PC mouse refresh rate

Atari talk, or the life and the universe and things. Just keep it clean!
User avatar
sandord
Trusted Guru
Trusted Guru
Posts: 470
Joined: Mon Aug 13, 2018 10:08 pm
Location: The Netherlands
Contact:

Re: PC mouse refresh rate

Post by sandord » Mon Nov 05, 2018 12:14 pm

sandord wrote:
Mon Nov 05, 2018 12:07 pm
This image (from the article linked above) basically sums it all up:


dn642112.ic458338(en-us,vs.85).gif
The most fundamental difference between CRT and LCD is that the latter (usually) doesn't redraw the complete image every frame. LCDs only update the difference. That's what sample-and-hold is about. Some LCDs have a strobed backlight (and some OLEDs have black frame insertion) which effectively does cause each frame to be completely redrawn. That is what the eye needs to see or otherwise it thinks that only the edges of the object have moved (https://en.wikipedia.org/wiki/Persistence_of_vision).
Steve wrote:
Mon Nov 05, 2018 12:07 pm
Yes it is the LCD 60hz that is causing your issue, even though 60fps might sound smooth your desktop most likely doesn't have v-sync enabled which causes a jittery mouse movement (which is the same for 99.9% of 'standard' pc configurations)
Yeah, that's also probably a big factor, though while making circular movements with my PC mouse, I don't see any judder. The number of simultaenously perceptive pointers (obviously, only one is displayed at the time but the eye tends to merge the pictures) is very stable (it's about 15).

Steve
Posts: 177
Joined: Fri Sep 15, 2017 11:49 am

Re: PC mouse refresh rate

Post by Steve » Mon Nov 05, 2018 2:25 pm

Believe me, 60hz is not suitable for FPS games. You need a 120/144hz capable LCD. Forget the BS, this is true.

Petari
Software Moderator
Software Moderator
Posts: 573
Joined: Tue Nov 28, 2017 1:32 pm

Re: PC mouse refresh rate

Post by Petari » Mon Nov 05, 2018 5:03 pm

"The most fundamental difference between CRT and LCD is that the latter (usually) doesn't redraw the complete image every frame. LCDs only update the difference. That's what sample-and-hold is about. Some LCDs have a strobed backlight (and some OLEDs have black frame insertion) which effectively does cause each frame to be completely redrawn. That is what the eye needs to see or otherwise it thinks that only the edges of the object have moved (https://en.wikipedia.org/wiki/Persistence_of_vision)."
That's complete BS. Inserting black frames means: stroboscope effect again (back to unhealthy CRT type displaying).
Sample and hold as is described is not real sample and hold. If some pixel does not change then it will be same after redraw and eye can not see difference - because there is no at all. Sample and hold actually sounds completely ridiculous here, it is for cases of taking some sample - like voltage in very short moment and then keep that value until next sample. Monitor does not sample - it gets RGB intensity levels for all pixels in row, sets them and after short delay (now usually couple mS) it is visible. And LCD redraws all pixels, just there is not black period between. Is there insertion of black periods in watching of real movement directly ?

Mouse flickering can be caused by crap video card (driver) for instance. Don't blame LCD monitors. What is done in last 12 years is fast enough for 60 Hz motion, to say so.
There is 2 kind of people: one thinking about moving to Mars after here becomes too bad, the others thinking about how to keep this planet habitable.

User avatar
sandord
Trusted Guru
Trusted Guru
Posts: 470
Joined: Mon Aug 13, 2018 10:08 pm
Location: The Netherlands
Contact:

Re: PC mouse refresh rate

Post by sandord » Mon Nov 05, 2018 5:35 pm

Petari wrote:
Mon Nov 05, 2018 5:03 pm
That's complete BS. Inserting black frames means: stroboscope effect again (back to unhealthy CRT type displaying).
It certainly is not BS. Black frame insertion indeed increases flickering but if applied at high frame rates (say 200 Hz) and the effective frame rate is still beyond 80 Hz or so (100 Hz in this example), the picture will show much sharper percieved (!) motion than without black frame insertion. This is the reason why the latest TV models (mostly OLED) are implementing BFI.

And some people are less sensitive to flicking, they may benefit from the increased motion sharpness without being annoyed by the flickering (if the effective frame rate is low).

If you do some research on perceived motion unsharpness, you'll find that the fact the (most) LCD backlight is stationary, will cause the eye to think that an object that is supposed to be perceived as moving, isn't moving.

To get a sense of how this works, imagine a solid gray square (without any texture) sitting in the middle of an LCD screen. Now, the square is moving to the right. What the eye actually sees is a small strip disappearing from the left of the square and a small strip appearing at the right side of the square.

If instead of that the backlight would be strobing (just like a CRT does, admitted it doesn't have a backlight but it draws the picture every frame after which the phosphor dies to black out well within the frame duration), the eye would receive a complete new picture and actually perceive the square as moving to the right.

Now, you may ask, why does the eye perceive things the way I described here. Well, it's due to a phenomena called 'persistence of vision'. You can see the effect for your self on this site: https://www.testufo.com/blackframes

The above experiment is exactly why non-strobing LCD TVs (and OLED to a lesser extent) suffer from motion unsharpness. And that is why frame generation techniques are deployed to reduce the effect. That works because the shorter the frame duration, the lesser the unsharpnes. If you imagine reality having an infinitely high frame rate, the motion unsharpness is requivalently low. But, in reality, the eye can't cope with very fast changes and does have it's own motion blur, which is hidden by the brain from your conciousness (Saccadic suppression) but that's a whole different story (see https://en.wikipedia.org/wiki/Saccadic_masking).
Petari wrote:
Mon Nov 05, 2018 5:03 pm
Sample and hold as is described is not real sample and hold. If some pixel does not change then it will be same after redraw and eye can not see difference - because there is no at all. Sample and hold actually sounds completely ridiculous here, it is for cases of taking some sample - like voltage in very short moment and then keep that value until next sample. Monitor does not sample - it gets RGB intensity levels for all pixels in row, sets them and after short delay (now usually couple mS) it is visible. And LCD redraws all pixels, just there is not black period between.
It doesn't matter at all whether the LCD draws the pixels again or not, that's just a technical detail that doesn't infuence what you see. Effectively, that's hold. Compare this to a CRT, which 'releases' the pixel so to speak (the phosphor dying out).
Petari wrote:
Mon Nov 05, 2018 5:03 pm
Is there insertion of black periods in watching of real movement directly ?
Of course not, since there is no frame rate in play that reduces the rate of perception.

The black frame itself isn't actually solving anything, it's the reduction of the frame duration that does. That's where you can make a comparison to reality: those frames are infintely short so the shorter the frame, the closer to reality they are perceived. And since we don't have many frames to display (the typical motion picture has only 24 of them per second), we either need some black period in between them or generate more frames (which leads to artifacts).
Petari wrote:
Mon Nov 05, 2018 5:03 pm
Mouse flickering can be caused by crap video card (driver) for instance. Don't blame LCD monitors. What is done in last 12 years is fast enough for 60 Hz motion, to say so.
It's not simply a matter of having X frames per second and that's that. Factors such as gray-to-gray response time, black-to-white response time, can make a display supposedly running at say 100 Hz still give perceived blurry motion. Sure, crappy drivers can mess up stuff but you can't just rule out motion unsharpness caused by LCDs from existence.

Some more info on motion unsharpness: https://www.blurbusters.com/faq/oled-motion-blur/

Petari
Software Moderator
Software Moderator
Posts: 573
Joined: Tue Nov 28, 2017 1:32 pm

Re: PC mouse refresh rate

Post by Petari » Mon Nov 05, 2018 6:26 pm

Human visual perception has limited speed. It needs time to process incoming visual. That's mostly some kind of intellectual process. And needs much more effort and energy than optical part of view. And that's why 24 fps makes already good impression of movement. But new tendencies are to go on more. So, there are movies shot at 60 fps. There are many sport (Formula 1 for instance) TV broadcasts at 50 fps, and it really looks better than 25 or 30, even on 'so slow' LCD.
I don't think that it has sense to go over 100 fps. Surely some people can to 'process' more, mostly younger ones, but that's rare.

And what I see is that TV manufacturers gone too far with all those implemented picture improving techniques. One of the reasons is to impress buyers - with some overdone things. So picture looking "more realistic" than seeing it in life. There are diverse systems to enhance contrast, edges - so it may look sharper and so on. But such picture often looks unnatural. My new Philips TV is full of such 'tricks' and I usually don't like them. In many cases it is too obvious that it is overprocessed. And same stays for motion.

And when image is overcompressed it losing details - that's what I see often - faces of commentators are without details, large areas are with same color, no 'texture' - just not realistic. And old, analog TV shown often better pic, even at lower resolution. In reality there are no objects with complete solid surfaces - there is always some dirt if nothing else differ. So, that example of moving square is not real. I would rather see it with better quality (less compression, not lost details) than some added black periods what is just not how we see things in reality.
But of course, manufacturers will always claim that their solution is best, something revolutionary and so on. No thanks Wiki, I don't need you. You got corrupted as so many places this times .
There is 2 kind of people: one thinking about moving to Mars after here becomes too bad, the others thinking about how to keep this planet habitable.

User avatar
sandord
Trusted Guru
Trusted Guru
Posts: 470
Joined: Mon Aug 13, 2018 10:08 pm
Location: The Netherlands
Contact:

Re: PC mouse refresh rate

Post by sandord » Mon Nov 05, 2018 7:18 pm

Petari wrote:
Mon Nov 05, 2018 6:26 pm
Human visual perception has limited speed. It needs time to process incoming visual. That's mostly some kind of intellectual process. And needs much more effort and energy than optical part of view. And that's why 24 fps makes already good impression of movement. But new tendencies are to go on more. So, there are movies shot at 60 fps. There are many sport (Formula 1 for instance) TV broadcasts at 50 fps, and it really looks better than 25 or 30, even on 'so slow' LCD.
I don't think that it has sense to go over 100 fps. Surely some people can to 'process' more, mostly younger ones, but that's rare.
There's this whole age-old discussion about the so called 'soap opera effect', some people dislike higher fps because it reminds them of video recording as opposed to film. I guess that's a cultural thing. But that's about the number of images per second, which can be regarded separate from the amount of video frames drawn per second. For example, if you watch a 24 fps recording on your PC, the monitor certainly isn't running in 24 Hz but the image frames are distributed over the video frames. Of course, this results in a mismatch, you can't fit 24 frames in 60 Hz with equal time between them. That's why you get this frame conversion judder, another problem amongst many.

Adverted high frame rates are often related to image processing. My old Panasonic Plasma states it supports 800 Hz, which obviously isn't displayed.
Lots of BS there indeed, especially since the picture looks pretty good with all the 'enhancements' disabled.
Petari wrote:
Mon Nov 05, 2018 6:26 pm
And what I see is that TV manufacturers gone too far with all those implemented picture improving techniques. One of the reasons is to impress buyers - with some overdone things. So picture looking "more realistic" than seeing it in life. There are diverse systems to enhance contrast, edges - so it may look sharper and so on. But such picture often looks unnatural. My new Philips TV is full of such 'tricks' and I usually don't like them. In many cases it is too obvious that it is overprocessed. And same stays for motion.

And when image is overcompressed it losing details - that's what I see often - faces of commentators are without details, large areas are with same color, no 'texture' - just not realistic. And old, analog TV shown often better pic, even at lower resolution.
Petari wrote:
Mon Nov 05, 2018 6:26 pm
In reality there are no objects with complete solid surfaces - there is always some dirt if nothing else differ. So, that example of moving square is not real.
That was a thought experiment to illustrate the effect of motion unsharpness, not something you can literally translate to something like a car moving across the screen.
Petari wrote:
Mon Nov 05, 2018 6:26 pm
I would rather see it with better quality (less compression, not lost details) than some added black periods what is just not how we see things in reality.
But of course, manufacturers will always claim that their solution is best, something revolutionary and so on. No thanks Wiki, I don't need you. You got corrupted as so many places this times .
The only way to get sharpness during motion is short frames. Every technique to solve that other than increasing the number of pictures per second in the video source, has its drawbacks, either strobing effects (including lower total brightness due to the black frames) or artifacts from generated frames (which I dislike much more because they're inherently unnatural).

Some movie directors are trying to improve things by recording more frames per second (like Peter Jackson with his The Hobbit at 48 fps) but it costs a lot more money (CGI has to be applied to twice as many frames for starters and equipment has to be compatible too) so I think we'll be stuck with the current situation for a while. Simply creating better display panels doesn't cut it, frames have to become shorter in one way or the other.

Of course the gaming situation is different, with the advent of 144 Hz displays and such with the equivalently short frames they can display. But that's games, they can easily render more frames per second since all of the frames are generated anyway.

tzok
Posts: 79
Joined: Sat Dec 30, 2017 2:27 pm

Re: PC mouse refresh rate

Post by tzok » Mon Nov 05, 2018 10:04 pm

That's not completely true that LCD has no flickering. It has, and it always had. In a huge simplification pixel transparency (brightness) is proportional to voltage (charge) applied to that pixel (cell). But due to the nature of the LCD displays the average cell charge have to be 0, to avoid LCD degradation. So even if pixel has a constant brightness from frame to frame it is being constantly switched from +C to -C. Voltage (charge) polarity doesn't have effect on transparency, yet it will have to cross 0 going from +C to -C and the opposite way. Polarized cell is opaque, unpolarized is transparent thus pixel goes white during this transient (that's why black color is so poorly represented on LCDs). In OLED displays the backlight is switched off during this polarization switchover phase to compensate this effect and improve contrast. The polarization switch is done in phase with refresh rate, but 2x or 4x times faster (2x/4x per frame).

Steve
Posts: 177
Joined: Fri Sep 15, 2017 11:49 am

Re: PC mouse refresh rate

Post by Steve » Tue Nov 06, 2018 12:15 am

You guys are going into wayyyy to much detail.

Jerky mouse pointer on modern computers: Combination of 60hz & lack of desktop Vsync

To fix: Get a 120hz monitor or higher

Done :)

Petari
Software Moderator
Software Moderator
Posts: 573
Joined: Tue Nov 28, 2017 1:32 pm

Re: PC mouse refresh rate

Post by Petari » Tue Nov 06, 2018 8:17 am

Steve wrote:
Tue Nov 06, 2018 12:15 am
You guys are going into wayyyy to much detail.
Jerky mouse pointer on modern computers: Combination of 60hz & lack of desktop Vsync
To fix: Get a 120hz monitor or higher
Done :)
I think that using better driver SW will cost 'little' less :D
Why should I get 120 Hz monitor when I don't have anything over 60 fps, don't play fps games ?
If I want to spend money on some better monitor it would be more resolution and true LED monitor (OLED) - but still don't see them on market here.

Btw. cheap video cards, integrated video can not 120 Hz.
There is 2 kind of people: one thinking about moving to Mars after here becomes too bad, the others thinking about how to keep this planet habitable.

User avatar
sandord
Trusted Guru
Trusted Guru
Posts: 470
Joined: Mon Aug 13, 2018 10:08 pm
Location: The Netherlands
Contact:

Re: PC mouse refresh rate

Post by sandord » Tue Nov 06, 2018 8:19 am

Steve wrote:
Tue Nov 06, 2018 12:15 am
You guys are going into wayyyy to much detail.

Jerky mouse pointer on modern computers: Combination of 60hz & lack of desktop Vsync

To fix: Get a 120hz monitor or higher

Done :)
Details are nice ;)

But yes, that would fix it. But I'd need to replace the three monitors I have connected to my PC, one of which is 4K. That's just too expensive! Replacing only one won't work of course, being reminded of the terribly slow 60Hz every time I move my mouse onto the other two...

Post Reply