Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
JonoT

Are Wireless Mice ACTUALLY Faster??

Recommended Posts

Posted · Original PosterOP

 

We tested five wired and wireless mice with the Phantom Flex 4K camera at 1000 FPS to see if wireless mice were faster than wired!

 

Buy: Logitech G703
On Amazon: https://geni.us/myuNy
On NewEgg: https://geni.us/i26tnPk

 

Buy: Logitech G MX518
On Amazon: https://geni.us/yHTc
On NewEgg: https://geni.us/uJYz

 

Buy: CORSAIR Dark Core SE
On Amazon: https://geni.us/CFbMAH
On NewEgg: https://geni.us/7wSz3

 

Buy: Logitech® MX Anywhere 2S
On Amazon: https://geni.us/AH2U
On NewEgg: https://geni.us/CpIa

 

Link to post
Share on other sites

Shooting the video at 24fps from a camera that's moving around a lot is a bad time. I noticed the choppiness a lot in the first 2 minutes. Increase the frame rate or keep the cameras more stable. 


Laptop: 2016 13" nTB MacBook Pro Core i5 | Phone: iPhone 8 Plus 64GB | Wearables: Apple Watch Sport Series 2 | CPU: R5 2600 | Mobo: ASRock B450M Pro4 | RAM: 16GB 2666 | GPU: ASRock RX 5700 8GB | Case: Apple PowerMac G5 | OS: Win 10 | Storage: 480GB PNY SSD & 2TB WD Green HDD | PSU: Corsair CX600M | Display: Dell 27 Gaming Monitor S2719DGF 1440p @155Hz, Dell UZ2215H 21.5" 1080p, ViewSonic VX2450wm-LED 23.6" 1080p | Cooling: Wraith Prism | Keyboard: G610 Orion Cherry MX Brown | Mouse: G303 | Audio: Audio Technica ATH-M50X & Blue Snowball
Link to post
Share on other sites

As I thought, you really bought that 4K 1k FPS camera? :D Geez. What's Brandon's opinion on it? ;D

 

... so will you do a PS/2 and USB comparison?

Link to post
Share on other sites

It seems like an opportunity was missed here to include a sample for those in the Razer product environment. Maybe they make fewer latency claims and get less pro play, but their consumer market share is sizeable.

Link to post
Share on other sites

For Corsair, Darkcore is advertised as a 1ms mouse, not sub 1ms.  The Ironclaw is advertised as sub 1ms. From their website:

 

 

DARK CORE RGB features ultra-fast 1ms wireless, Bluetooth® compatibility and wired mode for performance gaming with versatile connectivity. The custom, high-accuracy 16000 DPI optical sensor and comfort-contoured shape ensures precision and control.

 

The CORSAIR IRONCLAW RGB WIRELESS Gaming Mouse combines a native 18,000 DPI optical sensor with a comfortable palm grip, connecting to your PC via hyper-fast, sub-1ms SLIPSTREAM CORSAIR WIRELESS TECHNOLOGY, Bluetooth or USB wired connection.

 

 

Link to post
Share on other sites

Unless they come up with an internal pci-e card, you're still gonna be limited by the USB polling rate, which is 1ms at best (default is 125 Hz or an update every 8 ms)

Link to post
Share on other sites

Linus (sort of) justified the 16:10 video ratio (still not convinced), but there is no valid reason to start shooting in 24fps.

 

The fact that you chose to do it on a video featuring a high-speed camera recording reaction times just adds insult to injury.

Link to post
Share on other sites

Also to add... you could have soldered two wires directly under the left click button and use that as a way to detect when click happens or to simulate a click (and also use that as a trigger, for the camera... for example light a led or something when click happens)

 

Would be also interesting to have the mouse locked in place and just short the wires to make the mouse think you pressed the button, without actually touching the mouse.

The idea is that a wireless mouse may go to "sleep mode" to save battery power after a few seconds of no movement, so by shorting the wires to fake a mouse click you're guaranteeing you're not waking up the mouse before the actual click.

So, you can measure the latency between fake click on button and when the click is recognized in the operating system  - you'll have wake up latency, transmission, the 1ms polling time in Windows then whatever amount of processing time the driver needs inside Windows.

 

The video is kinda crappy, the motion of that hammer moving down towards the mouse could be enough to shake the bench or vibrate it enough for the mouse to wake up from battery saving mode.

Link to post
Share on other sites

Hey Linus Group, Did non of you Remember the 480Hz screen that you have in the storage somewhere? Or did you have to send it back?
https://youtu.be/rG_a7Qs2odc?t=325

Link to post
Share on other sites

So your going to milk this topic for another video or you really missed an opportunity to compare some more mice.  

 

Id love to see other mice tossed into this showdown

$7 amazon basic wired mouse

$40 razer death adder  (cough razer sponser cough)

 

What does bluetooth connection look like?

What does wireless receiver dongle plugged in under the desk 5' away look like?

What does plugged in to the slow-ass-USB-monitor-hub look like?  (both wired and wireless)

 

Looking forward to part 2 of this vid!

Link to post
Share on other sites
2 hours ago, JonoT said:

 

We tested five wired and wireless mice with the Phantom Flex 4K camera at 1000 FPS to see if wireless mice were faster than wired!

 

Buy: Logitech G703
On Amazon: https://geni.us/myuNy
On NewEgg: https://geni.us/i26tnPk

 

Buy: Logitech G MX518
On Amazon: https://geni.us/yHTc
On NewEgg: https://geni.us/uJYz

 

Buy: CORSAIR Dark Core SE
On Amazon: https://geni.us/CFbMAH
On NewEgg: https://geni.us/7wSz3

 

Buy: Logitech® MX Anywhere 2S
On Amazon: https://geni.us/AH2U
On NewEgg: https://geni.us/CpIa

 

Now do more mice and also test clicks! (there could be a difference between clicks versus sensor movement in the processing pipeline of the controller of the mouse and so on and so forth)

 

And also test different windows and GPU/CPU/mobo configs to see if any particular architecture/brand combination gives lower latency (linux would be interesting to be included as well)

 

It would be great to have a matrix with all the results of the most reputable mice out there and you will be the first ones to do it! :) 

Link to post
Share on other sites
1 hour ago, schmism said:

So your going to milk this topic for another video or you really missed an opportunity to compare some more mice.  

 

Id love to see other mice tossed into this showdown

$7 amazon basic wired mouse

$40 razer death adder  (cough razer sponser cough)

 

What does bluetooth connection look like?

What does wireless receiver dongle plugged in under the desk 5' away look like?

What does plugged in to the slow-ass-USB-monitor-hub look like?  (both wired and wireless)

 

Looking forward to part 2 of this vid!

 

I REALLY wanted to do more mice but it was a matter of restricted time. We had rented the camera for the shoot with Gav (day 1) and had to return it the next day. We said we'd return it by 3 (have it leave the office at 1). Which meant we had to do the spoken bit, do the tests, some analysis to make sure things were good, then finish the tests within a few hours AND ingest the footage (which is painfully slow).

 

=P NEXT time though!


Follow me on Twitter at: @edzelyago

Link to post
Share on other sites

I see input lag measured around the internet,

 

WydD (Medium.com)

  • Wired PS4 controller vs Wireless, measured via USB Host Shield (1ms device), Wireless on avg was 7.3 to 1.8 ms quicker depending on the dongle, coming in at 3ms avg with a good one. Plus other arcade sticks.

 

Rocket Science (youtube.com)

  • photosensitive diode vs 1K FPS camera setup, camera was 1ms quicker
  • razor naga vs arduino, arduino was 4.4ms quicker despite both being 1000Hz (1ms) devices
  • the arduino had 10.8ms of total response time

TFTCentral (TFTCentral.co.uk)

  • The quickest monitor measured with their own photosensitive diode&oscilloscope scope system, Asus ROG Swift PG279Q, 3.25ms total display lag
  • The PG35VQ@200Hz has a measured 2~3ms g2g at certain shades

LTT (youtube.com)

  • XL2730Z@200Hz avg of 21.5 frames/960FPS of input lag or, 21.5/960*1000, 22.4ms total

So this recent video of yours you've got the Acer FX250Q or A monitor which I couldn't find 3rd party measurements for, anyway you've got sub 14ms response times now.. so I've seen a fast clicking arduino with a 10.8ms total response time and your quickest mouse response of 13ms total response time, both on 240Hz monitors. 

 

TL;DR version - I'd say after a certain point the game engine (larger part?) and display must be the bottle neck if a USB Host Shield, intercept signals from devices like you described at the beginning of your video (check medium link for more info), can measure button response time avg of 3ms on a PS4 pad over a good bluetooth dongle.

 

Would that be fair to say?

Link to post
Share on other sites

I loved the video but you guys missed a TON of opportunities, like:

Testing with different types of mousepads, with different surfaces, such as Logitechs Powerplay, Artisan Shidenkai, Madcatz Glide 38, Steelseries Qck and etc. (maybe this is why u cant reach 1ms?)

Testing with different DPIs, 400 and 800 (used by most CSGO pros), then 1600, 3200 and up.

Testing microstutter using different Polling Rate. Yes it does make a huge difference using 1000mhz polling rate and you can actually see it.

Testing different games.

Link to post
Share on other sites

I found it funny that Edzel said it doesn't seem right, when he first saw the results. He's actually right about that.

 

I know that Acer KG251QD they used all too well, that was my first 240Hz monitor I bought, when I made the switch from 144Hz. Sadly it has noticeably more input lag than usual, compared to other 240Hz, or even decent 144Hz monitors.

 

When I first got it, I just fired up CS:GO to test it, and despite being really smooth, something about fast flick shots just immediately felt off, so I started testing its input lag and comparing it to my old 144Hz (Acer XF240H), and my tests immediately confirmed it.

I test input lag in almost exactly the same way as in this video. I would bump the mouse with another object, film in slow motion, and count frames between the mouse moving and the screen changing. I only have 240fps slow motion though which isn't as precise, but that's why I average at least 10-15 tries to get a more accurate number.

 

With that 240Hz Acer, I got an average around 12ms or so, which is very close to what Linus measured, especially given the fact that their game was running at a lower fps, while I always do these tests with 400-500fps in game. Now after I returned that monitor and got an Alienware AW2518HF instead, this one consistently measures around 8ms in the same test, and finally fast flick shots felt right again on it.

 

To be fair this doesn't affect their conclusion, because the monitor stayed the same across their tests, but the overall numbers they are giving are still a bit misleading, because on a decent monitor you would see less input lag than that.

 

Link to post
Share on other sites
15 hours ago, PeterBocan said:

As I thought, you really bought that 4K 1k FPS camera? :D Geez. What's Brandon's opinion on it? ;D

 

... so will you do a PS/2 and USB comparison?

He said at the beginning that they rented it.

Link to post
Share on other sites

I also wanted to point out some neat ideas I saw in the comments, like doing a video on screen tearing in slow motion would be nice. And also just testing other mice, such as cheap office ones, the MX series, and whatever regardless of whether they're wireless or wired. And finally testing in the Linux low-latency kernel.

Link to post
Share on other sites
16 hours ago, Miracle Nachos said:

The fact that you chose to do it on a video featuring a high-speed camera recording reaction times just adds insult to injury.

you know that makes no sense right? increasing the frame rate of the playback video reduces the effectiveness of the high speed,

 

from the man himself

 

 


Judge the product by it's own merits, not by the Company that created it.

 

 

Link to post
Share on other sites
8 hours ago, Arika S said:

you know that makes no sense right? increasing the frame rate of the playback video reduces the effectiveness of the high speed,

 

from the man himself

 

 

I did not know that, thank you for the info.

Link to post
Share on other sites

So, 500hz polling rate is actually a problem in certain scenarios. If you have a 240hz tn panel with strobing backlight, you have a fast enough panel with enough motion clarity that you can actually see the cursor stuttering from the low poll rate.

Link to post
Share on other sites
9 hours ago, Arika S said:

you know that makes no sense right? increasing the frame rate of the playback video reduces the effectiveness of the high speed,

 

from the man himself

 

I imagine you can just play the slow mo frames twice to effectively halve the frame rate of that part, while keeping a high frame rate for Linus.

Link to post
Share on other sites

I don't understand the point beyond interest.  Is there an observation of better outcomes from having reduced latency even at these levels?  surely it can't have anything to do with human involvement because the average reaction time to visual stimulus is 180ms and it seems we are talking lag in single digit ms.  That's more than ten times faster,  which can only increase reaction accuracy by less the 1% at best. 


QuicK and DirtY. Read the CoC it's like a guide on how not to be moron.  Also I don't have an issue with the VS series.

Sometimes I miss contractions like n't on the end of words like wouldn't, couldn't and shouldn't.    Please don't be a dick,  make allowances when reading my posts.

Link to post
Share on other sites

The test server had a tick-rate of 64. 

That means that all engine calculations were done 64 times per second.

Movements, positions, netcode, hitdetection almost everything in the game is updated once every 15.625 ms.

 

1 / 64 = 0.015625

15.625 ms of engine latency
+- 1 ms margin of error for the high speed camera.

 

All of the movements were capped by the engine. The net_graph keeps updating after the refresh-rate, it does not use the same clock. You will have to increase the tick-rate to get a result under 14.625 ms. The lesson is that you should use a tick-rate of at least 1000 when making a test like this.

Link to post
Share on other sites

All you get with a high refresh-rate monitor is more draws with a lower buffer latency. A 60 Hz monitor takes the latest frame every 16.66 ms in the buffer and draws it upon itself. A 240 Hz monitor takes the latest frame every 4.16 ms.

 

The rendering takes time but lets keep it simple

A frame can be 0.000000001 - 16.6666666 ms old in the buffer with a 60 Hz monitor. 

A frame can be 0.000000001 - 4.16666666 ms old in the buffer with a 240 Hz monitor. 

 

A higher refresh-rate will reduce the maximum time a frame can lay and wait, before it is grabbed and drawn on a monitor. Under the right circumstances, you can reduce this time by 12.5 ms. But you can also reduce the frame age by rendering more frames faster. Placing a new frame just before the monitor grabs the latest.

 

Old CRT monitors were particle accelerator where the accelerating anode turns the electrons into a high speed-beam of electrons. That gain around 600km/s for every volt they are accelerated through. A LCD monitor is unfortunately not as fast and the drawing process can take a lot of time. The picture can be very difficult to interpret during this period and should therefor be included as latency. Let's call it LCD blur.

 

There is a very noticeable reduction in delay by reducing rendering times and using a high refresh-rate monitor.


But it's far better to just ignore the game all together and make your own low latency application for testing a mouse. So you can test without additional latency and make it a battle of who has the lower latency, a mouse or the monitor? But it would be a waste of time, since Linus have already demonstrated how tick-rate holds you back where such gains matters, Gaming. 

 

You can get your 240 Hz and 1000 fps with no positional difference of the characters. You could not compare the gaming mouses due to the tick-rate. The Smoothness is capped by the tick-rate. The engine made its calculations 64 times per second.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×