Most ergonomic keyboard switches.

[czarek]

Guys I recently started to have symptoms of RSI in my right hand fingers.

I blame Realforce with 55G for it. I think they are very bad from ergonomics point of view. There is relatively high force on a very top of a key travel required to overcome, which causes your finger muscles to put a bit of stress. This is enough to push the key to the bottom (bottom out) and very hard not to hit the bottom, even after using those switches for a very long time (I used mine for over 1.5 year, before that I used 45G, which was easier not to bottom out for me).
That means you hit the steel plate (ok with 0.5mm of rubber cushioning) about 500 times a minute.
I'm not a medical doctor, only an engineer, but to me, that sounds like asking for injury sooner or later.

I did a little analysis with test drives of some other switches I had (or could have and built) keyboards with.

First one I gave a try after I could no longer use my beloved Topre 55G was Filco with MX Blue. Ligher springs are first thing I feel. Although it's very hard not to bottom out at first. After couple hours using it I bottom out less, and after few days I can type comfortably without bottoming out at decent speeds. At full speed though, pretty hard not to bottom out.

Second in line was another Filco with MX Browns. Just as light as blues, but very hard to feel tactile bump. When typing slowly I could avoid bottoming out but at any decent speed I just don't feel the bump. I tried using this keyboard as it was for 3 days and unfortunately, I find the bump too small to be useful. It actually gets in a way, makes the switch feel like it was broken or something, especially on modifiers.

After this experience there was a time for linear switch. I was too lazy to desolder vintage blacks from terminal board, and I didn't want to mod anything so I went with black gaterons which I put into the Filco (the one that had browns originally). First thing I noticed were heavier springs. At first I thought they would put more straing on fingers and until I adjusted to higher activation point I made a lot of mistakes while typing, and I did bottom out a lot. Surprisingly, very soon my fingers learned where the switch activates (by feeling the force pushing them up) and I started to type fluidly without bottoming out even at very high speeds (for me high speed begins at about 100 wpm), without any fatigue. That's quite fascinating as the springs (even though they're rated at 50g) feel much heavier than Cherry MX Blue and Brown.

That intrigued me and I decided to pull the shot and mod another keyboard (Code 87 that originally came with broken MX Blues - needed new switches anyways) with a switch I dreamed about since beginning of this experiment (similar to the one I use in my ErgoDox, just even better). I desoldered vintage blacks from old terminal board, 62g gold Korean springs, got them lubed and put inside the Code. It felt butter smooth, super light and I could type on this keyboard faster than any other I tried so far (over 110 wpm, which was max for me on Topre 55G). Unfortunately though at higher speeds it was hard to avoid bottoming out, due to the lightness of the springs. Possibly 65g springs could be better, but I haven't tried those. I did have 68g spring under space bar though and that felt a bit on a heavy side (but good for space bar).

As nice as those switches are (and feel almost identical to my ErgoDox anyways) I felt bored with them and decided to try one more switch I had home - Gateron Blue. It's actually pretty heavy switch. Just a hair lighter than MX Green, but with higher and more pronounced activation point. Reminds me blue/white alps a bit and is actually very fun and satisfying switch to type on. I have a lot of fun using keyboard with those but I can feel it's not doing any good for my fingers. It's almost impossible not to bottom out (due to high force required to overcome tactile point). Ergonomics aside though, if you haven't yet, try this switch. I guarantee that if you'd like your Cherry MX Blue to have a bit more pedigree and character (like Alps or BS), this is the switch for you.

There are 2 more keyboards that I want to mention. First one being my father in law's Dell AT101 (Black Alps) which I play with sometimes when I visit him. They don't feel as heavy as they are. I really like their character, but I just can't avoid bottoming out, even if I pay attention and try to type slowly. It's like it was meant to bottom out those switches.
Unfortunately I don't have any keyboard full of Matias switches and other Alps (I do have all Matias varieties, cream dampened alps, and blue complicated alps) but I have a feeling that all those would behave similarly.

The last one I want to talk about is obviously the Model M. I use it very often. It's always in rotation with any other keyboard I own and here's why. I find it to be very ergonomic indeed. You may think it's crazy because of it's heavy springs, long travel etc. But think about it. Pre-travel for BS to collapse is about 2mm. just letting your fingers go over that distance gives them enough inertia to overcome the (impossible not to notice) tactile point which is not hard (like in Gateron and Cherry Blue) but very satisfying, and after that you still have over 1mm of cushioning on already collapsed spring. To me this is genius idea and when really tired, I always get back to my trusty IBM. The only thing I could wish for is a bit lighter springs. Yes I know Model F sounds perfect, but I really can't stand their layouts (I could live with Unsaver or maybe F77, so thank you Ellipse ) and the sound of model F actually irritates me (while I like the clack of Model M).

I'm really sorry for putting my thoughts as they came to my mind, which may be hard to read as they're slightly convoluted, but I hope it will give you a bit of puzzle. Maybe you have/had similar thoughts and would like to share them?

[seebart]

So what are your RSI symptoms like? How intense? Have you recently changed anything in your setup? How may hours a week do you spend typing? Do you have any other physical disabilities? Do you exercise regularly? You might want to start thinking about other non-keyboard factors also. Have you tried variations on ergonomic setup at your desk?

For something like this goooogle is good:

https://www.google.de/search?q=Human+fa ... desk&hl=de&

[czarek]

So what are your RSI symptoms like? How intense? Have you recently changed anything in your setup? How may hours a week do you spend typing? Do you have any other physical disabilities? Do you exercise regularly? You might want to start thinking about other non-keyboard factors also.

I feel pain in my muscles and joints in middle, ring and little finger of the right hand.
I type around 10-12 hours a day / 5-6 days a week but I take breaks (coffee, food, bio, dog walk) and change posture often.
It's rather not possible that it's a mouse since I use decent mouse (DeathAdder) and M570 trackball interchangeably.
Also, my desk is huge, perfect height, with monitors shelf at perfect height, and I have very nice ergonomic chair (Ergomax Flexi).
Currently after I left Realforce in a box and used linear and light switches for last couple weeks I felt much better but after I started using blue Gaterons symptoms are coming back so I'm going to either go back to black Gaterons or if it gets really bad, get on Ergodox again. I keep it as last resort because it will make switching to Model M and laptop keyboards nearly impossible.

[seebart]

10-12 hours a day 6 days a week is a lot! If DeathAdder is a decent mouse is relative. Clocking the hours you do you need a perfect (perfect for your specific ergonomics) setup. Sounds like you're testing variations already. Good. Every human body is different, so it's really hard to give you any advice. I type much less, about 20-30 hours a week and never have any problems with any keyboards (I change keyboards often). Just try what works better for you, that's all I can tell you.Are you sure that your desk is "perfectly" set up for you? I have heard some people tell that standing at their desk works well for them, but at 10-12 hours a day?

[davkol]

derp

[HuBandiT]

First, as a temporary measure to alleviate your symptoms:

Matias unofficially has a specialty very light linear switch developed for a stenography board called Stenosaurus: http://stenosaurus.blogspot.hu/2014/07/ ... tches.html

You might want to build a board with those, and then add direct electronic audio feedback within the board. I remember having used a board with Cherry MX blacks twenty years ago. The board had audio feedback, with a buzzer built in, where each keypress made a short sound. I recall having the immediate audio feedback helped me stop the habit of bottoming out after a while, at which point it started feeling like 'typing on air", involving much less physical effort.

Second:

Do you hover-type?

I hold the opinion, that the most ergonomic keyboard is any keyboard which you hover-type on, instead of touch-typing. By hover-typing I mean extensively moving around your hands using your arms/elbows/shoulders to reach keys. This stands in contrast with the too often taught (and in my opinion so ergonomically and productivity-wise foolish) touch-typing, where you are trying to keep your hands in a fixed position while exerting your fingers extensively to try to reach keys.

So: do you hover-type? If not, I say you should seriously consider. Best hardware-investment you can make, since it is portable across all physical keyboards.

And finally, for a permanent (albeit off-topic on this forum) solution:

RSI is inflammation, and inflammation very often comes from a bad diet weakening the body, preventing it from recovering from levels of stress that it could normally handle without problems. I suppose you also have muscle stiffnesses. Have your CRP checked out by your physician, and seek out a specialist who knows where inflammation really comes from and can eliminate the root cause. Then your body will be back to being strong and robust and resilient again, and the matter of spring stiffness in your keyboard will not matter to it much, if at all, anymore.

[seebart]

Very good points HuBandiT, I totally forgot about hover-typing. Diet and exercise can always be an issue.

And welcome to DT!

[Muirium]

Indeed.

My trick is to hover type — I'm bloody useless at keeping my hands fixed in home row, it feels desperately unnatural — and to type a bit slower. The first can lead to the second if you're not used to it! >100 wpm is a lot of stress any way you look at it. I can go almost that fast in a typing test, but that's not how I actually work at all. I type the speed I think, which is in bursts, and averages more like 60 words per minute. None of my keyboards give me pain. Sitting, or indeed standing while I work.

[kbdfr]

[…] Do you hover-type?

I hold the opinion, that the most ergonomic keyboard is any keyboard which you hover-type on, instead of touch-typing. By hover-typing I mean extensively moving around your hands using your arms/elbows/shoulders to reach keys. This stands in contrast with the too often taught (and in my opinion so ergonomically and productivity-wise foolish) touch-typing, where you are trying to keep your hands in a fixed position while exerting your fingers extensively to try to reach keys.

This is a complete misunderstanding of what touch-typing is about.
Touch-typing has nothing to do with above all not moving your hands, but just with using certain fingers for certain keys instead of looking at the keyboard and hitting the next desired key with whichever finger is next. Of course when touch-typing you move your fingers and also your arms, elbows and shoulders. And of course when touch-typing your hands do not rest anywhere but hover over the keyboard and continuously bring your fingers in the best position to hit the next key with the least possible effort.
As an example, to hit the Backspace key, you don't just extend your right pinky upwards and sideways while resting your other fingers on J, K and L, but of course move your whole hand and just minimally extend your pinky. Muscle memory continuously tells you how to next move your hands and fingers.

In addition, I just do not believe that at halfway professional typing speed audio feedback makes any sense as the next key will have been pressed long before the sound info has been received, processed and reacted to - and this several times a second on a continuous basis.

[Muirium]

I agree with you about audio feedback. The beeper in my Kishsaver is cute and all, but quite useless. I'm glad there's a volume knob underneath the keyboard: I turn it up from time to time just to put a smile on my face, before going straight back to mute! Of course, added sound effects are entirely superfluous on buckling spring anyway…

Linear keyboards are where that feedback arguably makes sense, but does it really? Isn't your fingers own feeling of travel really the only thing that matters? I sure couldn't listen out for a click each and every time. It's just a marker, to train you. As is the string of text appearing on your screen! Not that I'm against clicky switches, I love the sharpness of the buggers, but linear is just as effective if you type lightly enough not to hammer them all home.

There are definitely different styles of touch typing out there. Which I think confuses the issue quite a bit. Many people say "touch typing" to distinguish between competent keyboard use and visual hunt and peck! In that sense I'm a touch typist, as I can use blanks just fine. But I don't even rely on homing bumps because my style is to fly all around the place. I routinely use two or more different fingers to hit any given key, depending on where my hands are coming from to hit it. I often even use fingers to hit space! The key is that I don't think about it. I'm typing by touch, sure, but not according to anyone's rulebook!

[kbdfr]

By the way,

Guys I recently started to have symptoms of RSI in my right hand fingers. […]

Quite obviously you should investigate things you do with your right hand but not with your left hand.
Possibly a mouse problem. Continuously extending your arm to grasp the mouse and in this extended position straining your fingers to click might be the (or at least one) cause.

There are definitely different styles of touch typing out there. Which I think confuses the issue quite a bit. Many people say "touch typing" to distinguish between competent keyboard use and visual hunt and peck! In that sense I'm a touch typist, as I can use blanks just fine. But I don't even rely on homing bumps because my style is to fly all around the place. I routinely use two or more different fingers to hit any given key, depending on where my hands are coming from to hit it. I often even use fingers to hit space! The key is that I don't think about it. I'm typing by touch, sure, but not according to anyone's rulebook!

Do you really type all that using different fingers for the same keys without having even a single look at your board?

[Muirium]

You'd be amazed. And likely a little horrified!

[oobacke]

Sad to hear you are having problems with RSI. The problem is the input device itself (Keyboard and Mouse) and the long hours you engage on it - even with pauses, they are just not long enough. So why not mix it up with completely eleminating finger driven HIDs - at least for 40-50% of the time? I was thinking about setting up Dragon. There is a very interesting video up on how to use it when programming:

Of course, this will take a lot of time getting used to and getting it to a point where it is ready to be efficient again. On the other side, you can be sure that your RSI will get worse with age! Some engineers and programmers have had to switch profession because of this. So consider long term solutions and not just a different mouse or keyboard for short term improvement.
I am still very interested in HIDs using brain waves. Problem is, of course, that there has been little research in those fields. If we could get to the point where a basic concept was developed in a team with linguists to create concepts we could interpret via signal input... it would be great, we would be on the way to GITS neck input ports.

With both, voice and brain waves, you need a smart database that grows over time with the person using it. Dragon does provide that, but still they do not ship it with anything related to coding. I do not know why they did not tackle that yet, there should be a giant market out there. I get that it is a learning experience of a few months to years since you are learning to speak programming language! But think of the benefits (and also the dangers: voice strain!). But at least one would be able to input 40-60% by hand and 40-60% by voice which ever you feel should do more.

Also, do not work when feeling pain! You will make it worse! I cannot stress this enough, but you are destroying more and more.

Regards,
Chris

[DanielT]

You'd be amazed. And likely a little horrified!

That sounds like my typing style too I'm all over the keyboard, I have tried to learn like in the "books" but not only it slowed me down but started to cause wrist pains it was a total failure.
Also what causes me pain are heavy switches like MX Clear, MX Black and if I use them longer IBM Buckling Spring. I use light switches and never had a problem.

[Wilkie]

For what it's worth, I regularly type a lot using (modern) MX Black (KUL ES-87, stock caps) and find it very comfortable. No doubt Gateron Black would be smoother, but I'm okay with this. More curious about MX Linear Grey. Heavy switches have always felt more comfortable to me over the long haul.

[kekman]

I think maybe the curved nature of the Model M could be helping a lot when it comes to relieving the strain on your wrists, maybe you should try some more keyboards with an ergonomic curve to see what works and what doesn't. Other than that, good luck battling it.

[HuBandiT]

This is a complete misunderstanding of what touch-typing is about.

Kindly talk to the authors of most touch typing materials I've seen.

Touch-typing has nothing to do with above all not moving your hands, but just with using certain fingers for certain keys instead of looking at the keyboard and hitting the next desired key with whichever finger is next. Of course when touch-typing you move your fingers and also your arms, elbows and shoulders. And of course when touch-typing your hands do not rest anywhere but hover over the keyboard and continuously bring your fingers in the best position to hit the next key with the least possible effort.

I don't think that's traditional touch-typing; at least not as described in the touch-typing materials I've seen (few and far inbetween). Most materials I've seen kinda pushed you into staying in the home position and rarely move for typing letters and space. (Functionals - for example your example backspace, or tab, or arrows, or F1-F12 etc. obviously do prompt a hand movement even in these schools. But the bulk of awkward finger-contortion is still there for typing letters.)

As an example, to hit the Backspace key, you don't just extend your right pinky upwards and sideways while resting your other fingers on J, K and L, but of course move your whole hand and just minimally extend your pinky. Muscle memory continuously tells you how to next move your hands and fingers.

I am not sure your example with Backspace is a good example, see above.

In addition, I just do not believe that at halfway professional typing speed audio feedback makes any sense as the next key will have been pressed long before the sound info has been received, processed and reacted to - and this several times a second on a continuous basis.

I think it does. Both in the training phase as well as during full speed typing:

1) During full-speed typing: There indeed is a latency between the audio cue arriving to your ears and you hearing and processing it, and coming back to it. But just like when typing fast you sometimes realize from haptic feedback you made a typing error several keystrokes after the fact, because that is how long it takes to process the haptic feedback cues from your fingers (as in "ahh, I accidentally hit the edges of two keys there, must have typed both cars - let's stop, take a look, and backspace and correct if needed"), audio feedback can provide the same ongoing check. In fact, I posit, that even on keys which do have a haptic cue, during full speed, trained typing you do not actually use that haptic cue in your servo system to stop finger motion for that particular finger - your sevo feedback loop is too slow to do that. But instead you only use it as a long-term learning/training tool in the calibration phase and the continuous ongoing recalibration process.
2) During calibrating your motor servos ("muscle memory") to your particular keyboard hardware: The audio feedback stands in for the missing haptic cue on linears. Remember, the goal was to make typing more ergonomic for czarek, one potential avenue for that is using light linear keys, where one emerging subgoal of that is by preventing him from bottoming out his keys, for which an audio feedback (or any other suitable sensory feedback) could be a potential training tool. Once the audio feedback taught you during slow typing where your keys activate and what force you need to exert to push them that far, then you can start increasing your speed, with the audio control.

A piano player (I don't know whether you ever played) doesn't hit the keys until they hear the sound and then stop/release - they train and play a lot on pianos in general and on the particular piano in question, to build up a model of the dymanics of the piano key action ("mechanism") in their servo system (muscle memory) in terms of multi-dimensional force-speed-distance maps (varying with pitch as well since keys are heavier and need more force/energy in lower octaves and are lighter and need less force/energy in higher octaves), and then during performance they highly rely on those maps for individual motions, while only slowly/gradually recalibrating themselves if needed.

In fact I would go further and posit, that that thought extends cleanly to all manually actuated acoustic instruments.

I don't see why the same would not apply to computer keyboards, just because the modality/genre of the final expression differs (words instead of music); music and language are similar in very many levels of thought process.

But, to cut things short, I am working on a prototype that - among other features - incorporates audio feedback over linear switches; once finished I will welcome you to try it out and share your experience on usefulness.

[HuBandiT]

What could be constructive to the discussion is to see videos of the hands of people typing.

[kekman]

What could be constructive to the discussion is to see videos of the hands of people typing.

What would you like to see from that? How different wrist position, typing style, typing layout, differently keyboard layout, switch type etc. can affect speeds or amount of wrist movement? If you can be more specific I can deliver a crude video.

[Miko]

I had some RSI about two years ago. What ever you decice to do, i'd suggest you to stop *any* usage of keyboards, mice and pencils until it gets better if possible. This only get's worse and in the long term it can have impact on your ability to use a computers at all. I'm in the very luxurious position that I, as a uni-student don't have a boss and I don't know what health-system you have access to.

Maybe you can call-in sick for the next and a half week (and then maybe you have holidays between christmas and new-year?).

For me, after 2-3 Weeks it went away. I changed my very unhealthy sitting position, and bought a cheap vertical mouse. I can recommend these a lot. They are cheap, the look cheap, they feel cheap, but are great. From my personal (totally unscientific) experience the position of the wrist joints is more imporatant than the switch. Just remember that people used typewriters some time ago, a highly unergonomic "switch" and could avoid RSI.

This is my mouse:
http://deskthority.net/other-devices-f3 ... t4433.html

This is how I sat:

Spoiler:

Maybe let somebody take a photo of your working position?

[Proword]

What could be constructive to the discussion is to see videos of the hands of people typing.

These three videos -

http://www.youtube.com/watch?v=pYJtF1I3PRs

https://www.youtube.com/watch?v=o4H931A3BDE

http://www.youtube.com/watch?v=hxH7Uith0tQ

may be instructive. The first one shows a typist using a Maltron keyboard with the Malt key distribution. The second shows a typist using another Maltron keyboard but with the QWERTY key distribution. Note the difference in hand movement(s). The QWERTY keyboard has the typist hovering her hands and moving them far more than with the Maltron layout, with the hands almost totally at rest on the platform(s) built into the keyboard case. This hovering slows the typist down as well as using much more muscular effort and is an indicator of the inefficiency of the QWERTY layout.

The third video shows a typist using a Kinesis Advantage keyboard, which is based upon the Maltron keyboard, but using the Dvorak key distribution. Again, compared to the QWERTY keyboard, there is much less muscular effort expended, increasing keying efficiency.

[kbdfr]

A few thoughts about the videos linked in the last (necro, but still interesting) post and the comments therein:
- The different camera perspective in video #2 vs. #1 and #3 seems to account for at least part of the perceived hand movement difference.
- Restricting typing to finger movements puts the whole stress on the fingers alone instead of distributing it over the whole arm.
- The whole hand/arm weight bears on the very point where the wrist rests on a surface, which will induce pain in the long run (the reason why wrist rests often are padded).
- If the wrist is not completely static but still moves (it always does, look at the videos), there will also be friction in addition to that.

[Proword]

- Restricting typing to finger movements puts the whole stress on the fingers alone instead of distributing it over the whole arm.

One of the things which the Maltron and Kinesis Advantage keyboards have in common compared to a "standard" or flat keyboard (regardless of which key distribution is used - ie, QWERTY, Dvorak or Colemak) is that the keys are laid out in straight vertical "columns" whereas the flat keyboards have the keys offset to left or right as a result of long held design practices going back to the manual typewriters. Because of this sideways hand movement within the columns is minimal. This means that the "fingers alone" are doing exactly what they evolved to do - extend and contract in straight lines - ie grasp and release, thus producing little or no stress compared to that generated on a flat keyboard with sideways hand and wrist movements.

- The whole hand/arm weight bears on the very point where the wrist rests on a surface, which will induce pain in the long run (the reason why wrist rests often are padded).

The "whole" weight does not bear upon the wrist. When the hand is not moving it is resting on the base of the palm, which has relatively soft padding. The area under the wrist is not touching any part of the keyboard. Further, on the case of Maltron keyboard there is a built-in platform which is angled and sculpted to allow this to occur. The Kinesis keyboard is a smooth surface without any sculpting. Thus there is no necessity for a "wrist rest". In measuring the weight of the palm during resting on this platform, it is probably no more than about 200 grams. (The two photographs below show the differences between the two cases.)

Prior to making the first (stenography) video I suffered mild sporting injuries to my wrists and I attempted rehabilitation by putting rubber padding on the palm rests, which you can just see at the very end of the video when I lift my left hand off the keyboard. However, this proved unnecessary and I removed it soon after.

- If the wrist is not completely static but still moves (it always does, look at the videos), there will also be friction in addition to that.

I've been using a Maltron keyboard since 1986 and I have detected no evidence of friction, either on the keyboard case or the palms of my hands.

[vvp]

It is possible to type comfortably on a contoured keyboard without moving wrist. Well some keys may need a slight movement. In case of Kinesis Advantage (and my hand), it is the two top pinkie keys and the Alt keys.

Anyway some people believe that not moving wrists when typing is slower because forearm provides a "lever". I.e. one should move forearm around when typing not because it is needed for reaching keys but because it speeds up typing. In my opinion, that does not sound plausible. Though I could believe that moving whole hands around is healthier.

[kbdfr]

[…] This means that the "fingers alone" are doing exactly what they evolved to do - extend and contract in straight lines - ie grasp and release […]

[…] like in the first pic below:



The ability to just "extend and contract in straight lines" is the starting point from where the evolution gave our hands the ability to perform elaborate movements.
Of course this does not mean that complex movements are per se better, but as our hands are literally made for them, contending that restricting to basic moves or positions is per se better is equally absurd.

[Proword]

Whether apocryphal or not, Einstein is quoted as saying "Everything should be as simple as it can be, but not simpler." Opening and closing the fingers achieves the end desired, so why go further.

[kbdfr]

Whether apocryphal or not, Einstein is quoted as saying "Everything should be as simple as it can be, but not simpler." Opening and closing the fingers achieves the end desired, so why go further.

Read your (apocryphal or not) Einstein quote again:
"Everything should be as simple as it can be, but not simpler."

Your solution seems to work for you,
but that does not make it universal nor intrinsically better:
https://foldorscrunch.wordpress.com/fol ... crunching/

[Proword]

Perhaps you should read my post again. I never mentioned the word "restricting". That is your terminology.

[vvp]

Take care Proword.
Kbdfr is a user of a keyboard with about 200 keys. He can hit all of them without looking and within 1 second.
On the same side he is surprised that Muirium (an avid 60% user) does not need to look at his keyboard. Kbdfr is not a common keyboard user one can find here. His needs are definitely extraordinary.

[Proword]

Thanks for the "heads up" I'll try to remember to tread carefully. I mean what would I know? I only began teaching word processing back in 1981, and then trained as a Court Reporter in 1990. This job involved transcribing court proceedings at 150 + words per minute for 5-6 hours a day. Not a really great achievement I suppose.