The Problem with Mechanical Switch Reviews (2016)
input.club
input.club
This realization was a big moment of disillusionment in (most) mechanical keyboards for me :)
Totally explains why I ended up settling on HHKB, which actually feels like it registers the button press right when you feel that "click" (given it is much closer to IBM Model M than to the usual mechanical switches).
Why do I like those linear switches? Because you can mash them hard (then you need no bump because the key bottoms out and that is your definite bump) or you can type them close to inaudibly silent. Which is kind of important if you e.g. do audio recordings.
Why black ones? I am a bass player, I have strong fingers and most other linear switches feel to light to me.
I guess some clicky optical switches may also behave this way?
I did once love my Model M, but it's more tiring to use than newer keyswitch designs are. Pushing enough to actuate the key means building up enough force to buckle the spring, at which point there's not really anything I can do to avoid bottoming out. It's sort of the double IPA of keyswitches: really, really, really good in small doses, but not necessarily something you want to be experiencing for 8 hours straight.
Then again, in my experience, it's not an issue. The resistance curve of the rubber domes makes it hard to feel the buckle yet not activate the switch. Once it buckles, it's hard to avoid the "thock".
I too like my Topre boards. But I also haven't gotten a chance to try any of the newer Cherry-style third-party switches.
I think there are plenty factors at play here. Spring resistance, wobbliness of the switch, activation distance, feel of the tactile bump/click…
Recently I've been using Otemu brown, violet, and orange and Gateron browns. All should be quite similar, yet in the end they feel totally diferent. I love Otemu oranges and dislike brown and violet. Gateron browns feel very good, but a bit wobbly for my taste.
I have tested more than 20 switch types, and most of my guesses based on someone else opinion or specification were usually wrong. The only thing I know is that I want tactile switches, with rather low activation force. Except of that, I have no idea if I'll like the switch or not.
I like the way that author approaches this, but I'm not sure we can conclude anything from a scientific approach here. Switches are just like art or music. Sometimes you like a song you think you should hate. Yet it makes something to you and you just can't stop listening ;)
Did anyone do a blinded mechanical switch ABX test? I bet most would fail to differentiate except between the major categories - linear, buckle, dome, ...
Feel is subjective, but it does matter, and it's something worth exploring for people who spend 8+ hours every day using a keyboard. It can also have important impacts on ergonomics.
Does this level of detail regarding feel really matter? I mean, all keyboards are usable and people do adapt to new/unfamiliar input devices, and it's not like laptops ship with high-action mechanical keyboards.
And regarding ergonomics, the standard keyboard layout is renowned for not being stellar ergonomics-wise, so I'm skeptical regarding how relevant ergonomics is in this field. In fact, isn't a standard Microsoft Natural ergonomics keyboard way better than your typical mechanical switch keyboard?
To whom?
This is a very subjective thing and the answer is that it matters to some people more than to others. The point is, that even spending 200 Euros on a thing that you use for hours every day can be justified if this is the thing you spend your life with if it feels better.
Note: it must not be objectively better, it just needs to feel better.
To everyone.
My point is that no one was born with an optimal keyboard feel and a crave for a particular type of clickiness. Most of the talk around mechanical keyboards always refer to acquired tastes, without any consideration that people can and do adapt to sub-milimiter differences in key layouts and range.
And that's perfectly fine.
Just don't try to justify buying grossly overpriced luxury goods due to whimsical personal tastes as a health and safety requirement.
Disclaimer: I own and use a mid-range mechanical keyboard as my main worktool.
Well, the existence proof that it does matter is the shitty Apple laptop keyboard that had no travel and broke if you got micron-sized dust particles in it.
In addition, I pound keyboards. So, the force of the spring absorbs force more gradually before the key bottoms out. That's probably better for my fingers.
Mostly among people who sell alternative keyboards. What little scientific research has looked into this, has, as far as I've seen, failed to find any measurable ergonomic problems with standard keyboards or the qwerty layout.
Which isn't to say that there can't be any adaptive value for people with specific needs. Years ago, I had to switch to a split tented keyboard after I broke my wrist and temporarily lost the ability to pronate it enough to use a standard keyboard. Now that I'm all healed up, though, I mostly stick with it as a matter of taste. I also use my laptop's built-in keyboard when I'm away from home, and it doesn't give me any problems. If anything, I would guess I'm getting some benefit from regularly switching between a split column staggered keyboard and a standard keyboard, since it exposes my hands to a wider variety of movements.
The switch type does not control the shape. In fact, one of the oldest and most respected ergonomic keyboard designs (Kinesis Advantage) uses mechanical switches. I use an IV-Works AV3 mechanical keyboard, which uses the Arisu layout, which is a modified Alice layout, which was modeled on the Microsoft Natural keyboard.
Many people use split keyboards for ergonomics since they provide the ability to adjust the angle, and these are quite common in the mechanical keyboard world, with some of the better known designs being the ErgoDox and the ZSA Moonlander.
Ergonomics is both a function of layout, as you point out, but also the key design because the key design determines key travel, weight, and whether or not you bottom out when typing, which can effect impact on your fingertips and finger joints, in addition to the impacts of layout on your wrists.
Except for Topre.
Not that I'm mad or anything.
Also I wonder how much of the preference is like old running shoes - I don't care how ratty my old shoes are or how nice the new ones are, the new ones will feel wrong compared with the old ones for a while.
Yes, newer Cherry switches are different from older switches. You'll see people in the community refer to older switches as "Vintage".
Also, if you use MX Browns, you might find you enjoy some other tactile switches more for newer boards. Browns these days generally feel a bit squishy/loose and somewhat scratchy due to the nylon components. I find I like POM stems, with minimal lubrication, and slow springs with a mid-weight (62g). Of course, all this is a matter of preference, but I'd encourage you to try some POM stem tactiles like Holy Pandas.
Edit: Glorious makes pre-made "Holy Pandas" under their own branding using the original INVYR tooling: https://www.pcgamingrace.com/products/glorious-panda-mechani...
Check out high end electric pianos at your local music shop, and compare to $100 Wal-Mart kid keyboards. The resistance and smoothness and feedback from the keys integrate with muscle memory better (and in the case of pianos, graduated activation helps determine the volume of a note and sustain.)
I don't know if typing with good switches helps with rsi, but it certainly helps with comfort and satisfaction.
A good clarinet or trumpet or saxophone will have similar features to the feel of different types of mechanical keyboard switches. Sure, you can play decent music on cheap mass produced elementary school instruments, but it's going to objectively feel better playing a high end bespoke instrument. No jarring or disruptive aspects in the motion to interfere with the process of putting characters on the screen.
While a for a PC keyboard, a basic membrane keyboard will do the job outside of maybe some very niche highest levels of competitive esports
Mech keyboards and obsession with them are purely a luxury, hobby thing
I agree. The fact that your typical keycap set typically sells for more than some full blown keyboards is already a clear tell.
However, I'd add that some properties of your typical mechanical keyboard are indeed very convenient. For instance, easily removable key caps and hotswappable switches do bring in the nice consequence of making keyboards easy to clean, maintain, and repair.
Some mechanical keyboards offer a symmetrical layout, and allow the thumb to use more than one key switch. Those are significant advantages for a keyboard user. (Since these are niche keyboards, they incidentally happen to have mechanical switches).
Though, I think most people interested in mechanical keyboards stick to a keyboard with the same layout as the basic membrane keyboard. (Each to their own, I guess).
And I am not a mechanical keyboard afficionado. I own one mechanical keyboard for well over a decade. But I had to do with cherry switches on a non-keyboard related electronics project.
From the OP, or from the discussion here? Serious question. Because it's genuinely not clear which you mean (for me it's the latter).
OP is fine, he's measuring things. The audiophile industry also always includes "frequency response" graphs and all kinds of other measurements in their reviews, yet the whole field is still a bit of a sham.
Personal preference being expressed here, while I can and have used buckling springs at length, they are not my favourite for two reasons. One, they click twice. Once when actuating and once again on the way back up so they are 2x as noisy. Two, they have a fair bit of resistance. Really, likely the best way to type on a buckling is to push down and let the key push your finger back up. It can be tiring, especially if you're not imprinted on this way of typing.
My favourites are complicated ALPS. Yes, I know they are no longer made with the exception of simplified whites (from Matias), but I have a lot of exposure to ALPS and am actually typing this on complicated orange ALPS right now. You can look at force curve graphs, but the typing experience is very different compared to buckling. When you press the key (the travel needed to actuate is not as far either compared to buckling), you get one click not two. The key does not need to push your finger back up because it's lower force, so when you hear or feel the click your finger can already be swinging away to another key.
MX is not as good as ALPS, but it has a large market share. Just like VHS wasn't better than Beta, but won anyway.
Cherry, Gator, Outemu etc so far have not shown such short life behavior...
If you don't want to bother with the tester, the grey quiet-click are roughly equivalent to cream dampeneds and are good for use in an office environment so that neighbours don't burst into tears at your acoustic assault in the open-concept floor plan. If this isn't an issue and noisier/bouncier is more fun, then white is what you want.
Tai-Hao also sometimes makes ALPS keycaps.
The Problem with Mechanical Switch Reviews - https://news.ycombinator.com/item?id=13006608 - Nov 2016 (5 comments)
> Graphs give me a much better idea of what I'd actually be getting
The issue with that is that you "have to have the feel" for the graph. Except for extreme circumstances, if you're not familiar with how to approximate the feel from the graph, it won't really help you.
I guess my point is that you still have to have some kind of actual, real-world experience.
In my case, I kind of deduced that I liked "bumpy" switches, so I went the "brown" route. I was happy with my Gateron Browns. When I wanted a second keyboard to not haul the first around, I figured Cherry Brown would be "close enough". They supposedly have the same specs and everything. Nope. They feel much "harder" to press.
I do like them enough, so I didn't send the KB back. But I do tend to miss some keys when I switch from Gateron to Cherry, and I do tend to press like crazy on the Gateron after switching from the Cherry.
This is true; it takes time and effort to understand anything quantifiable once it's been quantified. As somebody with extensive experience with math and physics, I didn't really grok the force/displacement plots on first glance. So, I read the axes. That got me most of the way there. Then, I looked at the force/displacement plots of various keyswitches, and reflected on my own experience with keyboards from clacky to mushy.
So, yes, you do need actual real-world experience with keyboards. With that experience, and the skill of reading a plot, you can get a good sense of what to expect without actual real-world experience with this specific keyswitch.
Got me curious, because cherry browns are considered "lighter" than gateron browns. Looking at the data, they take less force to register a keypress, but more force to bottom out. So, that's a perfect example of why you can't trust individual reviewers: some are interested in getting a keypress, some are more focused on bottom-out force.
https://input.club/the-comparative-guide-to-mechanical-switc...
https://input.club/the-comparative-guide-to-mechanical-switc...
But as someone who's not particularly well versed in physics, it's not clear to me how much of a difference 1 gf makes in terms of feel.
Plus, on the same pages, regarding actuation, it says that the Cherry is "light", whereas the Gateron is "medium". But a few lines below, the actuation force is actually higher on the Cherry (37 gf vs 36 gf). Also, the "spring force" is higher on the Cherry, so maybe it's expected for it be "stiffer"?
My pet peeve is fat fingering. To first order I think any keyboard sucks if I am afraid to use the backspace or home button because I risk toggling the "insert/replace".
The force curve might have something to do with that, you might also take some objective measurements about the key layout that are predictive of that, but I could waste a huge amount of time trying out "bad" keyboards without prediction of my personal fat finger rate.
- most can't be easily reprogrammed and some with only some programmable keys demand extra software at runtime to work;
- most are just designed out of IBM PC classic keyboard changing just the look&feel from one to another, there is exactly ZERO competition in keyboard features behind aesthetic/visual aspects/colorful backlit etc;
- keys tend to be smaller and narrow than the past, I suppose to save plastic for the OEM, nothing really for the end user;
- Cherry profile &c makes impossible to move keycaps freely around the keyboard (classic IBM parabolic design, with is very simple actually can perfectly: all keycaps are physically identical).
To be less off-topic: switches soldered in ways that's hard substitute them at home if some breaks after a bit of time. Switch with different keycaps connectors just to create incompatibility between vendors and no standard to rescue...
Long story short: real tactility and force might be interesting to discuss and compare people taste on those topics, noise levels are another topic, but the above issues well... Are IMO far more crucial and still ignored by most...
I'm living in Emacs (EXWM), with a binding strategy of a bind a single key hit or a mod+key because even with more spatial movement I'm still quick and with less mental context switch/overhead to move in space vs using "more mods"/keyboard layers or key chords...
In the past SUN Type 6 alike keyboards suffice, nowadays it's hard to find anything and start my own design demand simply too much time and effort to achieve daily usage quality for my tastes. That's about me of course, but the rant is in general: in the past some companies, even journals,have asked for custom keyboards to be more efficient, now the focus is just on consuming devices instead of producing ones. That's alarming IMVHO and worth to debate beside the rant...
In an extreme example, if it activates at 0.1 and the force required increases 10 fold at 0.2, the the force from 0.2…4.0 obviously doesn’t matter (because the user would feel the force increase and stop pressing).