Mouse latency demonstrated with slow motion - Overclock.net - An Overclocking Community

Forum Jump: 

Mouse latency demonstrated with slow motion

Reply
 
Thread Tools
post #1 of 11 (permalink) Old 11-18-2019, 04:30 PM - Thread Starter
New to Overclock.net
 
Jonagold's Avatar
 
Join Date: Oct 2014
Posts: 226
Rep: 12 (Unique: 7)
Mouse latency demonstrated with slow motion

Decided to make a visual representation of what latency really means in terms of mouse movement per millisecond in a fast motion.


I'd like to hear what you think.
Jonagold is offline  
Sponsored Links
Advertisement
 
post #2 of 11 (permalink) Old 11-18-2019, 04:58 PM
New to Overclock.net
 
TranquilTempest's Avatar
 
Join Date: Aug 2011
Posts: 1,722
Rep: 78 (Unique: 63)
I'm not sure what you think the video is demonstrating, but it definitely isn't demonstrating mouse latency. This probably should have compared the motion of the physical mouse to the resulting motion on screen. There are better ways to test the latency of a mouse, such as lagbox.


Last edited by TranquilTempest; 11-18-2019 at 05:01 PM.
TranquilTempest is offline  
post #3 of 11 (permalink) Old 11-18-2019, 06:12 PM
lololol
 
qsxcv's Avatar
 
Join Date: Feb 2014
Posts: 4,286
Rep: 370 (Unique: 153)
well i doubt you can consistently time your click within a 2ms window.
it's like trying to hit the finger of a running target...

it's more like
if you have a 50ms window to click, if there is 5ms delay or something, that's a 10% difference which can end up being significant.

too busy to check forums as regularly
pm me if i forget to respond
qsxcv is offline  
Sponsored Links
Advertisement
 
post #4 of 11 (permalink) Old 11-18-2019, 07:14 PM
New to Overclock.net
 
TranquilTempest's Avatar
 
Join Date: Aug 2011
Posts: 1,722
Rep: 78 (Unique: 63)
Quote: Originally Posted by qsxcv View Post
well i doubt you can consistently time your click within a 2ms window.
it's like trying to hit the finger of a running target...
That part is actually somewhat plausible from a human perspective, in this case it's the timing cue that isn't consistent enough(refresh rate too low).

Maybe not 2ms, but 5~10ish easily. That would be more relevant in a game like Osu!

TranquilTempest is offline  
post #5 of 11 (permalink) Old 11-19-2019, 01:27 AM - Thread Starter
New to Overclock.net
 
Jonagold's Avatar
 
Join Date: Oct 2014
Posts: 226
Rep: 12 (Unique: 7)
Quote: Originally Posted by TranquilTempest View Post
I'm not sure what you think the video is demonstrating, but it definitely isn't demonstrating mouse latency. This probably should have compared the motion of the physical mouse to the resulting motion on screen. There are better ways to test the latency of a mouse, such as lagbox.
I am not trying to calculate latency accurately or anything, there is better tools for that. It is really just a visual representation to make it more comprehensible of what it means in practice, degrees of rotation in milliseconds.
And it is quite accurate since I am averaging out 13 frames to calculate the rotation per 1. And I am looking at parts of the movement where the speed is not accelerating or decelerating..
Jonagold is offline  
post #6 of 11 (permalink) Old 11-19-2019, 04:37 AM
New to Overclock.net
 
TranquilTempest's Avatar
 
Join Date: Aug 2011
Posts: 1,722
Rep: 78 (Unique: 63)
When you're talking about latency, you need to be very clear about which events mark your start and end point, and what latency sources are in between. A muscle memory flick shot between two stationary points is a very different scenario than visually judging when a moving object crosses in front of you, and that's different again to tracking a moving target.

The first one has a plausible mouse-only timing consistency better than 1ms, The other two scenarios include the game engine and monitor as part of the feedback loop/timing cue, so are much worse off.

TranquilTempest is offline  
post #7 of 11 (permalink) Old 11-19-2019, 09:39 AM - Thread Starter
New to Overclock.net
 
Jonagold's Avatar
 
Join Date: Oct 2014
Posts: 226
Rep: 12 (Unique: 7)
Quote: Originally Posted by TranquilTempest View Post
When you're talking about latency, you need to be very clear about which events mark your start and end point, and what latency sources are in between. A muscle memory flick shot between two stationary points is a very different scenario than visually judging when a moving object crosses in front of you, and that's different again to tracking a moving target.

The first one has a plausible mouse-only timing consistency better than 1ms, The other two scenarios include the game engine and monitor as part of the feedback loop/timing cue, so are much worse off.
What I was doing there was just visualizing how many "pixels" or degrees you can move your mouse in 1 millisecond, therefore giving proof that milliseconds matter in extreme situations.

If we try to explain a situation where that relevance would peek. As I said in the video, when one is flicking reactively to either based on sound or visual stimulus. If you see the enemy anywhere in your screen, you can react to that based of muscle memory and some of the times you have to rely on muscle memory on the click timing as well, not being able to "land" the shot.

One can test his clicking muscle memory accuracy by shooting while wiping over the screen and trying to land every bullet on the same spot.

And even when you land your fast flick-shot, trying to do it fast will cause bullet to leave too early if one is used to a certain click latency but uses a mouse with less latency.

Also, sometimes it is faster to just press the button timingly rather than trying to land exactly. For example when both you and the enemy player are moving and trying to time the shot for the right moment. In those situations I don't really try to land anything but rather to throw my crosshair towards a spot where the head will likely be next and since I don't have time to stop the mouse movement, it is easier to time the click and continue movement towards next possible spot of head instead of completely stop mouse movements on each click.

And btw, flick without slowing down for the landing is much faster and useful in close combat situations where the reaction time is everything. When landing a shot it requires you to first accelerate, then decelerate. When just flicking, you only accelerate..

When doing fast flicks, monitor becomes irrelevant for the shot after initial frame where you see the enemy, since human reaction time is around 150-200 ms and fast flick is faster than that, even if you had time to react, you wouldn't be able to change the momentum of the mouse and your arm in time. You wouldn't be able to react to any stimulus after that point when you see the enemy and start the flick.. Muscle memory requires only the initial frame where your brain reads the distance between your crosshair and the enemy.

Last edited by Jonagold; 11-19-2019 at 02:02 PM.
Jonagold is offline  
post #8 of 11 (permalink) Old 11-19-2019, 10:15 PM
lololol
 
qsxcv's Avatar
 
Join Date: Feb 2014
Posts: 4,286
Rep: 370 (Unique: 153)
Quote: Originally Posted by Jonagold View Post
And btw, flick without slowing down for the landing is much faster and useful in close combat situations where the reaction time is everything. When landing a shot it requires you to first accelerate, then decelerate. When just flicking, you only accelerate..
in practice, at least for csgo awping, almost everyone is in the decelerating part of the flick when the shot is fired. generally you still fire while moving the mouse, but if you want any sort of aim, you're not like trying to move the crosshair over the target in the shortest time possible.

but yea basically your brain is working with data that is delayed by 150ms or so. it depends on how you mentally prepare, but in many (or even most) cases you track some 10s of milliseconds of motion after the enemy appears on the screen to extrapolate where to flick to. in csgo this would correspond to determining whether an enemy is running out a corner or shoulder peeking. if you just try to shoot as soon as you see any change on the screen (which is the right thing to do sometimes), you will get baited by many shoulder peeks.

too busy to check forums as regularly
pm me if i forget to respond
qsxcv is offline  
post #9 of 11 (permalink) Old 11-20-2019, 02:40 AM - Thread Starter
New to Overclock.net
 
Jonagold's Avatar
 
Join Date: Oct 2014
Posts: 226
Rep: 12 (Unique: 7)
Quote: Originally Posted by qsxcv View Post
in practice, at least for csgo awping, almost everyone is in the decelerating part of the flick when the shot is fired. generally you still fire while moving the mouse, but if you want any sort of aim, you're not like trying to move the crosshair over the target in the shortest time possible.

but yea basically your brain is working with data that is delayed by 150ms or so. it depends on how you mentally prepare, but in many (or even most) cases you track some 10s of milliseconds of motion after the enemy appears on the screen to extrapolate where to flick to. in csgo this would correspond to determining whether an enemy is running out a corner or shoulder peeking. if you just try to shoot as soon as you see any change on the screen (which is the right thing to do sometimes), you will get baited by many shoulder peeks.
I am talking about fast peeks, like no-scopes etc. You are clearly talking of holding an angle.

Holding an angle with awp has an advantage in reaction time because you can hit anywhere on the upper body to get the kill instead of only the head, in a same length of a peek, head is visible less time than the upper body. Also angles where people usually hold with awp has the angle advantage as well. Meaning that awp will see the enemy (enemy shoulder) sooner than the peeker sees anything about the awp.

All in all. Awps holding an angle have more time to react so they can "waste" some of the time, maybe 10ms, maybe more, for determining whether to hit closer to the corner or wider, or at all, depending of the peek.

Last edited by Jonagold; 11-20-2019 at 03:03 AM.
Jonagold is offline  
post #10 of 11 (permalink) Old 11-24-2019, 03:41 PM
lololol
 
qsxcv's Avatar
 
Join Date: Feb 2014
Posts: 4,286
Rep: 370 (Unique: 153)
well no scopes in csgo are inaccurate enough that this discussion doesn't matter (i.e. there is a finite probability that a X ms delay affected whether a no scope hits, but this probability is so small that it doesn't matter)

but yea shots like this one:
, where you drag through the target can be affected by a few ms.

from my own observations/video recordings, for holding angles, i waste about 30ms if i need to react to the direction or speed of motion rather than shooting as soon as possible, and then the flick takes 60-100ms itself. in total about 300ms from first pixels appearing on screen to shot fired. it's not always an advantage to hold angles with awp because a perfect prefire can give you only about 200ms to react. (which you can hit if you just shoot as soon as possible, but then that leaves you vulnerable to shoulder peeks)

too busy to check forums as regularly
pm me if i forget to respond

Last edited by qsxcv; 11-24-2019 at 03:50 PM.
qsxcv is offline  
Reply

Tags
delay , latency , mouse , slow motion , visualization

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off