question regarding keycap theory (history)

[Eszett]

Hi. Since words are not always clear enough, I drew a little illustration for my question:

[CPTBadAss]

I think it was to make it so it's relatively easy to type on and the arms would act as a lever when you pressed down on the switches. And it wouldn't jam the arms of the typewriter.

[Eszett]

Hi Captain! You describe the purpose of this arrangement on typewriters. What is the purpose nowadays with those inclined Cherry (or SA) keycaps?

[CPTBadAss]

Improved ergonomics over the typewriter arrangement.

[Eszett]

Can you elaborate why a staircase arrangement is ergonomically better?

[kurplop]

I don't think it always is. Sometimes it improves comfort and reachability.

I think it is mostly one of the many vestigial features left for the sake of familiarity. Things like the QWERTY layout and staggered rows are other features that we'd probably be better off without but attempts to change them hasn't been well received.

[Eszett]

@kurplop QWERTY, staggering and staircasing are certainly relicts from old times. But while QWERTY and staggering is useless nowadays, I'm not sure about staircasing. Some people I spoke to said, a staircase profile is at least better than a totally flat one. What do you think (?)

[kurplop]

I think a lot of it has to do with your typing style. I have not used a true stair step board but I do like a certain amount of row profiling that creates a bit of an arc. This allows better access to most of the keys without leaving home position. I tend to stay pretty well anchored to one position and this may not be as important to others whose style more resembles the conductor at the Philharmonic, gesticulating his way through Beethovans fifth.

The early IBM computer keyboards have somewhat of a stair step design and that may be telling. I think they were trying to replicate some of the traditional typewriter feel and sound and developed great elaborations to do so. I find it interesting that I've heard many mechanical keyboards lovers admit that they can type faster on RD's or flat laptop boards. It doesn't matter; it's more about the experience.

[jacobolus]

The shape of the keycaps makes particular typing styles / finger motions more or less convenient.

For most people, I suspect (I haven’t done a wide-scale formal study, this is just my observations; YMMV) the main finger motion use to press a key is flexion at the proximal finger joint (the metacarpo–phalangeal or MCP joint) where the finger meets the palm. Likewise, the main motions used to locate keys are flexion/extension of the further finger joints, the interphalangeal or IP joints, combined with some rotation of the wrist and movement of the whole forearm, depending on the key.

If the keycaps are all flat across the top, then to reach a further key (e.g. on the number row) requires either moving the hand or stretching the finger out forward, including a bit of flexion at the MCP joint. These are some combination of slightly slow and/or uncomfortable: if you move the whole hand around, then you need to move it back to strike the next key, and in extreme cases some key sequences become difficult to type; or if you stretch the finger out (still on a flat keyboard), the flexion in the MCP joint reduces the further range of motion in that joint, and the action of actually pressing the key by flexing the joint becomes weaker.

If instead the keys have a bit of a vertical step between rows, just extending the IP joints will place your finger closer to the key top, requiring little to no flexion in the MCP joint to reach the key top, and less whole-hand reaching.

This would all be easier to understand with a picture or with a physical demonstration using some keys and a finger, sorry. :-)

(I gave a lecture about this subject at a bay area keyboard meetup, and should really someday get around to writing up an extensive website about my thoughts on keyboard ergonomics complete with many diagrams and videos. Not today though.)

For anyone trying to learn about this for themselves, I recommend trying to get hold of some keycaps of different profiles, and directly compare some alternatives. They can be ****ty printed ABS ones, you can mix and match wildly different keycap profiles, or you can adjust the profile by sticking bits of modeling clay on the tops, etc., doesn’t have to be production quality. Pay careful attention to what motions your hands and fingers make as you type, and think about what muscles are involved, etc. Read a book on hand kinesiology if you feel especially motivated. If you have the equipment for it, try taking some high-speed videos.

[jacobolus]

I find it interesting that I've heard many mechanical keyboards lovers admit that they can type faster on RD's or flat laptop boards. It doesn't matter; it's more about the experience.

I think this has to do with overall familiarity and practice history, keyswitch stiffness, and switch travel distance, rather than the keycap profile per se. Would be interesting to run some large-sample-size formal studies.

[Oobly]

The shape of the keycaps makes particular typing styles / finger motions more or less convenient.

For most people, I suspect (I haven’t done a wide-scale formal study, this is just my observations; YMMV) the main finger motion use to press a key is flexion at the proximal finger joint (the metacarpo–phalangeal or MCP joint) where the finger meets the palm. Likewise, the main motions used to locate keys are flexion/extension of the further finger joints, the interphalangeal or IP joints, combined with some rotation of the wrist and movement of the whole forearm, depending on the key.

If the keycaps are all flat across the top, then to reach a further key (e.g. on the number row) requires either moving the hand or stretching the finger out forward, including a bit of flexion at the MCP joint. These are some combination of slightly slow and/or uncomfortable: if you move the whole hand around, then you need to move it back to strike the next key, and in extreme cases some key sequences become difficult to type; or if you stretch the finger out (still on a flat keyboard), the flexion in the MCP joint reduces the further range of motion in that joint, and the action of actually pressing the key by flexing the joint becomes weaker.

If instead the keys have a bit of a vertical step between rows, just extending the IP joints will place your finger closer to the key top, requiring little to no flexion in the MCP joint to reach the key top, and less whole-hand reaching.

This would all be easier to understand with a picture or with a physical demonstration using some keys and a finger, sorry. :-)

(I gave a lecture about this subject at a bay area keyboard meetup, and should really someday get around to writing up an extensive website about my thoughts on keyboard ergonomics complete with many diagrams and videos. Not today though.)

For anyone trying to learn about this for themselves, I recommend trying to get hold of some keycaps of different profiles, and directly compare some alternatives. They can be ****ty printed ABS ones, you can mix and match wildly different keycap profiles, or you can adjust the profile by sticking bits of modeling clay on the tops, etc., doesn’t have to be production quality. Pay careful attention to what motions your hands and fingers make as you type, and think about what muscles are involved, etc. Read a book on hand kinesiology if you feel especially motivated. If you have the equipment for it, try taking some high-speed videos.

The movements you're describing have more to do with keycap heights per row than keycap top angles.

I find flat keycap top angles to be very efficient and like to use SA profile with a flattish board. On a board like this I find "stepped" keycaps to be a hindrance since you have to raise your fingers over the step when changing rows. I first noticed this after using Penumbra caps on a board for a few months and then putting Cherry profile caps on (GMK Dolch). I noticed an increase in fatigue and a small decrease in typing accuracy and speed. The ideal for me is a curved keycap profile with a continuous curve from bottom row to top with no steps. A smaller diameter curve will give bigger height differences, larger diameter will give less.

If you don't want to move your hands around while typing and reduce the required movements to only the fingers, I find SA to be just a little less than ideal. It'd be nice to have a little more difference in heights and a tighter curve.

On my ergonomic board I achieve this by using only SA profile Row 1 and Row 3 caps with inverted Row 1 caps on the bottom row, Row 3 on the middle and normal Row 1 on the top row.

On a "normal" qwerty board I tend to move my hands around slightly as I type, so this reduces the need for such a difference in heights and I find standard contoured SA profile is nicest. This uses more muscle groups when typing, but by making only small movements of each, reduces possible fatigue and allows long typing sessions despite the extra fingertip travel compared to my ergo board (which uses an efficient layout and layers to reduce finger travel).

[jacobolus]

The movements you're describing have more to do with keycap heights per row than keycap top angles.

That’s correct, I think the top angle is mostly irrelevant, especially for keys past the home row. The important part is the height step.

For nearer keys, having a forward-sloped key that better catches the fingertip as it pulls back seems like it should be a slight improvement, but it would take doing some rigorous testing to figure out how much effect that had, and what the proper height/angle that should be.