying all possible filtering there is, to just being lucky in the end and thousands of euros less in wallet.

You prove it with measurements. Any sort of interference within the processor/OS/PC itself should be measurable with software without too much difficulty, probably even the GPU as well. You don't need a PCIE analyzer for that.

The only difference in these videos is the language; the person prefers the other-language FPS counter and a different layout, but this is unrelated to the other details. Yes, YouTube compressed the video to 60 fps.

As I already wrote, I see exactly the same thing with my eyes as in the video, no matter what camera, phone, or any other video recording device I use, but I get the same result. At home, the blurring of any moving objects is much more pronounced. And in my friend's video, there is only a small halo that moves behind the ball, and it is visible as a blue shadow.

Based on this, I don't see the point in analyzing this video at all, except as proof that there is a problem.

Therefore, I suggest focusing on the source of the problem and methods of finding it, because it is 10000% not the computer itself, its operating system, or the peripherals. This blurring will be equally visible at 60 fps 1080p and 1000 fps 8k. I say this because even my friend who owns a professional camera produced the same result.

[Inputdelay]

We had this thread before, may even be the same user: https://www.eevblog.com/forum/beginners/pc-performance-correlated-with-relative-humidity/

Basically nothing ever gets proven, because the problem they describe likely never existed.
Thats not to say there aren't 1000 possible annoyances that could exist in one room and not another (sound pollution, lights flickering, poor air quality, etc.).

You prove it with measurements. Any sort of interference within the processor/OS/PC itself should be measurable with software without too much difficulty, probably even the GPU as well. You don't need a PCIE analyzer for that.

It has nothing to do with sound pollution or poor air quality.
Unstable electricity, electromagnetic compatibility, and super harmonics-maybe. But it's definitely somewhere external. It's just impossible to have the same problems across the entire world, with the exact same symptoms.

I had a radio-monitoring inspection at my place with a spectrum analyzer. What kind of interference do I have to show as an example if there wasn't anything suspicious?

I'm still looking for any suggestions of how to find and fix the problem.

[dervu]

We had this thread before, may even be the same user: https://www.eevblog.com/forum/beginners/pc-performance-correlated-with-relative-humidity/

Basically nothing ever gets proven, because the problem they describe likely never existed.
Thats not to say there aren't 1000 possible annoyances that could exist in one room and not another (sound pollution, lights flickering, poor air quality, etc.).

Now tell me how can a regular person prove that to anyone? Most people wouldn't event touch 1% of what I did to try to figure out what it is. From trying to measure it, to trying all possible filtering there is, to just being lucky in the end and thousands of euros less in wallet.

You prove it with measurements. Any sort of interference within the processor/OS/PC itself should be measurable with software without too much difficulty, probably even the GPU as well. You don't need a PCIE analyzer for that.

Well, I did at least with one simple test as I mentioned in my previous post. I did not help me without proceeding to correlate that with some interference, as I would probably need to turn off that specific source of interference, but no one would allow me to do that. Otherwise it would be boring searching through differences between two locations.
While all other problems are quite easy to fix compared to this.

[thm_w]

The only difference in these videos is the language; the person prefers the other-language FPS counter and a different layout, but this is unrelated to the other details. Yes, YouTube compressed the video to 60 fps.

So why is the FPS significantly different (720fps vs 610fps), and why is it unstable in one case and stable in the other? Are these videos even from the same hardware?

Unstable electricity, electromagnetic compatibility, and super harmonics-maybe. But it's definitely somewhere external.

Then its effects on the computer would be easy to measure: data corruption, fps instability, performance decrease, etc.
Unless its an effect or setting on the monitor itself, thats harder to nail down, but you've said you already ruled it out.

[Inputdelay]

We had this thread before, may even be the same user: https://www.eevblog.com/forum/beginners/pc-performance-correlated-with-relative-humidity/

Basically nothing ever gets proven, because the problem they describe likely never existed.
Thats not to say there aren't 1000 possible annoyances that could exist in one room and not another (sound pollution, lights flickering, poor air quality, etc.).

Now tell me how can a regular person prove that to anyone? Most people wouldn't event touch 1% of what I did to try to figure out what it is. From trying to measure it, to trying all possible filtering there is, to just being lucky in the end and thousands of euros less in wallet.

You prove it with measurements. Any sort of interference within the processor/OS/PC itself should be measurable with software without too much difficulty, probably even the GPU as well. You don't need a PCIE analyzer for that.

Well, I did at least with one simple test as I mentioned in my previous post. I did not help me without proceeding to correlate that with some interference, as I would probably need to turn off that specific source of interference, but no one would allow me to do that. Otherwise it would be boring searching through differences between two locations.
While all other problems are quite easy to fix compared to this.

Lucky for me, I have done that test. In my area, I don't have electricity for many, many hours, and the situation stays the same. Even using portable power stations doesn't help the situation. Unless this specific source of interference is somewhere far or the electricity company doesn't turn it off for some reason.

[thm_w]

Well, I did at least with one simple test as I mentioned in my previous post. I did not help me without proceeding to correlate that with some interference, as I would probably need to turn off that specific source of interference, but no one would allow me to do that. Otherwise it would be boring searching through differences between two locations.
While all other problems are quite easy to fix compared to this.

What simple test was that? If OP runs the same test and its the same between locations, that would rule out the supposed interference issue you saw.
Seems it would be valuable, just run the test and start turning off circuits one by one.

Lucky for me, I have done that test. In my area, I don't have electricity for many, many hours, and the situation stays the same. Even using portable power stations doesn't help the situation. Unless this specific source of interference is somewhere far or the electricity company doesn't turn it off for some reason.

Then it has nothing to do with interference, and you've missed something basic in your analysis.

[Inputdelay]

Well, I did at least with one simple test as I mentioned in my previous post. I did not help me without proceeding to correlate that with some interference, as I would probably need to turn off that specific source of interference, but no one would allow me to do that. Otherwise it would be boring searching through differences between two locations.
While all other problems are quite easy to fix compared to this.

What simple test was that? If OP runs the same test and its the same between locations, that would rule out the supposed interference issue you saw.
Seems it would be valuable, just run the test and start turning off circuits one by one.

Lucky for me, I have done that test. In my area, I don't have electricity for many, many hours, and the situation stays the same. Even using portable power stations doesn't help the situation. Unless this specific source of interference is somewhere far or the electricity company doesn't turn it off for some reason.

Then it has nothing to do with interference, and you've missed something basic in your analysis.

The tests required turning off the entire half of the city from electricity. But due to the war, we have quite frequent power outages, more than 3 times per day for many hours. which means most of the interference sources were off. But as I said, it's unless this specific source of interference is somewhere far, or the electricity company doesn't turn it off for some reason, or if it's working even during power outages.

[Manul]

The thing is, such effects due to external interference are almost impossible, especially if consistent and not just a glitch in one frame out of a million. Any memory corruption events would manifest most notably as system crashes and are pretty random. Glitching digital video transmission (like hdmi link) would also manifest quite differently, by select variety of video defects or a total loss of video. I suggest to not go into "moon phases affect my frame rates" type of theories, it will be just a waste of time and a distraction from real issues.

[Inputdelay]

The thing is, such effects due to external interference are almost impossible, especially if consistent and not just a glitch in one frame out of a million. Any memory corruption events would manifest most notably as system crashes and are pretty random. Glitching digital video transmission (like hdmi link) would also manifest quite differently, by select variety of video defects or a total loss of video. I suggest to not go into "moon phases affect my frame rates" type of theories, it will be just a waste of time and a distraction from real issues.

it might be just an EMC, not the "moon phases affect my frame rates" type of theories.

[SteveThackery]

The thing is, such effects due to external interference are almost impossible, especially if consistent and not just a glitch in one frame out of a million. Any memory corruption events would manifest most notably as system crashes and are pretty random. Glitching digital video transmission (like hdmi link) would also manifest quite differently, by select variety of video defects or a total loss of video. I suggest to not go into "moon phases affect my frame rates" type of theories, it will be just a waste of time and a distraction from real issues.

This is exactly what I am thinking. For something in the environment to cause this specific problem is inconceivable. It just is. I do not accept that we are looking at exactly the same hardware and software - just in two different places.

Inputdelay: how do you explain the obvious judder in the second half of the video?