Author Topic: I'm from Logitech: What to use to test a Gaming keybaord  (Read 4628 times)

0 Members and 1 Guest are viewing this topic.

Offline LogiSly

  • Thread Starter
  • Posts: 2
I'm from Logitech: What to use to test a Gaming keybaord
« on: Tue, 10 January 2017, 10:51:49 »
Hello,

I'm a designer from Logitech. I have a question to the Gamers community :)

What would you do/use to compare different Gaming (mechanical) keyboards performance, on top of doing a typing test. Is there an online Game or any sort of performance online test Gamers are using to test and compare between each others how good or bad is their keyboards? There are plenty of typing tests available online but, but what about gaming test, where may be you use the WASD and space keys to see if the keyboard is good for gaming?

Thanks for answer!
LogiSly

Offline Folio

  • Posts: 214
  • Location: Chicago, IL
  • I'm a nice guy.
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #1 on: Tue, 10 January 2017, 11:30:52 »
I don't really use or own any "gaming" keyboards per se. However, I do use my mech boards to game on because they are very comfortable. Cherry MX Clear switches, imo, are optimal for me to game on. The board at that point should be heavy enough to give a good clack on the downstroke.

In terms of games to try new boards on, I always play my favorite twitch fps games. For example, Rainbow Six Siege.

Offline Sifo

  • Swaglord Yolomancer
  • * Exquisite Elder
  • Posts: 6109
  • Location: #GOLDSPRINGS, #LEGITBALLIN
  • beekeeper
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #2 on: Tue, 10 January 2017, 11:38:24 »
I mean... it doesn't really matter, it depends on the genre of game you play as well as your personal preference. That's what we have different switches for.

A lot of the key points of determining a quality keyboard can't be judged immediately through something like a typing test. You can get an idea of whether you like the switches or not, and maybe judge how "solid" the construction is, but other things like durability of the keycaps and quality control of parts, those things you can only really tell by using it for an extended period of time.

As an enthusiast, I'd probably open it up to take a look at how it's built. As a gamer? I'd go and play whatever game that I happen to play every day and see if I still like it after a week or two.

You could easily go about playing an FPS game with blues for about a month but then realize you hate it and want to try reds, and then realize that you make too many typos with reds so you try browns. Or conversely you could realize immediately upon taking it home that blues just aren't for you because they're too loud. And maybe you realize that your friend's different model keyboard with browns feels way better than the model you bought. Or maybe you realized you want to join the Topre circlejerk. I don't think there are many good methods for "first impressions" of a keyboard other than a typing test, and even then it won't tell you anything about the quality of the keyboard itself.


 
(Forever) Illustrious.

Offline Rayoui

  • Posts: 260
  • Location: Portland, OR
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #3 on: Tue, 10 January 2017, 11:43:53 »
I find that the "best" keyboard to use for gaming tends to be a subjective thing. You can make switches that have gimmicks like higher actuation points and less travel but at the end of the day, people just end up choosing a keyboard that has the tactile feel that they most prefer. The real-world performance difference between the switches is very small and I don't think it really impacts gaming much.

I wish that manufacturers making all these fancy new proprietary switches would focus more on making the switches feel better (i.e. smoother, more tactile). Including a nice set of thick keycaps would be a plus too, especially if the stems are incompatible with most keysets.
Clueboard  |  B.face  |  HHKB Type-S  |  Realforce RGB  |  Pok3r
Soon: MIRA SE

Online FoxWolf1

  • Posts: 835
  • 154
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #4 on: Tue, 10 January 2017, 12:22:36 »
When I get a new peripheral, I load up an Urban Terror jump-map and then see what kind of times I can set over a few runs of the course. That way, I have some numbers to look at, and not just a subjective impression. Then, once I have the numbers, I try and flesh out the subjective impression by switching back to whatever I was using previously for a run, and then back to the new keyboard.

The intro procedure also involves AquaKeyTest testing to ensure that I have functionality on all keys and am getting the advertised rollover, a typing test, and testing each key for consistency of sound and feel. Once I've done those things, then I'll load up various time-killer games and see how the keyboard feels there. After that, I'll use it normally for a while, but paying a bit more attention to my keyboard than usual until my impression is solidified.
Yeah, okay then.

Offline keytohopiness

  • Posts: 37
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #5 on: Tue, 10 January 2017, 12:30:08 »
I look for something that feels better than Omrons.

Offline chyros

  • a.k.a. Thomas
  • * Esteemed Elder
  • Posts: 2611
  • Location: The Netherlands
  • Hello and welcome.
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #6 on: Tue, 10 January 2017, 12:49:39 »
Ha, I wrote a review about a Logitech keyboard for Tom's Hardware a bit ago, don't know if they released it yet :) .
Check my keyboard video reviews:


Offline Hoaryhag

  • Posts: 14
  • Location: SW Missouri
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #7 on: Tue, 10 January 2017, 13:19:00 »
In today's market, a board that you can hot-swap cherry mx compatible switches, available in full size and tenkeyless layout at least.  Nice heavy keycaps with an attractive font for once and not all friggin black.  I don't wanna pay for a keyboard I'm gonna have to change the keys on right away and looks like it was hit with an ugly stick.  I would say most of the gamer marketed keyboards are flat out nasty looking.  Preference on switch feel is a personal thing.  Make it hot swap with your choice of switches installed and you cover all cherry mx preferences.
« Last Edit: Tue, 10 January 2017, 13:29:18 by Hoaryhag »
 

Online tp4tissue

  • * Destiny Supporter
  • Posts: 10821
  • Location: Official Geekhack Public Defender..
  • OmniExpert of: Rice, Top-Ramen, Ergodox, n Females
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #8 on: Tue, 10 January 2017, 15:34:35 »
Yo logisly..  I sent you a Private Message, it's at the top of the screen..- click message-


Offline klennkellon

  • Posts: 1168
  • Location: Southern California
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #9 on: Tue, 10 January 2017, 15:47:21 »
To test my keyboards roll-over ability I use this little tool from Microsoft. https://www.microsoft.com/appliedsciences/KeyboardGhostingDemo.mspx

But honestly for gaming all that really matters is that the switch is not too heavy or too light or easy to actuate.

Some gamers also really like to customize their keycaps and when the keyboard uses a non-standard bottom row like many Logitech boards it's what keeps me from recommending Logitech boards to other people.
Packard Bell KCLEA907L Alps SKCL Brown

KBParadise V80 Matias QC

IBM Model M 1391401

Keycool 87 PBT

Cherry are good Alps is just nicer

Online chuckdee

  • * Destiny Supporter
  • Posts: 842
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #10 on: Tue, 10 January 2017, 20:59:26 »
Honestly?  As far a Logitech specific, you make very good, but very focused products.  Too focused.

I think that klennkellon said something you should take to heart, i.e. for gaming all that really matters is that the switch is not too heavy or too light or easy to actuate.

That means different things to different people.  Instead of trying to reinvent one switch, why not give people different switches for different tastes?
Pok3r Mx Browns | Pok3r Mx Clears | Magicforce 68 Mx Browns | JD45 Mx Clears | Grid Planck Mx Browns | JD45 Mx Browns | miniVan Gateron Browns

Currently Supporting:

Offline supamesican

  • Posts: 188
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #11 on: Wed, 11 January 2017, 08:03:05 »
Define good for gaming? Good feel, responce time, good key roll over/blocking/anti ghosting, good sound(click)?

I usually start up a bot match in quake or ut to test if I like the way the wasd cluster feels when im moving between then fast. RTS stuff to test keymapping and a tiny bit of response. GTA 5 in a car horn blaring and driving to see if the a and d are blocked when holding w and e too. Click sound, I just type. Most importantly THICK PBT CAPS FOR THE KEYS ARE A MUST. And preferably for me something not a cherry mx or clone. I want to do different things those got boring for me years ago.
« Last Edit: Wed, 11 January 2017, 08:08:30 by supamesican »

Offline quasistellar

  • Posts: 167
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #12 on: Wed, 11 January 2017, 10:53:23 »
I would test on MOBAs and a fast FPS like DOOM or Overwatch.

Gamers typically smash their keys when they play from my experience, so the bottom-out feel should be good, and the overall key-travel should be around 3.5mm or maybe even a bit less.  The actuation (and tactile point) should be higher in the stroke, but not too high.  Between 1.5-2mm.  You should have both tactile and linear versions of a switch if possible, and shoot for an "alps-like" feel in the tactility, as opposed to a "cherry-like" feel.

The keyboard SIZE is important also.  Desk real-estate is at a premium, so please keep the size to just what is necessary, and if possible make the keyboards thinner.  The G810 is a fantastic step in the right direction.  The g410 is STUPIDLY oversized just for "style" and makes me keep my mouse way farther away than is needed.

The caps should be HIGH QUALITY.  That is the number one thing holding back almost all gaming keyboards for me.  They almost all have crappy ABS keycaps.  ABS in and of itself isn't bad, but you need to make them high quality.  Nice and thick doubleshots.  PBT would be great but not necessary.

Your current offering of Romer-G is good, but could be better.  The actuation point and overall key travel is pretty good.  The tactility leaves a bit to be desired, as it is very similar to Cherry MX Brown.

So, TL;DR:

Make new, silenced Orange Alps, with good keycaps (or MX-keycap compatibility), please  :p

Offline Findecanor

  • Posts: 3558
  • Location: Stockholm, Sweden
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #13 on: Wed, 11 January 2017, 11:25:31 »
People on gaming forums can sometimes be a bit obsessed about latency.. but I think that is overrated. Just try to engineer the firmware well. Latency is a combination of many factors.

N-key rollover can be achieved with Mac, Windows and Linux and work with crappy BIOS:es but many gaming-keyboard engineers don't know it, or they don't know how. There are various bugs in each of those operating systems and in many BIOS'es. Programmers in this community (Soarer, Hasu, Haata) have detected those and worked around them. Your programmers should check out the report descriptors in  Haata's Kiibohd firmware source code for an example that works.

The other things that I see people asking for the most in forums are:
* That the keyboard is stylish and does not look too gamer-y. I too think you did a great job with the aesthetics of the G810 and personally I appreciate that the backlighting does not bleed in-between the keys. I also see requests now and then for high-quality mechanical keyboards that are white instead of black.
* That the keyboard is not too loud. Noise does make something appear solid, but too much noise can make it appear clunky. But also, I often read that the gamer's girlfriend has complained about the noise (not that there aren't gamers who are girls, who have boyfriends... but anyway). Noise depends not only on the switch but also on the overall build of the keyboard.

As has been said before, switch type and key profile is a matter of personal preference. Best might be to give users options.
« Last Edit: Thu, 12 January 2017, 08:43:55 by Findecanor »
"It is no measure of health to be well adjusted to a profoundly sick society."
Daily driver: Phantom (Lubed Cherry MX Clear, Lasered Cherry PBT keycaps with Row A. Plastic "Frankencase". Custom firmware, Swedish layout)

Online OfTheWild

  • Formerly boostdemon
  • Posts: 786
  • Location: Cary, NC
  • Make things. Have fun.
    • Studios of the Wild
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #14 on: Wed, 11 January 2017, 12:33:39 »
Pretty much all the technical stuff has been said above.
As for design ideas however, I would love a gaming keyboard with something like mx reds or browns in a hot-swappable board (sip sockets) so i could change out the switches if needed or do something silly like put blacks in the wasd and reds in the alphas, browns at the shft keys and clears for the rest of the modifiers. Oh and I need a production keyboard that has better keycaps. Maybe even a set of rubberized grippy wasd caps if the user wants. I would have skipped all the fancy RGB distraction and design for just a well made board from solid materials and quality parts. So many gaming keyboards look like they're from a Giger drawing and lack the quality of a simple setup. As for added functionality for gaming, I really liked the macro keys of the Corsair K95 but they were in the way the rest of the time and honestly i never used them all. What would be really cool would be a programmable function layer where i can add my macros and assign them to keys for easy in-game layout changes.   

Just some thoughts.
-Dana

Doorbell

Online CJD_Anthony

  • Posts: 378
  • Location: US
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #15 on: Wed, 11 January 2017, 13:58:25 »
Well designed... not over the top gaudy looking cases..
Customizability (replaceable case, plate/pcb, etc)

A lot of people seem to value conservative types of designs with sleek edges (duck, otd, lz, etc....)

Literally have some logitech team members lurk geekhack/reddit/deskthority for a few weeks/months and youll see what enthusiast grade keyboards look like and feature. It's a lot deeper than just "what switches do you like?". Build quality, weight, plate types, keycap material and thickness all contribute to a comfortable typing experience.

Online kawasaki161

  • Posts: 905
  • Location: Germany
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #16 on: Wed, 11 January 2017, 14:43:31 »
Just go through the standard checklist:
-NKRO
-1000hz polling
-Latency not worse than the competitors'. I mean the latency on the PCB level not on the switch-itslef level, mainstream gamers might care about that 0.2mm higher actuation because it looks good on paper, but in the end it doesn't affect anything when the time from forming an electrical connection to having a keycode arrive at the PC is 20ms higher than everyone elses.

Now that we have a base that is at least as good as the competition, let's make a good keyboard:

-Don't go "ub3r1337g4m3r" design wise, look at the MasterKeys series from CM to get an inspiration on how to make a nice and minimalist keyboard
-Standard Layout and MX Keycap compatibility, preferably stock thick PBT caps. On your Romer G boards where MX compatibility is impossible at least use some standard profile caps made from thick PBT with a neutral font.
-Go Gateron instead of Cherry, just make a few engineering samples with Gateron switches (they are the same size as Cherry MX, so no redesigning) and let people test them next to the Cherry counterparts, they'll love it
-Make a software-less mode in case there is none
-Use thick plastic and make the board heavy, just add weights if necessary
-Detachable cable and cable routing options
-Hotswappable Switches, many people just want to go the safe route with Reds, why not give them the opportunity to try out every type for a lot less than buying a full keyboard each?
-No other gimmicks, RGB is okay, just don't add a non detachable wrist rest or a screen or something silly like that


But wait, you might have realized that this isn't gaming specified stuff. That is because as soon as the first 3 points are ticked you have everything you need for gaming if a keyboard passes a typing test and offers the three points mentioned at the beginning it's perfectly suitable for gaming. In fact I would honestly be surprised if you made a keyboard that offers that and passes the typing test, but is bad for gaming.
Just look at what is currently available, the average gamers don't have high standards, they just have deep enough pockets to buy what everyone else buys and to pay for a brand name rather than a product.

Offline csmertx

  • * inserts coffee
  • * Esteemed Elder
  • Posts: 2063
  • Location: Orlando, FL
    • Instagram | csmertx
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #17 on: Wed, 11 January 2017, 16:18:59 »
I test with the original Quake. That was the first PC game I played that got me hooked so I'll always go back to it to test out whatever. Quake 3 or Quake Live from what I understand are fantastic games to test out new keyboards and mice.

Still, most of what I play requires a decent mouse. I'm more concerned of smushing the plastic of a thin ABS keycap, wiping off legends, or busting a frail switch. The only problem I've had with gaming and keyboards was when Windows enabled sticky keys by default (ME or XP?). I still don't understand the benefits of NKRO :confused:

Offline falkentyne

  • Posts: 283
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #18 on: Wed, 11 January 2017, 22:47:56 »
Elite keyboards' switchhitter is the best tool to test for rollover and also for chatter (adjustable).  I prefer that over any other programs or tools.

Offline Data

  • Posts: 2393
  • Location: Orlando, FL
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #19 on: Thu, 12 January 2017, 06:44:47 »
There's not one specific game that applies to all gamers.  That's kind of like asking an audiophile community what one specific song they all use to test their HiFi systems, or asking the import tuner community what stretch of road they all use to test their latest upgrades.  It doesn't really work that way.

Every gamer is going to test on whatever game is relevant to him/her at that given moment, and that probably includes multiple games for most gamers.  It's highly subjective and situational.  Frankly, I don't consider a game to be a very good test for a keyboard -- and I consider myself a gamer.  Obviously you want your keyboard to perform well in whatever game(s) you play, but most of us expect more from the device than what we can get out of it in a typical play session.  In other words, it needs to do a lot more than just play games well.

Right now, I would probably test using Fallout 4, Doom, and Black Desert Online.  Those are the games I happen to be playing.

Offline LogiSly

  • Thread Starter
  • Posts: 2
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #20 on: Thu, 12 January 2017, 08:42:15 »
Hi All,

Thank you for your replies, I read them all.
It sounds like most people will use their preferred game to test a keyboard, makes complete sense! But this is not something I can run if I do a keyboard lab testing.
Thanks!

Offline chyros

  • a.k.a. Thomas
  • * Esteemed Elder
  • Posts: 2611
  • Location: The Netherlands
  • Hello and welcome.
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #21 on: Thu, 12 January 2017, 08:57:21 »
Hi All,

Thank you for your replies, I read them all.
It sounds like most people will use their preferred game to test a keyboard, makes complete sense! But this is not something I can run if I do a keyboard lab testing.
Thanks!
I use Switch Hitter from Elitekeyboards to test keyboards for Tom's Hardware reviews. I also use AquaKeys, but it does the same thing.
Check my keyboard video reviews:


Offline davkol

  • Posts: 4710
  • Location: CZ
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #22 on: Thu, 12 January 2017, 09:13:18 »
I have yet to see a *valid* (and reproducible) demonstration of any meaningful benefits from a "gaming" keyboard (or a keyboard used for gaming). Other than reliability problems. And obviously: standard keyboard layouts and shapes suck.

I mean, what can be expected from a keyboard? Reliability and ergonomics.
  • reliability: low, consistent latency on all levels from switches to connection; antighosting, a proper NKRO implementation; consistent switches/stabilizers without sticking/binding/hysteresis problems; bug-free firmware; longevity of the hardware itself
  • ergonomics: layout/shape, keycap shape/texture, key stiffness and tactility, noise levels/frequency
The latter requires (1) input from medical background and (2) user testing with proper methodology.

Offline Data

  • Posts: 2393
  • Location: Orlando, FL
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #23 on: Sat, 14 January 2017, 11:42:18 »
Hi All,

Thank you for your replies, I read them all.
It sounds like most people will use their preferred game to test a keyboard, makes complete sense! But this is not something I can run if I do a keyboard lab testing.
Thanks!

The practical lab tests you and other manufacturers are already doing -- switch longevity, actuation force, travel distance, audible feedback, etc -- those values are already important to gamers.  So I'd say you guys are probably doing it right, and to keep it up.  Feel free to print those results on the box.  :D  People are watching.  Gamers pay attention to that stuff too.

As a gamer, if I saw something like "Top performer in a head-to-head Call of Duty marathon shootout with 9 of its competitors" printed on the box I'd probably roll my eyes and quickly put it back on the shelf.  (1) That's not a data point that's particularly relevant to me (I don't play Call of Duty games), and (2) it screams "desperate marketing gimmick".  :P

Offline MperorM

  • Posts: 24
  • Location: Bagsværd - Denmark
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #24 on: Sat, 14 January 2017, 12:09:57 »
As someone who's played starcraft on a semi proffessional level, the most important factor for me is by far the form factor. Every logitech keyboard I've ever seen has been freaking huge. Having your mouse hand close to the keyboard hand removes a ton of strain that you get in your right hand from playing many hours every day without breaks. Even tenkeyless is too big for me, sadly I have to settle with that because 60 % removes important keys such as escape and the f-row. As a result, I've not even considered using a logitech board. Even the tenkeyless board is large compared to other tenkeyless boards.

I would love to see some experimentation with putting more keys on the left of the keyboard. Personally when I play I wish I could have more keys on the left of my board, since I definitely could make use of a few extra keys that are easy to reach. Every time I've seen boards with macro keys, they've always been placed fairly far to the left. I think macro keys would be much more useful if they were right next to the leftermost layer.

Regarding switches, I think there is no right answer. FPS gamers usually prefer linear switches, since it's easier to make miniscule adjustments with them, where I as an RTS gamer like my keyboard loud and clicky, as it is satisfying to hear the symphony of keys when you are playing with 300 actions per minue :)

Something I'd also love to see is a usb port or two in the keyboard. I don't own mouse bungee, and most people I know don't either. A usb port in the keyboard allows me to keep the cable on the table which reduces very annoying mousedrag.
Leopold fc980c with korean keycaps

KBParadise v80 with blank and original keycaps - FOR SALE

Offline falkentyne

  • Posts: 283
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #25 on: Sat, 14 January 2017, 13:28:16 »
Logitech should DEFINITELY consider making TKL gaming keyboards, as well as some Cherry MX RGB TKL keyboards.  Just make sure if you use MX BLUE Switches, that you do proper QA on the switches, as there have been some serious inconsistency issues with MX RGB blues that were far worse than the non RGB blues (and similar MX greens).  The only non RGB switch that was as flaky as the RGB MX blues were the "Milk white" MX clicky switches.  MX Browns were also on the bad side when you first released the G710+ (which all had really GOOD MX browns) as Logitech buying up all of Cherry's stock caused a worldwide shortage of MX brown keyboards, and then a bunch of "scratchy" no bump browns started being released (like on the early Ducky shine 3's, etc).

Stay away from 60% keyboards and 40% keyboards are a joke.  Most gamers don't want them, and we need function keys and arrow keys.  If people want a 60% keyboard they can just go buy a Vortex.  60% keyboards are only useful for carrying around in laptop bags.  TKL keyboards would be really appreciated, though.

Offline davkol

  • Posts: 4710
  • Location: CZ
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #26 on: Sat, 14 January 2017, 14:13:49 »
Even tenkeyless is too big for me, sadly I have to settle with that because 60 % removes important keys such as escape and the f-row.
No, you don't. There's a variety of form factors between 60% and tenkeyless. If you want pointers to specific products, see Noppoo/PLUM Nano75 (60% + function row) and 84key keyboards, such as Noppoo Choc Mini, PLUM 84 or Keycool 84.

BTW you're assuming default key bindings, although you don't have to stick to them (and you don't even have to go as far as TheCore).

Something I'd also love to see is a usb port or two in the keyboard. I don't own mouse bungee, and most people I know don't either. A usb port in the keyboard allows me to keep the cable on the table which reduces very annoying mousedrag.
Ugh. A ****ty hub just adds cost and doesn't really improve anything. Esp. if it's on the right, and it's way too often on the right, where it gets in the way of the mouse.

Meanwhile, the tape/paperclip tricks have been around for ages.

Logitech should DEFINITELY consider making TKL gaming keyboards
G310 and G410 have been around for quite a while.

as well as some Cherry MX RGB TKL keyboards.
Why would they compete against their own product line with arguably better engineered omrons?

60% keyboards are only useful for carrying around in laptop bags.
lol, no.

Stay away from 60% keyboards and 40% keyboards are a joke.  Most gamers don't want them, and we need function keys and arrow keys.
That's only a matter of how the keyboard is designed and marketed.

See early iterations of Roccat Sova, for example. That thing's amazing. It could be even combined with a design similar to Logitech's own G13 (semi-ergonomic keypad).

Offline dr-slump

  • Posts: 35
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #27 on: Sat, 14 January 2017, 17:19:58 »
In addition to my other keyboards I have a G710+, a G810 & a G910 from Logitech.


Of all 3 I definitely prefer the G710+ because it has the best wrist rest (G810 has none) & MX-Browns (personal preference), even though I use it just at work.(too much data entry for a 60%)
The 810 looks nice & minimalistic but it has the same flaw as the G910 : You can only control the lighting via the software (execpt for on/off) & there doesn't seem to be an onboard storage for profiles or at least one profile --> The software has to be in autostart( :-X --> suckz imo )

On my vacation, (last 2 weeks in December)I have exclusively used the G810 and got a nice and painful Tenosynovitis in my right hand because of the awkward, more than shoulder wide positioning of my mouse.

Since switching back to my WhiteFox & PurePro the symptoms have gotten better & I've gained much more appreciation of a smaller
footprint keyboard.

That said, I do now prefer the Romer G switches for gaming, but they're horrible for typing (imo :thumb:)

If you'd bring following specs to market, a lot of boxes would be ticked for most people:

- TKL in the G810/G610 design but non glossy on the sides (scratches way to easy)
- a magnetic, removable wrist rest (like Cherry )
- a metal volume-wheel (like Corsair --> the rubberized ones do get dusty pretty fast and are a pain to clean)
- onboard storage for at least one profile 
- removable cable
- doubleshot keycaps


Waiting for Miami Dolch . . .

Looking for: Titainium Spacebar 6.25

       

Offline chyros

  • a.k.a. Thomas
  • * Esteemed Elder
  • Posts: 2611
  • Location: The Netherlands
  • Hello and welcome.
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #28 on: Sat, 14 January 2017, 19:00:55 »
Hi All,

Thank you for your replies, I read them all.
It sounds like most people will use their preferred game to test a keyboard, makes complete sense! But this is not something I can run if I do a keyboard lab testing.
Thanks!

The practical lab tests you and other manufacturers are already doing -- switch longevity, actuation force, travel distance, audible feedback, etc -- those values are already important to gamers.  So I'd say you guys are probably doing it right, and to keep it up.  Feel free to print those results on the box.  :D  People are watching.  Gamers pay attention to that stuff too.

As a gamer, if I saw something like "Top performer in a head-to-head Call of Duty marathon shootout with 9 of its competitors" printed on the box I'd probably roll my eyes and quickly put it back on the shelf.  (1) That's not a data point that's particularly relevant to me (I don't play Call of Duty games), and (2) it screams "desperate marketing gimmick".  :P
I have enough experience with manufacturer specs that I don't take them at face value - I test each and every one of them to make sure. Also, a lot of buzzwords that manufacturers, including Logitech, use, are absolute bull****, like "anti-ghosting". Whenever I heard that I get suspicious because it shows a clear lack of understanding of the manufacturer's own products. Sadly, virtually everyone markets vacuous things like this nowadays. 
Check my keyboard video reviews:


Offline MasterBash

  • Posts: 34
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #29 on: Sat, 14 January 2017, 21:00:43 »
I am a gamer before anything else, I play games everyday and I am quite competitive in them.

1. 1000hz
2. NKRO
3. Low latency overall

Other things that many people mentioned here that I agree with :

- TKL
- Minimalist design (I personally prefer just a plate like Corsair keyboards... those are easier to clean), but minimalist is a must.
- Modular : able to change the switches in case one breaks or if we want to try new switches without changing the entire keyboard

I don't care about RGB, but I do know gamers want that stuff. I had the G910, which I sold because it was way too big and I am in the process of selling my G410 because of the design... It takes too much space, it is not minimalist.

I wish to have more switches choice when it comes to Romer-G. I like them overall but I wish there would be a Linear option and silenced (like the Cherry MX Silent switch). My G410 is the loudest keyboard I own. Louder than the G710+ and Coolermaster Quickfire TK with Reds.

Also, since I travel a lot with my computer, I would definitely love to see a wireless solution similar to the G403/G900... Extremely low latency with good battery life.

Offline MperorM

  • Posts: 24
  • Location: Bagsværd - Denmark
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #30 on: Sun, 15 January 2017, 03:10:29 »
Even tenkeyless is too big for me, sadly I have to settle with that because 60 % removes important keys such as escape and the f-row.
No, you don't. There's a variety of form factors between 60% and tenkeyless. If you want pointers to specific products, see Noppoo/PLUM Nano75 (60% + function row) and 84key keyboards, such as Noppoo Choc Mini, PLUM 84 or Keycool 84.

BTW you're assuming default key bindings, although you don't have to stick to them (and you don't even have to go as far as TheCore).

Something I'd also love to see is a usb port or two in the keyboard. I don't own mouse bungee, and most people I know don't either. A usb port in the keyboard allows me to keep the cable on the table which reduces very annoying mousedrag.
Ugh. A ****ty hub just adds cost and doesn't really improve anything. Esp. if it's on the right, and it's way too often on the right, where it gets in the way of the mouse.

Meanwhile, the tape/paperclip tricks have been around for ages.

Logitech should DEFINITELY consider making TKL gaming keyboards
G310 and G410 have been around for quite a while.

as well as some Cherry MX RGB TKL keyboards.
Why would they compete against their own product line with arguably better engineered omrons?

60% keyboards are only useful for carrying around in laptop bags.
lol, no.

Stay away from 60% keyboards and 40% keyboards are a joke.  Most gamers don't want them, and we need function keys and arrow keys.
That's only a matter of how the keyboard is designed and marketed.

See early iterations of Roccat Sova, for example. That thing's amazing. It could be even combined with a design similar to Logitech's own G13 (semi-ergonomic keypad).

ehh, I would love to get the layouts you've described, sadly it's impossible to get factory made with alps switches. Regarding USB costs, I can understand your point. I feel like both sides of the argument is very valid. Minimizing costs is nice, but as a competitive gamer, I would love it. Especially at events and lans it can get really bad where you sometimes have to sit across someone. Here I often get a setup where I experience cable drag, which is godawful to play with. I'm not sure how much extra usb ports cost, but if it's expensive then you've won me over. The core layout takes too long for me to learn compared to the payoff I would get. If I were to start learning starcraft again, I would definitely consider it, but at this point, I wouldn't gain a lot from it. My hotkey setup is based on standard, but with significant changes.

Haven't tried any homemade cable fixes, maybe I should do that. Can you provide a link or two maybe? :)
« Last Edit: Sun, 15 January 2017, 03:12:36 by MperorM »
Leopold fc980c with korean keycaps

KBParadise v80 with blank and original keycaps - FOR SALE

Offline davkol

  • Posts: 4710
  • Location: CZ
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #31 on: Sun, 15 January 2017, 15:32:27 »
ehh, I would love to get the layouts you've described, sadly it's impossible to get factory made with alps switches.
If you mean genuine Alps, there were some Ortek or Siig MiniTouch keyboards with 75% layouts.

If you mean Matias switches, the company itself has Matias Mini * Pro keyboards in their product line. They even have a hub, if you like it that much.


The core layout takes too long for me to learn compared to the payoff I would get. If I were to start learning starcraft again, I would definitely consider it, but at this point, I wouldn't gain a lot from it. My hotkey setup is based on standard, but with significant changes.
As I wrote, you don't have to go that far. I used to play quite a bit of AoE2 and the competitive players' bindings were a nice example.

I'm not sure how much extra usb ports cost, but if it's expensive then you've won me over.
It's more complex. Complex, that's the key word.

If you want an integrated hub, you need a specifically engineered PCB (or daughterboard) & case (thus molds). That adds cost by itself. And then you have the hub itself. Unfortunately, mainstream companies often cut costs by using very cheap chips; I can't think of a specific example off the top of my head, but it's the same about so-called audio extensions in keyboards—it's typically an extremely cheap sound card (I remember testing some gaming Logitech, probably G11, and seeing a C-Tech chip, that you could buy for $2.5 as an USB soundcard on eBay).

At last but not least, there's the cable, which is typically thick and inflexible... sometimes there are even two cables to deliver more power. Which leads to another point: power requirements. Such a feature-rich keyboard might have power requirement, that make it unusable with some hubs or portable devices (e.g., tablets).

Getting a little, decent hub of your choice makes things much easier overall.

Haven't tried any homemade cable fixes, maybe I should do that. Can you provide a link or two maybe? :)
Actually, I use either trackballs, or Roccat Apuri (hub+bungee) above a tray in my setups, thus I don't need those "ghetto solutions, but I'm sure you find some at forums. I've just googled "ghetto mouse bungee tape" and a bunch of results came up (e.g., this).

Offline lowkey

  • Posts: 44
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #32 on: Mon, 16 January 2017, 03:45:09 »
Speaking as someone who played quakeworld, q3, and cpma for years, where a complex movement system requires extremely precise inputs, here's what's important that I think a lot of OEMs seem to miss:

1. Bottom out feel and general studiness.  We press down hard, we hold the key, and we alternate those presses.  A solid thunk, reassuring you that the switch has been activated and will stay activated until you've let up.  Cheaper keyboards flex, shake, vibrate, ping and resonate through the backplate/case, or move in general--the keyboard needs to feel stable 100% of the time.  More weight helps here.  People might prefer o-rings for typing, but when it comes to gaming, I think just about everyone prefers the accurate feel of a solid thunk. 

2. Keycap quality/texture.  I don't really care for keycaps with cut ridges or anything like that, and rubberized keycaps feel pretty terrible in general and wear like ****.  Thick, rough textured PBT is optimal here, and this is where I feel like quality slips the most.  This applies to the spacebar too.  Before I had mechs, I would shine up a spacebar and WASD in about six months, which was extremely frustrating.  When a keycap wears, it shines, and when it's got that shine, your fingers are gonna slip.  No laser-etched caps either.  Here was my Quickfire TKL after like, three months of playing quake? - http://imgur.com/a/rwsLd - looks like ****. 

3. Rollover/ghosting.  This should be obvious in 2017. 

4. Detachable wrist rests.  Some people prefer them, some don't.  I can't use a keyboard without a rest, it just doesn't feel natural to me.  Leather would be best here, possibly a plastic, detachable wrist rest that is covered in a piece of 3oz vegetable tanned full grain leather.  Manufacturing costs might be prohibitive, but in my opinion there is nothing that would last longer and feel better.  Emphasis on detachable though. 

5. Size/looks.  The G410 is a weird shape.  Wrist rest on the left, nothing on the right?  Weird protruding handle on the left too, why?  I can't use my own wist rest, and I can't detach the left side rest if I want to.  It's just a mess, and I want nothing to do with it.  Make a normal TKL board with some extras.  Volume nobs are appreciated, weird shaped cases that are bigger and more unwieldy than they need to be are not. 

Offline chyros

  • a.k.a. Thomas
  • * Esteemed Elder
  • Posts: 2611
  • Location: The Netherlands
  • Hello and welcome.
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #33 on: Mon, 16 January 2017, 05:50:06 »
1. 1000hz
[...]
3. Low latency overall
Could you elaborate on why you would need the former and what you mean by the latter please?
Check my keyboard video reviews:


Offline Rayoui

  • Posts: 260
  • Location: Portland, OR
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #34 on: Mon, 16 January 2017, 08:15:59 »
1. 1000hz
[...]
3. Low latency overall
Could you elaborate on why you would need the former and what you mean by the latter please?

A 1000Hz polling rate has the potential for the system to poll a keypress 8x faster depending on when in the polling cycle the key was pressed. As we are talking specifically about gaming keyboards here, and it's something fairly standard that isn't difficult or expensive to implement, it seems like something that should be in every purpose-built gaming keyboard.

Other factors will, of course, compound latency. For example, the amount of time it takes for the controller to actually recognize a keypress that can be polled by the system. It doesn't sound like much, but shaving a few milliseconds in multiple places can add up to big gains.

Whether or not the difference is perceptible to the user in typing or gaming is immaterial. If the overall latency is lower, the gains will still be realized. If two players press a button at the same time and one system registers the keypress 5ms faster, guess who is going to get the kill? I certainly don't take my gaming seriously enough for these kinds of measurements to make much of a difference, but those milliseconds are very important to serious competitive gamers. If these things can be implemented easily and cheaply, why wouldn't you want them?
Clueboard  |  B.face  |  HHKB Type-S  |  Realforce RGB  |  Pok3r
Soon: MIRA SE

Offline chyros

  • a.k.a. Thomas
  • * Esteemed Elder
  • Posts: 2611
  • Location: The Netherlands
  • Hello and welcome.
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #35 on: Mon, 16 January 2017, 09:53:59 »
1. 1000hz
[...]
3. Low latency overall
Could you elaborate on why you would need the former and what you mean by the latter please?

A 1000Hz polling rate has the potential for the system to poll a keypress 8x faster depending on when in the polling cycle the key was pressed. As we are talking specifically about gaming keyboards here, and it's something fairly standard that isn't difficult or expensive to implement, it seems like something that should be in every purpose-built gaming keyboard.

Other factors will, of course, compound latency. For example, the amount of time it takes for the controller to actually recognize a keypress that can be polled by the system. It doesn't sound like much, but shaving a few milliseconds in multiple places can add up to big gains.

Whether or not the difference is perceptible to the user in typing or gaming is immaterial. If the overall latency is lower, the gains will still be realized. If two players press a button at the same time and one system registers the keypress 5ms faster, guess who is going to get the kill? I certainly don't take my gaming seriously enough for these kinds of measurements to make much of a difference, but those milliseconds are very important to serious competitive gamers. If these things can be implemented easily and cheaply, why wouldn't you want them?
I'm asking because I've gotten familiar with the issue when I started writing gaming keyboard reviews for Tom's Hardware, and I'm intrigued by the perspective of those that care about these things. The problem is that, just like the term "anti-ghosting", this appears to be a completely vacuous issue, as the polling rate isn't actually the limiting factor of a keyboard's response time, therefore shortening it will not achieve anything other than making your computer slightly slower. For those who genuinely want a high polling rate, USB isn't the interface of choice anyway, making the feature even more useless. However, so many things are hilariously overmarketed towards gamers that it's hard to see where the bull**** starts and where it ends, sometimes. I'm preparing a review in which I'm probably going to tear the hide off the manufacturer because their marketing is almost utter nonsense to those that know how a keyboard ACTUALLY works xD .
Check my keyboard video reviews:


Online FoxWolf1

  • Posts: 835
  • 154
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #36 on: Mon, 16 January 2017, 10:32:16 »
the polling rate isn't actually the limiting factor of a keyboard's response time

That's not strictly true, though, is it? I mean, there are different technologies in different keyboards in terms of scan rate and debounce method (if it's even applicable to the keyboard in question), and some of these are already recording response times well under the 8ms mark. If you take a device that's otherwise capable of a 1ms response time, whether because of non-contact sensing (optical, Hall effect, etc.), something like Cherry RealKey, or just having a debounce system that doesn't introduce an initial delay coupled with a cranked-up scan rate (some Chinese and Korean keyboards do this, and perhaps Logitech as well-- the measurements I've seen indicate very low latency for at least some Logitech models, comparable to Cherry RealKey, even), and then you poll that device at 125 hz, then surely the polling rate will be the limiting factor.
Yeah, okay then.

Offline Rayoui

  • Posts: 260
  • Location: Portland, OR
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #37 on: Mon, 16 January 2017, 10:43:45 »
1. 1000hz
[...]
3. Low latency overall
Could you elaborate on why you would need the former and what you mean by the latter please?

A 1000Hz polling rate has the potential for the system to poll a keypress 8x faster depending on when in the polling cycle the key was pressed. As we are talking specifically about gaming keyboards here, and it's something fairly standard that isn't difficult or expensive to implement, it seems like something that should be in every purpose-built gaming keyboard.

Other factors will, of course, compound latency. For example, the amount of time it takes for the controller to actually recognize a keypress that can be polled by the system. It doesn't sound like much, but shaving a few milliseconds in multiple places can add up to big gains.

Whether or not the difference is perceptible to the user in typing or gaming is immaterial. If the overall latency is lower, the gains will still be realized. If two players press a button at the same time and one system registers the keypress 5ms faster, guess who is going to get the kill? I certainly don't take my gaming seriously enough for these kinds of measurements to make much of a difference, but those milliseconds are very important to serious competitive gamers. If these things can be implemented easily and cheaply, why wouldn't you want them?
I'm asking because I've gotten familiar with the issue when I started writing gaming keyboard reviews for Tom's Hardware, and I'm intrigued by the perspective of those that care about these things. The problem is that, just like the term "anti-ghosting", this appears to be a completely vacuous issue, as the polling rate isn't actually the limiting factor of a keyboard's response time, therefore shortening it will not achieve anything other than making your computer slightly slower. For those who genuinely want a high polling rate, USB isn't the interface of choice anyway, making the feature even more useless. However, so many things are hilariously overmarketed towards gamers that it's hard to see where the bull**** starts and where it ends, sometimes. I'm preparing a review in which I'm probably going to tear the hide off the manufacturer because their marketing is almost utter nonsense to those that know how a keyboard ACTUALLY works xD .

Indeed there are other factors that cause more latency than polling. However, as I mentioned in my previous post, shaving milliseconds in every place possible matters to some people. Assume it takes ~30 ms for the controller to register a key has been pressed in the matrix (factoring in scan rate, debounce, etc). If you are only polling every 8 ms and the controller registers that keypress at the beginning of the polling cycle, you have an additional ~8 ms of latency added to your keystroke for a total of 38 ms (a 26% increase). If you are polling every 1 ms, the maximum additional latency you will add is 1 ms. Latency from polling rate isn't negated just because other parts of the chain add more latency. It still factors into the total.

With modern systems, having a keyboard polling rate of 1000Hz over 125Hz isn't going to measurably slow down the system, certainly not enough to negate the gains you will get from the decreased latency.

PS/2 would be ideal but isn't always a convenient solution. Many modern motherboards no longer include PS/2 ports in their I/O.

I do agree with one thing, there is a HUGE amount of marketing fluff surrounding any kind of "gamer" hardware. You end up with buzzwords and a list of bullet points that every manufacturer has to include on their product or else consumers will say things like "only 30 million clicks? Deal breaker!"
Clueboard  |  B.face  |  HHKB Type-S  |  Realforce RGB  |  Pok3r
Soon: MIRA SE

Offline chyros

  • a.k.a. Thomas
  • * Esteemed Elder
  • Posts: 2611
  • Location: The Netherlands
  • Hello and welcome.
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #38 on: Mon, 16 January 2017, 12:28:49 »
1. 1000hz
[...]
3. Low latency overall
Could you elaborate on why you would need the former and what you mean by the latter please?

A 1000Hz polling rate has the potential for the system to poll a keypress 8x faster depending on when in the polling cycle the key was pressed. As we are talking specifically about gaming keyboards here, and it's something fairly standard that isn't difficult or expensive to implement, it seems like something that should be in every purpose-built gaming keyboard.

Other factors will, of course, compound latency. For example, the amount of time it takes for the controller to actually recognize a keypress that can be polled by the system. It doesn't sound like much, but shaving a few milliseconds in multiple places can add up to big gains.

Whether or not the difference is perceptible to the user in typing or gaming is immaterial. If the overall latency is lower, the gains will still be realized. If two players press a button at the same time and one system registers the keypress 5ms faster, guess who is going to get the kill? I certainly don't take my gaming seriously enough for these kinds of measurements to make much of a difference, but those milliseconds are very important to serious competitive gamers. If these things can be implemented easily and cheaply, why wouldn't you want them?
I'm asking because I've gotten familiar with the issue when I started writing gaming keyboard reviews for Tom's Hardware, and I'm intrigued by the perspective of those that care about these things. The problem is that, just like the term "anti-ghosting", this appears to be a completely vacuous issue, as the polling rate isn't actually the limiting factor of a keyboard's response time, therefore shortening it will not achieve anything other than making your computer slightly slower. For those who genuinely want a high polling rate, USB isn't the interface of choice anyway, making the feature even more useless. However, so many things are hilariously overmarketed towards gamers that it's hard to see where the bull**** starts and where it ends, sometimes. I'm preparing a review in which I'm probably going to tear the hide off the manufacturer because their marketing is almost utter nonsense to those that know how a keyboard ACTUALLY works xD .

Indeed there are other factors that cause more latency than polling. However, as I mentioned in my previous post, shaving milliseconds in every place possible matters to some people. Assume it takes ~30 ms for the controller to register a key has been pressed in the matrix (factoring in scan rate, debounce, etc). If you are only polling every 8 ms and the controller registers that keypress at the beginning of the polling cycle, you have an additional ~8 ms of latency added to your keystroke for a total of 38 ms (a 26% increase). If you are polling every 1 ms, the maximum additional latency you will add is 1 ms. Latency from polling rate isn't negated just because other parts of the chain add more latency. It still factors into the total.

With modern systems, having a keyboard polling rate of 1000Hz over 125Hz isn't going to measurably slow down the system, certainly not enough to negate the gains you will get from the decreased latency.

PS/2 would be ideal but isn't always a convenient solution. Many modern motherboards no longer include PS/2 ports in their I/O.
This is why I figure this is bull****. People care about polling rates because companies tell them to so they can charge more for their products. If people GENUINELY cared about polling rates, they'd get a PS/2 port, but they don't. I get comments on my videos about old keyboards where people go "it looks slow so it can't be good for gaming" without them even knowing PS/2 is actually faster. If they cared, or actually knew how the things worked, they'd know about PS/2.

In addition, the debounce latency does definitely factor in here. Generally bounce time is specified as <5 ms, so let's assume that's what the debounce is set to. The slowest keyboards poll at 125 Hz, right? So that's 8 ms of delay at most. The debounce of 5 ms means there's a 5 ms delay at least. Assuming the very worst-case scenario, assume you press the key 1 ms before the polling is complete, that means you'll need to wait 9 ms until the next poll when your keypress will be done. 5 ms of this is due to debouncing, so you're incurring an actual delay of 4 ms, while you'd have a delay of 5 ms if you hadn't. That's a difference of 4 ms at the absolute worst - in other cases you'll have a smaller difference or even no difference at all. At 4 ms, you're below the latency of the actual computer itself, the network, the switch operation time, or possibly even the margin of error in switch production. It simply isn't significant by the standard of ANY of the parts you're working with.

Quote
I do agree with one thing, there is a HUGE amount of marketing fluff surrounding any kind of "gamer" hardware. You end up with buzzwords and a list of bullet points that every manufacturer has to include on their product or else consumers will say things like "only 30 million clicks? Deal breaker!"
Tbh the switch lifetime is one of the more useful things they market IMO. That at least has a definitive consequence - the lifetime of the keyboard. I get people saying "it doesn't have LEDs so I can't game on it" xD .
Check my keyboard video reviews:


Offline Rayoui

  • Posts: 260
  • Location: Portland, OR
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #39 on: Mon, 16 January 2017, 12:58:49 »
Indeed there are other factors that cause more latency than polling. However, as I mentioned in my previous post, shaving milliseconds in every place possible matters to some people. Assume it takes ~30 ms for the controller to register a key has been pressed in the matrix (factoring in scan rate, debounce, etc). If you are only polling every 8 ms and the controller registers that keypress at the beginning of the polling cycle, you have an additional ~8 ms of latency added to your keystroke for a total of 38 ms (a 26% increase). If you are polling every 1 ms, the maximum additional latency you will add is 1 ms. Latency from polling rate isn't negated just because other parts of the chain add more latency. It still factors into the total.

With modern systems, having a keyboard polling rate of 1000Hz over 125Hz isn't going to measurably slow down the system, certainly not enough to negate the gains you will get from the decreased latency.

PS/2 would be ideal but isn't always a convenient solution. Many modern motherboards no longer include PS/2 ports in their I/O.
This is why I figure this is bull****. People care about polling rates because companies tell them to so they can charge more for their products. If people GENUINELY cared about polling rates, they'd get a PS/2 port, but they don't. I get comments on my videos about old keyboards where people go "it looks slow so it can't be good for gaming" without them even knowing PS/2 is actually faster. If they cared, or actually knew how the things worked, they'd know about PS/2.

I don't think it's that people don't know about PS/2. If the motherboard manufacturer doesn't put a PS/2 port on the motherboard, you don't really have much choice. Many modern chipsets don't have PS/2 ports on any of the SKUs.

Besides, the difference between a 1ms polling rate and the "immediate" PS/2 hardware interrupt is, at best, less than 1ms.

Quote
In addition, the debounce latency does definitely factor in here. Generally bounce time is specified as <5 ms, so let's assume that's what the debounce is set to. The slowest keyboards poll at 125 Hz, right? So that's 8 ms of delay at most. The debounce of 5 ms means there's a 5 ms delay at least. Assuming the very worst-case scenario, assume you press the key 1 ms before the polling is complete, that means you'll need to wait 9 ms until the next poll when your keypress will be done. 5 ms of this is due to debouncing, so you're incurring an actual delay of 4 ms, while you'd have a delay of 5 ms if you hadn't. That's a difference of 4 ms at the absolute worst - in other cases you'll have a smaller difference or even no difference at all. At 4 ms, you're below the latency of the actual computer itself, the network, the switch operation time, or possibly even the margin of error in switch production. It simply isn't significant by the standard of ANY of the parts you're working with.

I see people making this mistake quite often. You are operating under the assumption that the system is polling the switch and not the controller. For example, assume you press the switch down about 3ms into the polling cycle. The controller waits the 5ms for debounce, and then says "this key has been pressed." But oh no, the system just finished polling right before the controller registered the keypress. Now we need to wait an additional 8ms for the system to poll again, for a total latency of 13ms. If we had a 1ms polling rate, the total latency would have been, at most, 6ms (the debounce time + polling cycle time).

Point I was trying to make is, having a 1000Hz polling rate vs a 125Hz polling rate doesn't really increase the cost of the board and it does have advantages (however small), so why wouldn't we want this as a standard feature in all boards? USB is the current standard and has been for over a decade, so saying we should all be using the obsolete PS/2 is a bit silly, even if it is technically better.
Clueboard  |  B.face  |  HHKB Type-S  |  Realforce RGB  |  Pok3r
Soon: MIRA SE

Offline chyros

  • a.k.a. Thomas
  • * Esteemed Elder
  • Posts: 2611
  • Location: The Netherlands
  • Hello and welcome.
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #40 on: Mon, 16 January 2017, 13:19:34 »
Indeed there are other factors that cause more latency than polling. However, as I mentioned in my previous post, shaving milliseconds in every place possible matters to some people. Assume it takes ~30 ms for the controller to register a key has been pressed in the matrix (factoring in scan rate, debounce, etc). If you are only polling every 8 ms and the controller registers that keypress at the beginning of the polling cycle, you have an additional ~8 ms of latency added to your keystroke for a total of 38 ms (a 26% increase). If you are polling every 1 ms, the maximum additional latency you will add is 1 ms. Latency from polling rate isn't negated just because other parts of the chain add more latency. It still factors into the total.

With modern systems, having a keyboard polling rate of 1000Hz over 125Hz isn't going to measurably slow down the system, certainly not enough to negate the gains you will get from the decreased latency.

PS/2 would be ideal but isn't always a convenient solution. Many modern motherboards no longer include PS/2 ports in their I/O.
This is why I figure this is bull****. People care about polling rates because companies tell them to so they can charge more for their products. If people GENUINELY cared about polling rates, they'd get a PS/2 port, but they don't. I get comments on my videos about old keyboards where people go "it looks slow so it can't be good for gaming" without them even knowing PS/2 is actually faster. If they cared, or actually knew how the things worked, they'd know about PS/2.

I don't think it's that people don't know about PS/2. If the motherboard manufacturer doesn't put a PS/2 port on the motherboard, you don't really have much choice. Many modern chipsets don't have PS/2 ports on any of the SKUs.

Besides, the difference between a 1ms polling rate and the "immediate" PS/2 hardware interrupt is, at best, less than 1ms.
Well, you can't have it both ways :p . You can't say it's mandatory to shave single milliseconds off response time, but NOT to save the same time by going PS/2 :p .

Quote
Quote
In addition, the debounce latency does definitely factor in here. Generally bounce time is specified as <5 ms, so let's assume that's what the debounce is set to. The slowest keyboards poll at 125 Hz, right? So that's 8 ms of delay at most. The debounce of 5 ms means there's a 5 ms delay at least. Assuming the very worst-case scenario, assume you press the key 1 ms before the polling is complete, that means you'll need to wait 9 ms until the next poll when your keypress will be done. 5 ms of this is due to debouncing, so you're incurring an actual delay of 4 ms, while you'd have a delay of 5 ms if you hadn't. That's a difference of 4 ms at the absolute worst - in other cases you'll have a smaller difference or even no difference at all. At 4 ms, you're below the latency of the actual computer itself, the network, the switch operation time, or possibly even the margin of error in switch production. It simply isn't significant by the standard of ANY of the parts you're working with.

I see people making this mistake quite often. You are operating under the assumption that the system is polling the switch and not the controller. For example, assume you press the switch down about 3ms into the polling cycle. The controller waits the 5ms for debounce, and then says "this key has been pressed." But oh no, the system just finished polling right before the controller registered the keypress. Now we need to wait an additional 8ms for the system to poll again, for a total latency of 13ms. If we had a 1ms polling rate, the total latency would have been, at most, 6ms (the debounce time + polling cycle time).
This is actually the exact situation I was tying to illustrate except 7 ms into the keypress, but maybe it wasn't completely clear. I chose 7 ms in instead of 3 because if at 3 ms in the controller would register at the precise time the switch finished debouncing, so there should be no delay whatsoever. I should add that for convenience, I considered the baud rate high enough that the time between switch and controller is negligible.

Quote
Point I was trying to make is, having a 1000Hz polling rate vs a 125Hz polling rate doesn't really increase the cost of the board and it does have advantages (however small), so why wouldn't we want this as a standard feature in all boards? USB is the current standard and has been for over a decade, so saying we should all be using the obsolete PS/2 is a bit silly, even if it is technically better.
First, the feature is being used to justify high prices for products, so while it may not cost the manufacturer much to implement, it still costs you at the end of the day. Also the polling rate can cause compatibility issues with some systems. But the thing that bothers me most tbh is that people make this out to be something massively more important than it could ever be. The difference in timing is so small that other factors cause delays even bigger than this, but no-one seems to give a damn about them. That to me smells of aping marketing speak, and not independent thought, and I hate it when that manifests in a market.
Check my keyboard video reviews:


Online tp4tissue

  • * Destiny Supporter
  • Posts: 10821
  • Location: Official Geekhack Public Defender..
  • OmniExpert of: Rice, Top-Ramen, Ergodox, n Females
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #41 on: Mon, 16 January 2017, 13:42:15 »
This is like one of those huddles..

Where there was 1 cool kid..

Then the Cool kid left.. 


And all the uncool kids pretend like they still want to hang around with each other..


hahahahahahahahahahahha

Offline Findecanor

  • Posts: 3558
  • Location: Stockholm, Sweden
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #42 on: Mon, 16 January 2017, 15:01:38 »
Logitech should DEFINITELY consider making TKL gaming keyboards
G310 and G410 have been around for quite a while.
BTW, please Logitech, do get rid of the ridiculous bezel. Those negate the purpose of having a tenkeyless form factor - which is to use less space, and they don't look good.

For those who genuinely want a high polling rate, USB isn't the interface of choice anyway, making the feature even more useless.
The USB protocol can have practically as little latency as PS/2, in theory, on the protocol level ... But in practice it requires that the operating system prioritizes low latency, that the controller on the motherboard is up to the task and that there is no USB hub in the chain.
There are motherboards right now, advertised as always having with 1000Hz polling rate, made especially for gaming mice. The USB protocol for mice is the same as for keyboards.... so if motherboard manufacturers would make a "gaming keyboard" USB port as well, I.e. two ports of the same type instead of just one then the problem would be solved.

Yes, I believe debouncing latency is the limiting factor otherwise. I believe manufacturers use a longer debouncing delay than 5ms to account for the odd switch now and then that may not have been 100% perfect.
« Last Edit: Mon, 16 January 2017, 15:03:09 by Findecanor »
"It is no measure of health to be well adjusted to a profoundly sick society."
Daily driver: Phantom (Lubed Cherry MX Clear, Lasered Cherry PBT keycaps with Row A. Plastic "Frankencase". Custom firmware, Swedish layout)

Offline Rayoui

  • Posts: 260
  • Location: Portland, OR
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #43 on: Mon, 16 January 2017, 21:05:02 »
Well, you can't have it both ways :p . You can't say it's mandatory to shave single milliseconds off response time, but NOT to save the same time by going PS/2 :p .

I'd argue that there's a big difference between an 8ms delay and a 1ms delay. The difference between a 1000Hz polling rate and PS/2 would be virtually immeasurable as there is no such thing as zero latency. It will still take the system a certain amount of time to recognize the hardware interrupt after it is sent to the system.

Quote
This is actually the exact situation I was tying to illustrate except 7 ms into the keypress, but maybe it wasn't completely clear. I chose 7 ms in instead of 3 because if at 3 ms in the controller would register at the precise time the switch finished debouncing, so there should be no delay whatsoever. I should add that for convenience, I considered the baud rate high enough that the time between switch and controller is negligible.

I think you're misunderstanding how the delays interact with each other. The debounce delay would not be subtracted from the polling delay, it would be added since the system only polls the controller, which is after all the other functions that the controller performs (scanning, debounce delay, etc).

Regardless of what numbers you choose to use in an example, you are correct that if you press a key and the debounce delay finishes right before the polling cycle, it will register with virtually no delay induced by polling. However, if you press a key and the debounce delay finishes right after the polling cycle, the total delay will be the debounce (and all other controller-induced latency) with the remaining polling delay added. Granted, this would not occur every time, but using a polling rate of 1000Hz would negate the problem altogether.



Sorry for the crude graphic; on my way out the door for work.
Clueboard  |  B.face  |  HHKB Type-S  |  Realforce RGB  |  Pok3r
Soon: MIRA SE

Offline chyros

  • a.k.a. Thomas
  • * Esteemed Elder
  • Posts: 2611
  • Location: The Netherlands
  • Hello and welcome.
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #44 on: Tue, 17 January 2017, 04:59:26 »
I'm subtracting it because the debounce is inherent; it has nothing to do with the polling rate. If you want to measure the difference, afterwards you can just subtract the values because both debounce. A 5.25 ms delay is only 0.25 ms when considering the polling rate.

And whether the difference between 8 ms and 1 ms is big; well, it's 7 ms to be exact :P . That's the kind of latency you'd expect on your own system when hosting a game - and I don't think anyone complains about that.
Check my keyboard video reviews:


Offline Rayoui

  • Posts: 260
  • Location: Portland, OR
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #45 on: Tue, 17 January 2017, 05:59:45 »
I'm subtracting it because the debounce is inherent; it has nothing to do with the polling rate. If you want to measure the difference, afterwards you can just subtract the values because both debounce. A 5.25 ms delay is only 0.25 ms when considering the polling rate.

I'm not sure what you mean. A 5.25 ms delay is a 5.25 ms delay. Debounce has nothing to do with polling rate, but both affect total latency. The debounce delay is always there and is always the same. The amount of latency you get from the polling rate is determined by when in the polling cycle the key was pressed. Neither one should be subtracted from the other.

The scanning of the matrix and the debounce delay is performed by the keyboard's controller. The system never sees any of these operations. It is only after these operations are complete and the keypress is finally registered by the controller that it is available to the system at the next polling cycle. The remaining time until the next polling cycle completes will always be additional latency. It does not cancel out.

Quote
And whether the difference between 8 ms and 1 ms is big; well, it's 7 ms to be exact :P . That's the kind of latency you'd expect on your own system when hosting a game - and I don't think anyone complains about that.

You say 7ms, I say 800%.  ;)

That 7ms may not mean much to your or I, but to a competitive gamer, that 7ms could be the difference between scoring a kill or being killed. Everyone has different priorities. We shouldn't declare a feature to be superfluous just because we don't personally need it, particularly if it doesn't add to the overall cost of the product.

In fact, a lot of the "gaymur" mechanical keyboards are some of the cheapest mech boards available. I mean, GHers are regularly spending $500-1000 usd on kustoms, and those don't even have 1000Hz polling rate!  :p
Clueboard  |  B.face  |  HHKB Type-S  |  Realforce RGB  |  Pok3r
Soon: MIRA SE

Offline chyros

  • a.k.a. Thomas
  • * Esteemed Elder
  • Posts: 2611
  • Location: The Netherlands
  • Hello and welcome.
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #46 on: Tue, 17 January 2017, 06:25:09 »
I'm subtracting it because the debounce is inherent; it has nothing to do with the polling rate. If you want to measure the difference, afterwards you can just subtract the values because both debounce. A 5.25 ms delay is only 0.25 ms when considering the polling rate.

I'm not sure what you mean. A 5.25 ms delay is a 5.25 ms delay. Debounce has nothing to do with polling rate, but both affect total latency. The debounce delay is always there and is always the same. The amount of latency you get from the polling rate is determined by when in the polling cycle the key was pressed. Neither one should be subtracted from the other.

The scanning of the matrix and the debounce delay is performed by the keyboard's controller. The system never sees any of these operations. It is only after these operations are complete and the keypress is finally registered by the controller that it is available to the system at the next polling cycle. The remaining time until the next polling cycle completes will always be additional latency. It does not cancel out.
The debouncing delay is present on both systems and it's not the issue at hand. Surely what you're interested in is the DIFFERENCE between the two, right?

Quote
Quote
And whether the difference between 8 ms and 1 ms is big; well, it's 7 ms to be exact :P . That's the kind of latency you'd expect on your own system when hosting a game - and I don't think anyone complains about that.

You say 7ms, I say 800%.  ;)

That 7ms may not mean much to your or I, but to a competitive gamer, that 7ms could be the difference between scoring a kill or being killed. Everyone has different priorities. We shouldn't declare a feature to be superfluous just because we don't personally need it, particularly if it doesn't add to the overall cost of the product.
Like I said, it might not cost them much, but it does cost us in the end. And the differences is delay (mind you, 7 ms is the highest it can be) are so small that if this were a factor, competitive gaming would be dominated by much smaller things. If this were significant, you couldn't go to a LAN party and have a fair game because of the inherent different latencies on the network, for example. What I'm trying to say is that if this is an issue, why aren't any of the others that contribute a much larger delay?
Check my keyboard video reviews:


Offline Rayoui

  • Posts: 260
  • Location: Portland, OR
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #47 on: Tue, 17 January 2017, 07:08:58 »
The debouncing delay is present on both systems and it's not the issue at hand. Surely what you're interested in is the DIFFERENCE between the two, right?

What do you mean by both systems? I'm referring to a single system, with one keyboard, where the keyboard's controller is performing the debounce delay. The PC itself does not have any kind of a debounce delay.

Quote
Like I said, it might not cost them much, but it does cost us in the end. And the differences is delay (mind you, 7 ms is the highest it can be) are so small that if this were a factor, competitive gaming would be dominated by much smaller things. If this were significant, you couldn't go to a LAN party and have a fair game because of the inherent different latencies on the network, for example. What I'm trying to say is that if this is an issue, why aren't any of the others that contribute a much larger delay?

But they are  :)

That's why we have things like high-performance, low-latency mice with no measurable acceleration up to speeds that no human could ever move, 1ms response time monitors with super high refresh rates which are overdriven to reduce motion abberations, technologies such as G-sync to further reduce latency by syncing GPU output to display refresh... you get the idea.

You are correct; there are certainly larger sources of latency. That is exactly why gamers find it important to reduce latency wherever possible, even a few milliseconds at a time. All those milliseconds add up to larger gains if you reduce them at multiple links in the chain.
« Last Edit: Tue, 17 January 2017, 07:12:26 by Rayoui »
Clueboard  |  B.face  |  HHKB Type-S  |  Realforce RGB  |  Pok3r
Soon: MIRA SE

Offline Tangtawan

  • Posts: 87
  • Location: Bangkok, Thailand
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #48 on: Tue, 17 January 2017, 13:17:31 »
I wonder if anyone has ever benefit from more than 6 keys rollover ? or 1000Hz polling rate? Personally I never felt any differences.
My boards : Model F AT (6450200), Industrial M (1394950), Model M SSK (1391472), Model M13 (13H6705), Model M (1390131), Focus FK-2001 (ALPS SKCM Blue), HHKB Pro 2, Realforce 87U 55g, KBP V80 Matias Click, TADA68 (Gateron Reds), Filco Majestouch 2 TKL (67g Ergo Clears), Cherry G80-11800 (Ergo Clears), KBT One 108 (MX White), Ducky One TKL RGB (MX Black), Filco Majestouch Tenkeypad (MX Brown), Leopold FC210TP (MX Red)

Offline MasterBash

  • Posts: 34
Re: I'm from Logitech: What to use to test a Gaming keybaord
« Reply #49 on: Tue, 17 January 2017, 15:43:57 »
you don't need to notice the difference. Its just there... The lower the better.

A mouse with 5ms latency vs one with 10ms... Would you notice it? Maybe not, but the mouse with the lower latency responds faster.

A 120hz vs 144hz monitor. Would you notice the extra 24hz? Its possible that you won't, but it has 24hz more anyway.

You don't have to notice it, its just better because the latency is lower. Also with NKRO... Well, its also there. To be honest, I don't want to know if I can tell the difference between 6KRO and NKRO in a game. I don't want my keyboard to not register something, so I go straight to NKRO for the peace of mind. You know, I don't tell myself "Lets see if I can buy a keyboard that could effect my performance in a negative way"... I prefer to have a low latency keyboard with NKRO, so I don't have to worry about it.
« Last Edit: Tue, 17 January 2017, 15:47:31 by MasterBash »