Mechanical Keyboard Sound Isn't That Simple
Figure 1: I couldn't think of a more literal way to represent this article if I tried... Looking back just a few years ago, there’s no doubt that the huge influx of people that joined the hobby at the peak of the COVID pandemic were drawn to keyboards by way of YouTube, TikTok, and other audio-visual content platforms. Even as the output from these content creators has waned in recent months, their collective impact and legacy on the keyboard hobby is rather firmly etched in the history books. As a result of all of their sound tests, build logs, and opinion videos, the message is clear to any new person joining the hobby: mechanical keyboards are all about the sound. Thock this, clack that. Whether it’s keyboards, keycaps, or even singular switches, seemingly everyone new to the hobby meticulously pores over each component of their keyboard not in an attempt to figure out how it will feel in hand, but how it will sound as they’re furiously grinding their way out from...
Mar 27, 2024
On the input club site it mentioned Cherry MX Clears as a possible switch option for this board. Since Cherry currently doesn't make RGB compatible Clears with transparent housing, will this still be an option, and if so will they be moded in a way that makes them RGB compatible?
I think the more people try the Kailh switches the more people will come to see the wisdom of the choice. Any outliers who absolutely hate them can always easily swap them out later.
tl:dr - Kailh switches are very good!
Everything was negative, except one semi-recent keyboard review in YouTube, which mentioned they have been improving the quality of their switches in newer batches. Your comments and that video helped confirm I'm not losing it, and I'm going to pull the trigger for a build I'm working on. Yes, they are pretty damn good!
For linear I prefer Gateron, and I do love my Topre switches.
May you explain what you mean with better quality control? Did you guys tested it over time, for example with a machine that press the key milion times?
I think the biggest thing you'll notice is that the cherry switches feel kind of scratchy in comparison, and the way the key actuate feels different. I'll try to explain The second point, but keep in mind that this is all subjective...
The cherry switches basically have two steps when you press down on it. When you press a key, it will drop down a bit with almost no pressure, then there's the click and avkind of a release (this is the actuation point) and the key will bottom out quickly after that. You won't notice this when you're typing but if you press it down slowly you'll feel it, and if you open up the switch you'll see that it's the way it's designed.
Gateron switches also have this step before the actuation point, but there's additional resistance when you first press down so the change in force required is not as erratic. It jury feels more natural, and you don't feel like there's metal scraping metal like with the Cherry switch. It feels smoother, almost buttery in comparison.
That being said...
I still prefer Cherry tactical switches over Gateron. The smooth buttery feel is great on linear switches, but for the clicky switches Cherry just feels right. That might be because I'm used to Cherry switches, but I'm looking for something else so it's fair to say I'm not in love with either.
The Kailh switches are somewhere in the middle of the two. In the end they're both good and the differences are not overly noticeable. It ultimately it comes down to preference. One thing I would suggest is to look up typing videos - if you listen closely you can really hear the difference. Hopefully that helps, good luck!
To be honest, I found earlier comments by some folks off-putting. "Kailh sucks.." or "Kailhs are the best..." They both sound gibberish. There are so many subjective elements to this, and some objective criteria on very small sample is just a data point. I think it's better to put data and hypothesis forth, and let people make their own choices. Swappable switch design goes a mile for that and I appreciate that.
In any case this will be an excellent opportunity for Kailh to demonstrate the quality of their product.
>"They just had a good price for switch + sockets from kailh, so now they are telling people that kailh are good"
I even saw their page with kailh graphs, but nothing obvious showed high-higher QC, also that comparison based on new Kailh switched popped out when they probably already planned to use kailh switches (graph popped out recently), so i don't know, i'm not saying "they are lying", but i would really like to understand why Kailh are better in their opinion (and they can be totally right on this, who knows...)
Also, a real quality test should include a long-term use, aka a machine that press the switch repeatedly to see how the switch behaves after millions of pressures.
I hope that @OverKill can elaborate a bit on it and tell us why they find Kailh the "best switches" (over the graph they already showed to us that honestly did not proved it, at least to me)
So what makes a switch good? Feel and longevity/durability?
Feel is entirely subjective, of course, but what I can say is that in a blind test I did at the last meetup I was at I called some of the newest Kailh brown switches "tuned clears" on the stock card describing the board, and people were very fond of them. I don't think I heard a single complaint about them. The actuation is very smooth, the feeling of the switches was totally consistent across the board, and the tactile bump was more pronounced than the Cherry Browns.
This alone tells me a HUGE issue here is just peoples preconceived ideas or assumptions about Kailh switches. Some days, frankly, I'm shocked Gateron gets the praise they do from the community. They are SO inconsistent between batches (especially with weight). They are very smooth, but at this point the new Cherry MX and Kailh switches are comparably smooth.
Now, a lot of the complains coming from people having "failing switches" with older Kailh was not actually the switch itself failing, it was the boards they were in. And we all know who is most known for using Kailh, Razer. I have had pretty much every Razer out there, and will say they easily have the highest failure rate of any keyboard brand I've used yet. Out of 8 Razer boards I have, 3 don't work after VERY moderate use. In two of the cases some switches weren't working, but it was actually PCB issues causing the malfunction, not a failed switch. End of the day, switches can fail, but it is extremely rare. I can count the number of failed MX switches I've had on a single hand, from all brands combined. And those that failed chattered. Half were Zealios from the first round at that.
The light pipes are awesome as well, and are something nobody else currently has. Check out this picture from my coming review showing how much better/more even the light pipe switches distribute the light from the SMD LEDS:
Kailh switches do have a different style housing, so you cant mix/match housings from other MX variants, which could be a downside to some.
I adamantly think the hatred of Kailh is unwarranted, especially for the new switches at this point. Frankly, the majority of folks commenting on them haven't even tried the new ones out yet. Really the biggest determining factor on if they are "good" or not is simply feel here, and to make that judgement call you really have to test them for yourself. Personally I think if you put your assumptions aside and really test them with an open mind most will enjoy them equally with any other MX variant at this point.
The tl;dr is consistency. Gateron has *terrible* consistency and Cherry has ancient and gritty tooling.
You will have switches that measure very differently even within the same batch. Even the well regarded Zealios switches (which are quite good, and imo the best things that Gateron produces, much thanks to Zeal), have issues between batches. This points to some combination of tooling, plastic choice and melting profile that they can't keep consistent for some reason. Gateron doesn't even have proper part numbers for all of their switches. They basically just yolo them and hope people buy them. Most of the Chinese market is very confused that the US (and EU) prefer them. Many keyboard factories in China won't even use them due to QC issues (Input Club and Massdrop have talked with most of them). Some Gateron switches use non-heat resistant plastics in combination with cheaper contacts. This the plastic to be warped by the soldering process, causing the electrical contacts to spread apart unevenly. Basically this would result in brand new Gateron switches becoming chatty and requiring significantly more debounce than the standard sub 5 ms. To give you an example, take this graph of a Gateron Red -> https://plot.ly/~haata/150/gateron-red/ Go and click on the legends for the different presses. You'll notice that each press is very different near the end of the press. That's pretty unacceptable imo. Not only should each switch be consistent with each other, but they should also be consistent between presses as well. And, while I can go measure hundreds of switches for you. This switch passed Gateron QC, this means it was deemed OK. And this is not OK. That is the problem with Gateron. Not that the switches can be quite good, but that they can also be bad and inconsistent.
Cherry on the other hand has been using the same tooling for a very long time and are running into the issues Unicomp has with their tooling. Burrs. Ever wonder why some Cherry switches have a wear-in period? Switches are molded, not sanded or machined. A wear-in is just Cherry being cheap and lazy with their tooling. A side note about Cherry. Until very recently, Cherry has been very very difficult to source. In the early days, Massdrop had to do very interesting things so source them at all, and it's always been hard to get the quantity they've needed without ordering a year in advance... Cherry has changed though. I've started seeing them trying to help the enthusiast market, so perhaps this will start to get better (maybe even fix their tooling too :D). https://flic.kr/p/UDCfVX Here's a microscope photo. The surface should be shiny and smooth, instead we have all these indents caused by slow damage to the tooling (or poor milling of the tooling).
I'm not sure it's well known but Greetech, Gateron and Kaihua (Kailh) all make switches for Razor. Basically they're just doing what any behemoth is does and sources from multiple factories. So while the tooling may be the same, the process and QC may not.
Now, before I go more into Kaihua, I will say that I really like competition in the switch space. Having more companies than just Cherry is the reason that Cherry is even doing anything mildly interesting with their switches now. Keyboards (and switches) are truly a personal preference thing and for enthusiasts in general a "perfect switch" just does not exist, period. In my own designs I try to balance taking excellent ideas from a variety of switch manufacturers in combination with trade-offs and limitations of my manufacturing partners. Because even if you design an amazing switch, if you can source or make them, it's meaningless at the end of the day.
I will be the first to admit, that I was very skeptical of Kaihua. The original K-Type prototypes, back in January 2016 used Greetech switches (which are probably one of the more interesting switches, but they do have a few QC concerns as well) as they provided a much more tactile experience than other switches available (for example https://plot.ly/~haata/113/greetech-brown/ notice how high the tactile peak is). I first tried Kaihua switches back in early 2014, and tbh, they were awful. When one of my colleagues in the keyboard space asked them why they didn't just clone Cherry, they had an interesting response. Their goal was to make a more reliable Cherry-type switch. This is why you see 70 million keypresses for Kailh switches and around 50 million for everyone else. Oh and Kaihua really does test their springs to 70 million keypresses (they're doing just that for our custom switch). Selecting the default switch types in a keyboard is a question of trade-offs. And really, it came down to Kailh coming out on top. * Easy to source * Good force curve and very similar to Cherry MX * Actual light pipe (one of the upcoming reviews is going to have a real lighting comparison) * Consistent and excellent QC
Now the story doesn't just end there. In terms of keyboard switch factories Massdrop went to visit the Kaihua factory and were extremely impressed with how clean their factory was. Now this might not sound like much, but this is one of the key components to consistency and quality. Yes, you *can* make excellent things in a dirty factory, but it's definitely a lot harder. If it wasn't for Kaihua we wouldn't even be able to use the (pretty amazing) hotswap inserts. They've also taken a chance on me doing non-standard things with the upcoming custom switch (definitely pushing the limits of what you can do with Cherry style switches).
Admittedly much of my personal skepticism of Kailh switches stems from their partnership with Razer and Razer's notoriously unreliable mechs (as you described). I suppose it couldn't hurt to give them another try, especially when this is a hot-swappable board.
For switches, yeah i wanted to put my assumptions aside, i didn't started this conversation with a square mind, I just wanted some more infos on them (i believe that many other people want more infos on these new kailh too, without having to test them, even if testing it is a good way to judge). I was talking about an objective test, not based on how they feel. For example the pic you showed above is a nice start of "proofs" I was talking about and damn it seems good!
Also, for old Kailh, i don't believe they were good, i don't want to say "i tested it and it was bad" because my opinion will be based on a super small sample, not accurate. But if i remember correctly, even Input Club, before picking the NEW kailh as switch for the k-type, it wasn't talking very well about the old kailh. I now see the page edited so i can't remember exactly what it was wrote, but i remember that i spent hours on these pages and Kailh in general was the less liked.
Basically, i'm asking , "Hey Input-Club, you stated something big by saying that Kailh is in your opinions the best switch, i would like to know more since you are going to sell these switches on the k-type as stock switches, can you tell us more and give us some proofs of what you stated to convince us? (since we are used to not trust Kailh switches)"
You, @livingspeedbump, started a good point tho, with at least your opinion and a pic, i was expecting something bigger in this direction from Input-Club, to help us to get convinced that Kailh is in fact a good switch.
If I can i would like to ask other 3 questions related to this:
1) Why the Kailh does not use the common switch open mechanism? This is not a problem on hotswappable keyboards, but is a problem on non-hotswappable keyboards! Also, it was a pain to open a Kailh :D 2) I read many times that Cherry has new molds for switches and seems like you confirmed it, you know how to recognize these new Cherry batches or where to buy new Cherrys? Also, are Cherry Silent/speed/natural white or in general rgb version a safe bet if you want to get the new-mold Cherry? 3) I'm sorry but i have to ask, i'm hyped for the new switch made by you and you just said that this will happen thanks to Kailh. It will be possible to buy the Haata switch without purchasing the K-type or will be only possible to bundle it with the K-Type? Also, when will be announced?
And a last thing, thanks for the time you dedicated to answer to me with that wall of text, i really appreciated it and hope it will be useful to many other people here! I was really hoping in a detailed explanation like this. Thanks again Haata / InputClub :)
PS: Seems like i will have to edit my profile picture soon. Damn.