overly twitchy/shaky/not smooth mouse at 144hz?

[wiseude]

hen swinging my mouse left and right/up and down I sometimes see my mouse jitter/twitch and overly shaky.At times it also doesn't look as smooth as it should on 144hz when moving around.
I have v-sync set to "ON" on global and I'm using driver 417.22

I'm also thinking it could be a mousepad/mouse issue since they are like 3-5 years old (hard mat and a gs400)Yea I did try to clean them but no luck.

Then I thought... what if it's the monitor? I tried https://testufo.com/frameskipping noticed my fps either spikes momenterally to 145 then stabilized to 144 or in some rare occasions it got stuck at 143fps until I refreshed.I do also see the occasional dip in fps to 140 or even 130s in rare occasions.

my monitor is a AOC G2460PG 144hz g-sync monitor That has dark bright spots shaking at 144hz but not on 120hz.(it's like on 144hz the brightness increases but when you switch to 120hz all that brightness goes away.)

Monitor is 3.5 years old.Could it be a cable issue? or a mouse/mouse mat issue?

picture below showing the occasional stuck at 143 fps.

[Chief Blur Buster]

picture below showing the occasional stuck at 143 fps.

For www.testufo.com/frameskipping -- can you please attach a photograph and not a screenshot?
I want to see the frameskip pattern (bright spots or dark spots)
Thanks!

As for 143fps versus 145fps, that is sometimes normal if it's a fractional 143.4Hz being rounded-off to 144Hz or 143fps, as there's some caveats with web browser ability to properly detect refresh rates. Try testing www.testufo.com/refreshrate and seeing what your fractional refresh rate looks like.

[wiseude]

picture below showing the occasional stuck at 143 fps.

For http://www.testufo.com/frameskipping -- can you please attach a photograph and not a screenshot?
I want to see the frameskip pattern (bright spots or dark spots)
Thanks!

As for 143fps versus 145fps, that is sometimes normal if it's a fractional 143.4Hz being rounded-off to 144Hz or 143fps, as there's some caveats with web browser ability to properly detect refresh rates. Try testing http://www.testufo.com/refreshrate and seeing what your fractional refresh rate looks like.

I used my brothers go pro and took multiple pictures and I pretty much saw this.Couldn't get him to send me the pictures because it was low on battery and the charger was no where to be found but I drew a crude version of it in paint xD

Basically what was shown is like the picture below but happening in each row from the go pro view.(turns out a gopro can take 10 pictures a second)

So I guess there is no frame skipping.Just a shitty mousepad/mouse.

[RealNC]

What mouse do you have? Are you using too high DPI maybe? Or too low DPI but too high in-game mouse sensitivity?

[wiseude]

What mouse do you have? Are you using too high DPI maybe? Or too low DPI but too high in-game mouse sensitivity?

I have a 4 year old logitech g400s.Using default dpi.It's not really an ingame issue.More like general desktop issue.Like when I'm watching streams or having a tab open.I tried using isopropyl alchohol to clean the sensor and compressed air but no dice.Could be something a cleaning can't remove.Maybe some sort of stain or spec of something thats too far in.

Another culprit could be the hard mouse pad I use.I can feel drag spots in the middle and smooth spots on the edges and left side.(it's 3-4 years old too)I think I wore the plastic down or something because even cleaning it doesn't remove the drag spots.


I ordered a new mouse pad and mouse just incase.

[Chief Blur Buster]

It's a good idea. Some newer mice are really good.

If motion fluidity is numero uno (in all sync modes including strobed VSYNC ON), I suggest testing out the new sensors, and maybe even a true-2000Hz poll rate (e.g. Cougar Minos X5). But some swear by lower pollrate like 500Hz, though sensors have improved so much that I don't recommend under 1000Hz poll rate if you care about mouse fluidity.

[wiseude]

It's a good idea. Some newer mice are really good.

If motion fluidity is numero uno (in all sync modes including strobed VSYNC ON), I suggest testing out the new sensors, and maybe even a true-2000Hz poll rate (e.g. Cougar Minos X5). But some swear by lower pollrate like 500Hz, though sensors have improved so much that I don't recommend under 1000Hz poll rate if you care about mouse fluidity.

Yep.Hopefully a new mouse/mat fixes my problem...it kinda pisses me off how my mouse pointer goes from fluid 144hz to something that doesn't feel like 144hz fluidity.Back and forth.

if it doesn't it has to be a driver problem but no one has had any mouse problems with 417.22 so I would be facing a wall.

[Chief Blur Buster]

Adjusting the DPI-versus-sensitivity balance helps a lot. In many games I find fluidity improves with "higher mouse DPI-low game sensitivity" combinations especially for slow mouseturn situations.

Basically, when I turn extremely slowly while admiring scenery in a solo FPS adventure game, I want my slow mouseturns to be smooooth, not steppy-steppy-steppy turns. That's why I like 1600dpi (and sometimes 2400dpi or 3200dpi) during these times, I can get "TestUFO-silky-smooth mouseturns and pans" even for slow mouseturn and slow mousedrags, if that sorta thing is important to you.

Importing the "TestUFO stutterless, stepless smooth mouse turns" into FPS games -- at all slow/med/fast turnspeeds -- can require ultrahigh DPI + ultralow ingame sensitivity combinations, but that can sometimes (slightly) distort fast mouseturns for twitch eSports, depending on how the mouse/drivers/game interacts. Historically, gaming mice were crappy at 1600dpi+ or 3200dpi+. That's less and less the case.

The new sensors have fixed a lot of problems at 1600dpi and now more eSports players have switched to 1600dpi to allow stepless slowturns without sacrificing flick fastturns.

However, the aimfeel can feel "wrong" if you're in twitch games and you need a more balanced sensitivity-to-DPI ratio, common is 400dpi or 800dpi, but that's waaaaaay too low for me for strobed ULMB VSYNC ON when I want stepless mouseturns even for sloooooow mouseturns while gawking scenery or such.

Now, that's a totally different use case than paid eSports competitive. Mousefeel is sometimes better at lower DPI such 800dpi, though some champion players like sprEEEzy (I moderate a channel on his Discord server now, #monitor-tech-talk) now use 1600dpi for their mouse. It depends on what kind of gamer you are.

Hopefully your new mouse has one of those new ultraaccurate sensors that are perfectly great at 1600dpi. For such insane quadruple-digit DPI, you NEED a really good mousepad engineered to have low tracking error rates at high DPI -- and replace the mousepad occasionally due to wear-tear like you have.

Yes, yes, desktop mouse arrow will be too fast. But you want "TestUFO-smoooooth mouseturns" in FPS at all turnspeeds, slow turns and fast flick turns? Sorry, no choice. Use mouse profiles (e.g. Razer, Logitech) to automatically jack up the DPI when steam.exe launches, and automatically decrease DPI when steam.exe exits. ... And slow down the game mouse turns by sliding the in-game sensitivity slider very low in the game to compensate for ultra-high DPI.

That way, you get the same mouse turnspeed at 3200dpi or 6400dpi as you used to at 800dpi.... Except ultra-slow-turns and ultra-slow mousepanning are massively smoother. Stepless slow turns where your mouse really slowly millimeters across your mousepad. That said, extreme DPI that are not well implemented can degrade accuracy of flick turns, so it's extremely important that the mouse is as flawless as possible at the quad-digit DPIs -- that includes the mouse, sensor, feet, mousepad... Then you've got the silk feel at all slow/fast turns.

If you have wear-tear, you may get more accurate aiming at lower DPI like 800dpi. But eliminate the weak links (sensor, pad, mousefeet, etc), and higher DPI ends up better, giving you ultrasmooth slowturns without sacrificing the quality of your fast flickturns. Sometimes I feel the advantages of >3200dpi during slowturns (especially in strobed VSYNC ON!) -- but usually I stick to 1600dpi depending on the mouse brand, and many of the top eSports players now are increasingly switching to 1600dpi nowadays. It can require more frequent mousefeet cleanings / mousepad replacements to keep 1600dpi or 3200dpi or higher staying topnotch though!

In fact, some of the newer gaming mice no longer have an 800dpi setting that used to be the mantra -- minimum starting at the quadruple digits.

[RealNC]

You can use a paint program (the old Paint application in Windows works.) Use a 1 pixel thick pencil and draw lines and spirals. See how jittery they are.

See:

https://forums.anandtech.com/threads/mo ... st.2348268

[Chief Blur Buster]

You can use a paint program (the old Paint application in Windows works.) Use a 1 pixel thick pencil and draw lines and spirals. See how jittery they are.

That's a good test but it's beginning to fail.

If you use full 3200dpi or 6400dpi mouse sensitivity, it can be impossible to control the Windows Desktop.... almost. 6400 pixels of mouse arrow movement over one inch, means a 1080p display traversal on only one-sixth of an inch vertically. Your mouse arrow goes rocket sled zoom from top edge to bottom edge of the screen in less than 10 millimeters at 6400dpi, preventing you from doing a proper accurate paint-program test of 6400dpi, so this old test begins to fail. Don't even get me started on the sheer insanity of 12800dpi where mere millimeters of mouse movement can potentially cause a full screenwidth rocket -- with acceleration disabled. Some older games don't even have enough mouse sensitivity range (bottom of slider bar) to compensate for such white lightning. You don't have enough screen resolution anymore for spirals for DPI testing unless you're on an 8K display! We need other alternative ways to measure accuracy of high-DPI. If only we had a 163840x8640 display specifically for 6400dpi mouse testing in an old paint program (and even with that, limiting to mouse spirals only one inch high on the mouse pad), then old paint programs may still be practical for mouse DPI testing. But we're now in the DPI stratosphere where the paint program accuracy testing begins to fail because we can't do big spirals for sprial testing. We're starting to hit limiting factors of precision for DPI but high-Hz and the need for stepless slowturns

Now, the thing is 1 pixel on desktop can mean 10 pixels mouseturn in a game (steppy, steppy, steppy mouse turns when you drag mouse one millimeter per second in a FPS game). Now, you jack up mouse DPI and dramatically lower in-game sensitivity. Then you've might have a situation of 10 pixels on desktop and 1 pixel mouseturn in a game (ultrafine mouseturns).

Now a slow 240 pixels/sec ultrasmooth mouseturn (tracking a slow moving object) by dragging the mouse a few millimeters per second, becomes stepless.

But wait! Mouseturns are subpixel accuracy.

Even your slowest mouseslide (a few desktop pixels per second) becomes 240 unique positions per second, at 240 frames per second on a 240Hz monitor. Instead of the steppy-effect, you're getting that "TestUFO smooth effect" on all slowturns. Since mouseturns are subpixel capable because of how scenes can render (floating point coordinates), that means a 50 screen-pixel mouseturn is 240 unique subpixel-moved frames rather than a steppy effect of 5 frames 10 pixels apart (from a low DPI mouse)

What you formerly single-pixelled in PhotoShop or Paint (aiming for a single-pixel), you're potentially rendering only 10 new mouseturns frames per second that is coarsely stepping every 1/10sec. So your 240fps monitor is playing only "10fps" because you've only generated 10 new mouse positions on a low-DPI mouse trying to readjust your mouse arrow onto a tiny slow-moving faraway target. That's poor slowaim feedback.

Now you've got the PhotoShop "trying-to-aim-for-a single-pixel" coarseness in your mouse, except amplified somewhat because the whole screen is moving in 10-pixel-steps at a time. Ouch.

Most people adjust better than that, but you still have the steppy-steppy effect of slow mouseturns, which makes precision slowaims a bit more challenging if you've optimized your low-DPI mouse for reliable flickturns. Ouch. I can't have both the cake and eat it too.

That coarse stepping effect is a form of mouse microstutter caused by DPI limitations. That stepping effect is a DPI limitation and requires ultra-DPI and lowered ingame sensitivity. With enough DPI, you can have even have 240 separate subpixel mouse positions over the entire millimeter (or few) of mouse movement (game sensitivity dependent) with perfectly smooth mouseturn (on a stiction-free mousepad with new mouse feet), producing a TestUFO smooth effect of a step-free 1 pixel movement per refresh cycle (or even subpixel) over a tiny smooth mouse move.

Although tempted to say "1600dpi ought to be enough" in the vein of "640K ought to be enough" or "512Hz ought to be enough", I'm not even going there. The diminishing returns are still beyond all of the above.

Now, the beaut of mouseturning 30-to-50 pixels with 240 unique subpixel-moved frames per second, is a silky effect to behold -- completely stepfree -- milked out of a ~3200dpi mode on a 12800dpi mouse (not even 1600dpi is enough to go fully TestUFO-smooth in all slowturns). You definitely need as frictionless as possible for that.

Sure, it's a very subtle thing. But now we're having cake and eating it too. Full fast flickturns (at the same velocity we're used to) with no degradation. Same mouseturn speeds we're used to. Except with high-quality ultra DPI, we are now automatically gaining completely stepfree stutterfree distractionfree ultraslow precision mouse movements, for those "tiny aim" situations that are so common in certain games such as Fortnite.

Like precision machine-mowing down 10 distant wooden boards left-to-right in one slow sweep, not being bothered by the steppy-motion of 400dpi or 800dpi. Sure, your hands may be shaky trying to mow those tiny targets, but at 240Hz (with high-DPI mouse + low ingame sensitivity) you're getting 240 fully-feedbacking-to-your-eyes new fullscreen repositions per second rather than just 10 granular turn positions. From a slow sweep you're trying to do mowing all that built Fortnite stuff down. You're got quite a very noticeably better slowturn feel even at 20 pixels/second mouse movement, with those 240 subpixel repositioning rather than just 5 or 10 or 20.

Sure, Fortnite players don't really think about the benefits that the modern gaming mice has now afforded them, but it's quite so obvious watching Twitch. How "cake-and-eat" slowturns+fastturn capability is needed -- and the DPI by the truckload -- has given Fortnite players. It's now taken granted-for, but the benefit is there.

Such uncanny precision some Fortnite players have achieved, and many are using quad-digit DPI now. You can use a bigger mousepad (3 footer mousepads!) to compensate for the low DPI to give you precision mouseturns at low speeds, and that's another technique.
Now if your hands are shaky for any reason, you'll need to go for the "3 foot mousepad" technique instead. That may be better for you.
That means you have to swipe the mouse a longer distance for a 180, some players prefer that. Your hand accuracy limitation is a limiting factor, obviously. (Even high DPI can still help with that too, just not nearly as much).

But others prefer small flicks for the fast 180-degrees turns. Less distance means flicks can be much faster if your hand is accurate. That's where ultra-DPI comes to the rescue for stepfree slowturns that often gets thrown out of the window for players who prefer small wrist flicks for fast turns. Good news, not anymore! Sensors are good enough. You can now finally fix the precision at the bottom of the mouseturn speed spectrum too, without changing the velocity nor accuracy of your full-speed flickturns, by a carefully symmetric DPI-vs-mouse-sensitivity rebalance.

The diminishing points of returns still surprise somewhat, and the obvious benefits of ultra-DPI for precision slowtracking. The human-worth headroom is far beyond 1600dpi in some situations. The limiting factor is simply trusting the mouse sensor quality and reliability, and the precision/quality of the mousepad/mousefeet/stiction/etc since those start to become serious limiting factors for ultra-DPI.

In the past, more than 10 years ago, 1600dpi was just...unreliable/inaccurate/or had compromises (e.g. limited mouse velocity for flick turns) etc. But 1600dpi technology has improved to the point where it's becoming quite a standard setting by many competitive players I'm talking to. The key is using the tool properly.

Now instead of feeling like "you're aiming at a single pixel" trying to aim at 10-pixel targets because of the steppy-steppy effect of 400dpi (even 800dpi feels quite noticeably steppy in Fortnite) at certain game sensitivity settings (because you calibrated it for your favourite mouse flick turn speeds)... You feel like you're aiming at real 10-pixel targets when precision aiming at those 10-pixel targets (or continually slewing slowly to mow distant Fortnite structures). Great full flickturn accuracy. Great slowturn accuracy all the way down to your hand's accuracy limitations. Now you're eating cake!

The good news is that the modern new sensors are feeding out what seems to be quite accurate quad-digit DPI (given the right mouse, pad, feet, lack of stiction, etc) without sacrificing flickturn reliability. And now we have to figure out how to milk that properly.

Done well (the "high DPI, low ingame sensitivity" rebalance), you gain ultra-precision slowtrack capability while maintaining the same flickturn accuracy we're used to.


We need a rawinput-compatible paint program with subpixel accuracy for the famous "spiral test"