[EXPERIMENT]Keyboard latency at different debounce settings

Everything about latency. Tips, testing methods, mouse lag, display lag, game engine lag, network lag, whole input lag chain, VSYNC OFF vs VSYNC ON, and more! Input Lag Articles on Blur Busters.
User avatar
ko4
Posts: 126
Joined: 06 Jul 2018, 16:14

[EXPERIMENT]Keyboard latency at different debounce settings

Post by ko4 » 09 Feb 2020, 18:38

testing methodology

Game: Fortnite 240fps locked in-game exclusive fullscreen
Keyboard: Ducky One 2 Mini (adjustable debounce) 1000ms polling rate
Camera: Note 9 960fps
Monitor: Dell s2716dg

I am hitting the key by flicking my finger. Easy method to get the point of actuation with id say 4-5ms accuracy
I stop counting frames at first on screen reaction so the 144hz refresh rate should not cause any unwanted variations

sample video:
phpBB [video]


Results:
results.PNG
results.PNG (10.33 KiB) Viewed 22027 times
https://docs.google.com/spreadsheets/d/ ... sp=sharing


These values do not represent the actual latency of the keyboard but the whole chain from keypress to display on my particular setup
However, the results still show that the debounce settings does in fact have an effect on keypress to usb latency

EDIT:
Decided to test some of my other boards and got a few surprising results
results2.PNG
results2.PNG (7.15 KiB) Viewed 21994 times
Surprised to see that the Corsair is faster than all of the ones i've tested so far because you'd think theyd use oem pcb's like any other just like they do with their power supplies

I will test the Logitech G-pro wired keyboard when i get my hands on it
Should get some interesting results because this board was shown to have 9 ms on a Razer keyboard
phpBB [video]

UPDATE:
finally got my hands on g pro romer-G

here are the results
results 5.PNG
results 5.PNG (2.19 KiB) Viewed 21862 times
Last edited by ko4 on 12 Feb 2020, 16:19, edited 3 times in total.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: [EXPERIMENT]Keyboard latency at different debounce settings

Post by Chief Blur Buster » 09 Feb 2020, 19:35

That's pretty neat!

Thank you for sharing your results.

The debounce latency is definitely statistically showing up in your results!

Immediately, right off the bat, I can see that the 240fps + 240Hz (1/240sec granularity) is probably one of the biggest error margin -- possibly bigger margin than finger flick interpretation.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

MatrixQW
Posts: 278
Joined: 07 Jan 2019, 10:01

Re: [EXPERIMENT]Keyboard latency at different debounce settings

Post by MatrixQW » 11 Feb 2020, 09:20

Isn't it better to just use ps/2 keyboard?

Krizak
Posts: 11
Joined: 05 Feb 2020, 18:56

Re: [EXPERIMENT]Keyboard latency at different debounce settings

Post by Krizak » 13 Feb 2020, 23:28

Any chance you could try testing a keyboard with the optical switches. Even though your testing is capturing the whole chain and is subject to some error, would still be interested in seeing the results. The following are 2 keyboards I know that use optical switches even though slightly different technologies:

Bloody B930/B970
https://www.amazon.com/Tenkeyless-Keybo ... 02&sr=8-11
https://www.amazon.com/Keyboard-Bloody- ... 446&sr=8-3

Wooting One/Two
https://www.amazon.com/WOOTING-Linea%E2 ... 484&sr=8-2
Game PC: ASUS PG258Q 240hz, 8700K 5.1ghz OC , ASUS MAXIMUS IX APEX Z270, 2 x 8GB 4200Mhz 16-16-16-36, Nvidia 1080FE
Work PC: Dell 2716DG 144hz, 6850K 4.1ghz OC, Asus Rampage Extreme X Motherboard, 4 x 8GB 3200mhz 15-15-15-35 , Nvidia 1050TI

User avatar
ko4
Posts: 126
Joined: 06 Jul 2018, 16:14

Re: [EXPERIMENT]Keyboard latency at different debounce settings

Post by ko4 » 17 Feb 2020, 23:12

Krizak wrote:
13 Feb 2020, 23:28
Any chance you could try testing a keyboard with the optical switches. Even though your testing is capturing the whole chain and is subject to some error, would still be interested in seeing the results. The following are 2 keyboards I know that use optical switches even though slightly different technologies:

Bloody B930/B970
https://www.amazon.com/Tenkeyless-Keybo ... 02&sr=8-11
https://www.amazon.com/Keyboard-Bloody- ... 446&sr=8-3

Wooting One/Two
https://www.amazon.com/WOOTING-Linea%E2 ... 484&sr=8-2
Hey,

I was going to purchase the Bloody B930 because I don't like how the romer-G's feel and linears are my favorite
I did test out this 50$ keyboard iGK61p that uses optical Gateron switches
here are the results
igk61p.PNG
igk61p.PNG (2.89 KiB) Viewed 21577 times
Relatively high despite being optical.
It's a fairly cheap keyboard so I suspect they never optimized the electronics for lower latency

Looks like i'm going to try the B930 next

as for the Wooting boards
somebody already tested their latency
https://twitter.com/sunjun_kim/status/9 ... 7173860352

I think the wooting is on par with the logitech g pro

User avatar
ko4
Posts: 126
Joined: 06 Jul 2018, 16:14

Re: [EXPERIMENT]Keyboard latency at different debounce settings

Post by ko4 » 26 Feb 2020, 15:06

just tested the Bloody B930 at around 11ms
My test isn't accurate enough to find the winner between the logitech G pro and B930 but the linear switches on the B930 are a big step up from the Romer-G's. They are really smooth and light which might make the B930 feel faster even though the object latency is around the same

User avatar
revonoc
Posts: 2
Joined: 05 May 2020, 12:28

Re: [EXPERIMENT]Keyboard latency at different debounce settings

Post by revonoc » 05 May 2020, 12:30

Which switch is on your Ducky? I feel like actuation distance would add a few ms to this. wondering as I'm trying to decide whether to buy one!

User avatar
ko4
Posts: 126
Joined: 06 Jul 2018, 16:14

Re: [EXPERIMENT]Keyboard latency at different debounce settings

Post by ko4 » 14 May 2020, 00:29

revonoc wrote:
05 May 2020, 12:30
Which switch is on your Ducky? I feel like actuation distance would add a few ms to this. wondering as I'm trying to decide whether to buy one!
I had cherry mx reds on the ducky. I returned it because of the crazy spacebar wobble
I'm using a k65 rapidfire right now. Low input lag and silver switches. Had to get used to the low actuation force though

MegaWatt
Posts: 21
Joined: 20 Apr 2020, 09:35

Re: [EXPERIMENT]Keyboard latency at different debounce settings

Post by MegaWatt » 15 May 2020, 10:35

Interesting results, I've owned and sent back some time ago a Razor huntsman elite - optical buttons, I found surprisingly there was a debounce delay and even more surprisingly key chatter when you held the key just on the actuation point, I'd guess because the light level wasn't fully on or fully off.

So their claim to be fast to me didn't stack up at all, wasn't any better than a standard mechanical switch, maybe worse with the button chatter, perhaps they've fixed that now ? I don't know.

I've recently bought a Fnatic keyboard with cherry browns, it's definitely responsive, from the below link I've managed to get a key press down to 3ms. Their claim for no BS gaming gear seems fair, I like their gear, same for their mouse, one of be best mice I've ever used (currently I'm using a Razor death adder v2).

http://blog.seethis.link/scan-rate-estimator/

The keyboard I'm really interested in testing and / or buying is the Steel series Apex pro, I wonder if this could be any better / faster than my Fnatic board, it's not cheap.

https://steelseries.com/gaming-keyboards/apex-pro-tkl

User avatar
ko4
Posts: 126
Joined: 06 Jul 2018, 16:14

Re: [EXPERIMENT]Keyboard latency at different debounce settings

Post by ko4 » 16 May 2020, 08:55

MegaWatt wrote:
15 May 2020, 10:35
Interesting results, I've owned and sent back some time ago a Razor huntsman elite - optical buttons, I found surprisingly there was a debounce delay and even more surprisingly key chatter when you held the key just on the actuation point, I'd guess because the light level wasn't fully on or fully off.
Optical switches doesn't automatically mean low lag as seen from my igk64 testing. If the electronics aren't optimized there can still be a high amount of lag from keypress to usb packet

I did test the Bloody b930 which uses the same optical switches as the huntsman elite and its pretty quick according to my testing, which shows that they did a good job optimizing the electronics with latency in mind

Unfortunately i returned that board because although it is quick, I also faced the chatter issue you faced with the huntsman elite where there is a slight chance of the board registering a double press if the plastic part that blocks the light ends up in the "hot zone"(light level neither 0 or 1) right before the key goes back up(velocity is zero)
This can happen when light tapping the keys while typing instead of pressing it fully down.
I first noticed the chatter while playing Typeracer and confirmed it with this useful program called EK SwitchHitter that detects double presses by recording the time between keystrokes. The amount of ms before it detects a doublepress is configurable too.
MegaWatt wrote:
15 May 2020, 10:35
've recently bought a Fnatic keyboard with cherry browns, it's definitely responsive, from the below link I've managed to get a key press down to 3ms. Their claim for no BS gaming gear seems fair, I like their gear, same for their mouse, one of be best mice I've ever used (currently I'm using a Razor death adder v2).

http://blog.seethis.link/scan-rate-estimator/
http://blog.seethis.link/scan-rate-estimator/
This is useful for finding the internal scan rate of the keyboard but doesn't say anything about keypress to usb packet latency

This tool only shows the latency between the press event and release event which isn't necessarily reflective of the internal scan rate depending on the debouncing method

An example of this is the Logitech g-pro keyboard. This keyboard is incapable of registering a release event under 20ms after the press event no matter how quick you flick the switch. The lowest will always be 20ms. I think this is because their method of debouncing involves applying it only after the initial press event, but im not sure.
But regardless of that, the g pro is still one of the fastest mechs when it comes to press to usb packet latency. It's just that there will always be a minimum delay of 20ms between press and release which absolutely does not affect gameplay at all.

A better tool/program to get the internal scan rate of a keyboard is Passmarks KeyTest or https://github.com/ad8e/input-polling-test
These show the latency between press down events instead of press down and release. (you have to use multiple keys)
1ms lowest between press down events= 1000hz internal scanning
8ms lowest = 125hz internal scanning
MegaWatt wrote:
15 May 2020, 10:35
The keyboard I'm really interested in testing and / or buying is the Steel series Apex pro, I wonder if this could be any better / faster than my Fnatic board, it's not cheap.

https://steelseries.com/gaming-keyboards/apex-pro-tkl
Yea the hall effect switches are definitely interesting
https://steelseries.com/blog/worlds-fas ... switch-105
this blogpost seems to show that they might have optimized their electronics for low latency as well but it might also just be marketing BS
if im not mistaken Wooting is also in the process of developing their own hall effect switch

Post Reply