Wooting One keyboard tested against Logitech G Pro Keyboard

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.
dhaine
Posts: 49
Joined: 22 Nov 2014, 00:25

Re: Wooting One keyboard tested against Logitech G Pro Keybo

Post by dhaine » 18 Jul 2019, 05:49

The SteelSeries Apex Pro is interesting with it's customizable switch going as low as 0.4 mm actuation point, there's also the razer huntsman with a laser system which "seems" better than older keyboard with similar tech like Bloody keyboards

Morkai
Posts: 10
Joined: 23 Jan 2014, 15:15

Re: Wooting One keyboard tested against Logitech G Pro Keybo

Post by Morkai » 18 Jul 2019, 18:21

Yes, the apex pro does seem like it could be good if their info is true.
Unlike most keyboard manufacturers they actually describe the input functionality of the electronics and not just speed of the switch. https://steelseries.com/blog/worlds-fas ... switch-105

"With it's analog input, it only requires 1 scan of the keys to gather an accurate reading. No more multiple scans due to debounce.
A full scan of all keys happens every 700 millionths of a second.
The analog to digital converter is located directly on the keyboard's main chip (instead of the switches), so there is virtually no delay in converting the signal.

These factors add up to a never-before-seen 0.7ms response time"

Could try one when available actually, wooting sounds good too, but unavailable. Would really like a low profile keyboard though, but doesn't seem like anything really good exists. Maybe the cm sk630, but no info on latency for that anywhere really.

ad8e
Posts: 68
Joined: 18 Sep 2018, 00:29

Re: Wooting One keyboard tested against Logitech G Pro Keybo

Post by ad8e » 19 Jul 2019, 03:58

The Hall Effect keys in the Apex Pro should theoretically be better than Flaretech switches, and Wooting is exploring the same style (Lekker keys, to be released in 2020 or maybe never). Even then, their Lekker keyboard is planned in full size, and I don't want to make space for a full size keyboard. I'll probably wait six months after release for other people to receive their keyboards and iron out the bugs. That delays any exposure I'll have to their Lekker keys to at least 2021.

I would really like an analog keyboard that works; that would change my approach to music. The Apex Pro's keys support it but maybe their electronics/firmware doesn't. If it changes, I'll consider buying one.

EDIT: I just noticed the Apex Pro TKL removed their Print Screen button in favor of a mini display??? I guess I can afford a keyboard that's only good for one thing (music), but it makes me a lot less happy about it.

MrPleasant_exe
Posts: 1
Joined: 23 Jul 2019, 06:22

Re: Wooting One keyboard tested against Logitech G Pro Keybo

Post by MrPleasant_exe » 23 Jul 2019, 06:42

From what i have read, their customer support is great, they reply quickly and you can even submit ideas and suggestions. They also seem to be working on and constantly improving their software called Wootility. Also, you could probably even invite their designer and engineer to talk about input lag measurements here on the forums.

https://www.reddit.com/r/WootingKB/
https://trello.com/b/NpKEdAgB/wooting-roadmap
contact: [email protected]
Appreciate the love here ❤️, we try our best to help provide answers and fix issues whenever we can on discord, Reddit etc and when we get compliments, it really brightens our day. Also thank you for linking the subreddit. The discord is open to anyone also if you want to ask any questions there! https://discord.io/wooting/

-WootHelp

And yes, i totally just made this account to say thanks haha

xmysticgohanx
Posts: 4
Joined: 26 Jul 2019, 01:39

Re: Wooting One keyboard tested against Logitech G Pro Keybo

Post by xmysticgohanx » 27 Jul 2019, 18:15

Xtrfy claims their response time is 0.001: https://xtrfy.com/keyboards/k2-rgb-white-edition/

ad8e
Posts: 68
Joined: 18 Sep 2018, 00:29

Re: Wooting One keyboard tested against Logitech G Pro Keybo

Post by ad8e » 30 Jul 2019, 00:30

I changed my mind about the Wooting One. I was noticing some issues, so I benchmarked it to get data. The Wooting One's electronics are not good. My Logitech G-Pro's electronics are better, although its binding wobble is still a major physical problem.

Data: https://pastebin.com/qxVpgm9W

Test method: I press two keys simultaneously, D and F. Each line shows the key pressed and the time since the last keypress. The large numbers separate simultaneous presses, and the small numbers are within simultaneous presses.

Notice that on the Logitech G-Pro, I get a good spread of values: 0 ms, 1 ms, 2 ms. After a while, my fingers drift out of sync and reach 4-6 ms, but it's still perfectly consistent so you can tell it's finger drift, not the electronics' fault. After I notice it, I correct and things go back to 0-2 ms.

On the Wooting One, the consistency is completely gone. The 1 ms timepoint is not achieved, only the 0 ms and 1.6 ms - 2 ms timepoints. After that comes sudden jumps to 8 ms. This is on Tachyon mode with RGB off and Xinput off, their highest performance mode. It looks like their electronics becomes clogged every so often. They're unable to achieve a 1000 Hz report rate.

For my Wooting One failing to work after wakeup, that was not fixed by the devs and it does not look like they will fix it either. It is said to be BIOS-specific, although my laptop is quite common (Lenovo Thinkpad T-series). I can work around it by waking up my system through some other device than my keyboard. It is an annoying problem but a livable one.

As for Xtrfy's claim, they are straight up lying. That means they are not trustworthy in their other claims either. I wouldn't bother with them.

Right now, the keyboards of most interest are the Apex Pro and the Bloody B975.

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

Re: Wooting One keyboard tested against Logitech G Pro Keybo

Post by Chief Blur Buster » 30 Jul 2019, 16:12

Interesting about simultaneous-key consistency -- that is very important for many gamers.

Multiple methods of keyboard benchmarking could be designed -- latency, consistency, simultaneous keys, etc.

Does any websites currently do this?
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!

xmysticgohanx
Posts: 4
Joined: 26 Jul 2019, 01:39

Re: Wooting One keyboard tested against Logitech G Pro Keybo

Post by xmysticgohanx » 30 Jul 2019, 20:55

ad8e wrote:As for Xtrfy's claim, they are straight up lying. That means they are not trustworthy in their other claims either. I wouldn't bother with them.

Right now, the keyboards of most interest are the Apex Pro and the Bloody B975.
How do you know they are lying?

Do you think it’s likely that the apex pro and b975 (and their tkl version b930) will outperform the gpro?

Rocky4
Posts: 2
Joined: 30 Jul 2019, 19:31

Re: Wooting One keyboard tested against Logitech G Pro Keybo

Post by Rocky4 » 30 Jul 2019, 20:56

Unfortunately the whole test is useless. You aren't measuring the input speed of a device, you are only measuring the lag and inconsistence of a system in combination with human error.


I can proof my claim quit easily.

Proof 1:
It's impossible to have a lower delay than around 1 ms (without exactly 0 ms), because the fastest polling rate for USB device is at 1 ms and I don't think you had overclock your USB controller. So the difference between two inputs from the same device must be exactly 0 ms, when both inputs are in the same package, or it must be around multiple times of the polling rate of the device. For example when the device has a polling rate of 8 ms, the delay between two inputs which aren't in the same package must be at around 8, 16, 24, 32... ms and so on. But if you see a value which is in between like 12 ms you know that your system has added a delay which is the difference between this value and the previous timeslot from it, which is 4 ms in this example. The reason for this delay, which the system is causing, is that the CPU needs to poll the device, but it can be that he needs to perform more important task first.

Proof 2:
Just by analysing your code I can tell that your program isn't suitable for testing inputs exactly. Your program is only using one thread there the measuring of the inputs isn't really even decoupled from the printing of the outputs, but this alone isn't a big deal. The printing of the outputs to the screen takes a very long time and is very inconstant. Because of this, the results from the test method one was always inconsistant, because the output was done after every key event. But even with the test method two, you will add everytime you print out the results from the previous measurements, the time it takes for it to the then last value. This will happen for sure everytime the inputs are handled. You must clear the entire list after handling the results, instead of always keeping the last key. But then you must handle it blockwise, to be able to even calculate something. Because if you clear the entire list it can be happen that you never have two values to compare.

Proof 3:
Use a mouse which is able to really perform a 1 kHz polling rate and compare the values of your program with the RAW USB values. I using my Logitech G903 and Wireshark to capture the RAW USB data. I can see that my mouse sends a package constantly every rounded millisecond to my PC, but the results of your program are very different than the data I can see with Wireshark. Because Wireshark captures the data within the USB controler and so, on the nearest possible point, the system side has a lower inpack to adulterate the values.


In other words, the only nearly true test for measuring the input speed with software is to have a look on the RAW USB data. Everything else is not even worth to try it out. To really test the input speed you would need an oscilloscope, a logicanalyzer and a dedicated fast test controller as the host for the device which is to test.


@Edit:
With this you are only able to test the rate how fast the device is able to handle separated distinct inputs. This wont tell you the delay how log it takes for the device to register an input and to inform the system of it.


@Edit 2:
Next, because how your program works, it disadvantaged devices which really send values very fast and preferred devices which aren't. A modern gaming keyboard uses bitflags to indicate the pressed keys, where every key has its own state to determine if the key is pressed or not, this is called nKRO and the Wooting is such a device. Because of this it's impossible to tell which key was pressed first, if two keydowns are registered in the same package the first time. An old 6KRO keyboard works differently, becaue the pressed keys are listed in a query of 6 keys (without the modifiers) the PC knows which key was pressed first and keyboards with this type are normally slow.
The Logitech Pro claims to be a 26KRO, so I would assume that this device is more like a 6KRO with just a larger query and with a much slower scan rate than the Wooting, but because the pressed keys aren't handled altogether at the same time and your PC instead worked through the query from one single package, one key after another, you misleadingly think that the signals are sended through different packages.


Subnote:
This is my own opinion from a technical standpoint only. Nonetheless I own a Wooting One and I love it so much that I helped them as a volunteer for quiet a long time now.
The keyboard isn't perfect and it still has some bugs, like the problem with the sleep mode on some systems. Unfortunately it takes a long time for them to add new feature and to address minor bugs.

ad8e
Posts: 68
Joined: 18 Sep 2018, 00:29

Re: Wooting One keyboard tested against Logitech G Pro Keybo

Post by ad8e » 31 Jul 2019, 00:25

xmysticgohanx wrote:How do you know they are lying?

Do you think it’s likely that the apex pro and b975 (and their tkl version b930) will outperform the gpro?
We know they're lying because there's a minimum latency of 0.5 ms caused by the USB polling, which can be taken downwards using sketchy techniques but still won't reach their values. Their claim isn't physically possible. Then, thinking of all the other parts of their system which I find unbelievable to reach that latency, and which will all add up in a latency chain, the only logical conclusion is that they are talking nonsense.

In terms of pure latency, I think it's pretty likely the Apex Pro and B975 will outperform the G Pro. In latency, the Apex Pro will likely outperform the Wooting by a tiny margin, and the B975 will underperform the Wooting by a tiny margin. This is just electronics though, and physical problems can be big.
Rocky4 wrote:
I won't take the time to address your points in detail, because it doesn't seem your analysis is good enough to be worthy of that.

Your Proof 1 has no relationship.

Your Proof 2 is wrong.

Your Proof 3 is "wrong"; we are testing keyboard-to-computer latency, and I already measured the error of my program, which is why I put the Logitech keyboard in there for comparison, so that other people can see the (lack of) error too. We've benchmarked my program's accuracy on >20 widely varying machines and USB devices, so I have a very good idea of the error my program introduces, which is excellent. It's true that oscilloscopes will be able to capture things that the USB won't. For example, it would be great if someone could oscilloscope the Wooting, since its 1.6 ms - 2.0 ms timepoint is inconsistent and the USB polling rate is hiding whatever is happening there.

Your edits are wrong, but you should read the OP in this thread. The point of the program is to measure inconsistency; the latency is in the first post.

Most of the early Wooting bugs have been ironed out over time. I know that other people experience a different subset of bugs than I do, but at least the devs seem aware of the various bugs, just prioritize different things. I think that they still care, which lends hope that these lower-priority bugs will eventually be resolved after their product launches settle down, maybe after a year.

Post Reply