geekhack
geekhack Community => Keyboards => Topic started by: aviphysics on Fri, 22 February 2013, 11:22:50
-
I was just thinking about the normal human response times and how long you have to respond to a tactile or audio que from a switch.
It seems like if you type at a somewhat slow 30 WPM, you have an average of about 150 keystroke per a minute. To put that another way, a key press needs to take less than 400 ms. Of course the key doesn't click until about halfway down, so you have about 200 ms or so to respond to the tactile or audio feedback and stop your finger from bottoming out. Just a rough approximation of course. The few papers I managed to find on the subject suggested that in 200 ms, a younger individual might have sufficient time to respond to the click but an older person wouldn't and much above that neither would.
To me this suggests that the tactile and audio que probably don't really do much to keep you from bottoming out, unless you are a slower and younger typist. For everyone else, it is just about keyboard feel.
-
Nope, they're not useful at all...
Also, fast typists ALWAYS bottom out.
I stand that MX blacks are probably the best Technical switch, because it does away with all the fluff.
-
I was just thinking about the normal human response times and how long you have to respond to a tactile or audio que from a switch.
It seems like if you type at a somewhat slow 30 WPM, you have an average of about 150 keystroke per a minute. To put that another way, a key press needs to take less than 400 ms. Of course the key doesn't click until about halfway down, so you have about 200 ms or so to respond to the tactile or audio feedback and stop your finger from bottoming out. Just a rough approximation of course. The few papers I managed to find on the subject suggested that in 200 ms, a younger individual might have sufficient time to respond to the click but an older person wouldn't and much above that neither would.
To me this suggests that the tactile and audio que probably don't really do much to keep you from bottoming out, unless you are a slower and younger typist. For everyone else, it is just about keyboard feel.
Tactile and Audio feedback would help a lot if the keyboard industry would give us long switches.
But since they have all colluded to provide us only with super short little switches there is no time for the feedback to do its job. There is no time for the spring to do its job. 4mm of travel is entirely toooooo tiny.
I would love to build a long switch and build a keyboard out of them.
-
I was just thinking about the normal human response times and how long you have to respond to a tactile or audio que from a switch.
It seems like if you type at a somewhat slow 30 WPM, you have an average of about 150 keystroke per a minute. To put that another way, a key press needs to take less than 400 ms. Of course the key doesn't click until about halfway down, so you have about 200 ms or so to respond to the tactile or audio feedback and stop your finger from bottoming out. Just a rough approximation of course. The few papers I managed to find on the subject suggested that in 200 ms, a younger individual might have sufficient time to respond to the click but an older person wouldn't and much above that neither would.
To me this suggests that the tactile and audio que probably don't really do much to keep you from bottoming out, unless you are a slower and younger typist. For everyone else, it is just about keyboard feel.
Tactile and Audio feedback would help a lot if the keyboard industry would give us long switches.
But since they have all colluded to provide us only with super short little switches there is no time for the feedback to do its job. There is no time for the spring to do its job. 4mm of travel is entirely toooooo tiny.
I would love to build a long switch and build a keyboard out of them.
Ah... I could understand Longer than 4mm, but "not much longer" because there's a a certain point where, if you have the travel be too long, your fingers would be in a "hole" whereby finger lifts would push against the shafts of the "above" switch or cap
-
I was just thinking about the normal human response times and how long you have to respond to a tactile or audio que from a switch.
It seems like if you type at a somewhat slow 30 WPM, you have an average of about 150 keystroke per a minute. To put that another way, a key press needs to take less than 400 ms. Of course the key doesn't click until about halfway down, so you have about 200 ms or so to respond to the tactile or audio feedback and stop your finger from bottoming out. Just a rough approximation of course. The few papers I managed to find on the subject suggested that in 200 ms, a younger individual might have sufficient time to respond to the click but an older person wouldn't and much above that neither would.
To me this suggests that the tactile and audio que probably don't really do much to keep you from bottoming out, unless you are a slower and younger typist. For everyone else, it is just about keyboard feel.
Tactile and Audio feedback would help a lot if the keyboard industry would give us long switches.
But since they have all colluded to provide us only with super short little switches there is no time for the feedback to do its job. There is no time for the spring to do its job. 4mm of travel is entirely toooooo tiny.
I would love to build a long switch and build a keyboard out of them.
Ah... I could understand Longer than 4mm, but "not much longer" because there's a a certain point where, if you have the travel be too long, your fingers would be in a "hole" whereby finger lifts would push against the shafts of the "above" switch or cap
Hey that is a very good point! But surely we could get an extra 2mm safely? That would be 50% more travel.
OTOH I have another theory that says crashing (bottoming out) doesn't matter that much as long as you have something soft and squishy and ergonomically safe to crash into.
You can fall off a 100 story building and move towards the concrete street at 120MPH, crash, get up and walk away like nothing ever happened as long you have something soft and squishy to crash into like an air bag deployed by the fire department.
But if you take the same fall and land on the steel plate of a Rosewill keyboard you die. Immediately.
As long as you crash into something soft and harmless you get your tactility without the nerve damage.
The faster you type, the more important it is to have the soft safety-zone crash barrier for your fingers because, the faster you type, the faster your fingers are moving and the faster they move, the more kinetic energy they have when they smash into the bottom.
Making a custom keyboard with padding and off-the-shelf short 4mm switches is something like $100,000.00 less than making one with new long switches. So I may hafta give up my dream of a long switch keyboard...
-
There are (now old) scientific papers with data that back that tactile and audio feedback is better in some way.
I type faster on tactile switches than on linear. A fast typist that is used to tactile switches types fast on tactile switches. It is a matter of habit, and each person has different habits.
-
I just like the sound of a click
I have browns but prefer blues
browns are still satisfying :P
-
There are (now old) scientific papers with data that back that tactile and audio feedback is better in some way.
I type faster on tactile switches than on linear. A fast typist that is used to tactile switches types fast on tactile switches. It is a matter of habit, and each person has different habits.
I think this is probably a little like talking with earplugs in. It just screws you up because you don't get the feedback you expect. I suspect that someone who had always typed on linear switches might not notice the difference.
BTW, those papers can't be that old. Typing on switches hasn't been around that long. Until about 30 years ago, most everyone typed on a typewriter. IIRC the only feedback my mother and father's type writers gave was a nice hefty *THWAK* as they hit the page. They are actually a lot of fun to type on but slower and a lot more work.
-
I learned to type on a manual typewriter and still use buckling springs, so you know where I stand.
However, I would love an extra couple of millimeters of travel, and I love O-rings, so an added 3-4 mm of length with 1-2 mm of it taken up with an O-ring would be exquisite!
-
I don't think they are usefull to know if you have actuated a key or not, but they have a real impact on how typing "feels". So it's not about typing faster, it's about enjoying more or less typing on a keyboard.
-
IBM spent a LOT of money researching buckling springs because they wanted to replicate the feel of a typewriter because without the tactile feel, the typists were not as fast. If tactile feedback were not important, they wouldn't have done it, and Model M's would have faded into obscurity long ago. Instead IBM Model M's are considered the very best keyboard for typing on. Most parts on your computer last a few years, meanwhile early M's are nearing 30. It's not hype, hype doesn't last 30 years.
As for the reaction time thing, this is true, you might go faster than the brain can process the click, but you do develop a sensory or muscle memory of where that click happens, and it is reinforced on every keystroke. Repeat it enough times and you will be able to anticipate it. You don't have to react to it, you learn to anticipate it through the click.
Not every fast typists bottoms out, I don't.
I run extremely short travel switches for that reason. I have just a little over 1mm travel before actuation and about 3mm after (unless I install o-rings, which these keycaps don't work well with).
Fingers are EXTREMELY sensitive, you just have to learn to use it. Look across a panel and it may look smooth and clean but drag your finger across and it may be extremely rough and dirty. Fingers can detect thousandths of an inch differences that you eye cannot see without help such as a magnifying glass or straight edge.
Due to good finger dexterity, even with my short switches, during games, I can get the key just on the cusp of the actuation point and make it float without topping or bottoming out. Handy for those times when you need to just inch forward or feather a throttle. If you doubt this is possible, think about the Playstation controllers with the sensitive buttons. Same thing. It's just a matter of dexterity, however it's something you have to work at, especially at the small distances we are talking.
-
Yes, mouse switches have very short travel and provide feedback. Anybody know why no keyboards with mouse like switches? (besides: "I don't like them" OR "they shouldn't work").
-
Yes, mouse switches have very short travel and provide feedback. Anybody know why no keyboards with mouse like switches? (besides: "I don't like them" OR "they shouldn't work").
Would that be a little like ML switches?
-
IBM spent a LOT of money researching buckling springs because they wanted to replicate the feel of a typewriter because without the tactile feel, the typists were not as fast. If tactile feedback were not important, they wouldn't have done it, and Model M's would have faded into obscurity long ago. Instead IBM Model M's are considered the very best keyboard for typing on. Most parts on your computer last a few years, meanwhile early M's are nearing 30. It's not hype, hype doesn't last 30 years.
I think the part in bold has a lot to do with it. The keyboard was optimized for typists that were used to type writers. In the same way, it seems like a lot of kids these days want desktop keyboards that feel like their laptop keyboards, because that is what they are use to. IMO, people probably type fastest on what they are used to. Assuming, what they used to isn't incredibly terrible.
-
Yes, mouse switches have very short travel and provide feedback. Anybody know why no keyboards with mouse like switches? (besides: "I don't like them" OR "they shouldn't work").
Would that be a little like ML switches?
Definitely not. ML in specs and my opinion feels like a lighter, and shorter travel, Red MX. I hate the clack of the bottom, and it's very light. It is not a membrane either... I still use and love my G84-4100 or ML-4100 (but I don't like the spacing and some of the key placements)
-
Yes, mouse switches have very short travel and provide feedback. Anybody know why no keyboards with mouse like switches? (besides: "I don't like them" OR "they shouldn't work").
Would that be a little like ML switches?
Definitely not. ML in specs and my opinion feels like a lighter, and shorter travel, Red MX. I hate the clack of the bottom, and it's very light. It is not a membrane either... I still use and love my G84-4100 or ML-4100 (but I don't like the spacing and some of the key placements)
Not totally certain but I believe the Omron microswitches in my mouse in principle operate most closely to a buckling spring.
There is a guy out there who decided to make an entire keyboard using microswitches. He describes his switch of choice as feeling somewhat similar to a very light weight buckling spring switch.
Here he discusses his switch choice http://mykeyboard.co.uk/microswitches/
Here is crazy looking home made microswitch based borad http://mykeyboard.co.uk/microswitch/
For some reason I am not surprised that a British person built this thing.
-
Yes, mouse switches have very short travel and provide feedback. Anybody know why no keyboards with mouse like switches? (besides: "I don't like them" OR "they shouldn't work").
Would that be a little like ML switches?
Definitely not. ML in specs and my opinion feels like a lighter, and shorter travel, Red MX. I hate the clack of the bottom, and it's very light. It is not a membrane either... I still use and love my G84-4100 or ML-4100 (but I don't like the spacing and some of the key placements)
Not totally certain but I believe the Omron microswitches in my mouse in principle operate most closely to a buckling spring.
There is a guy out there who decided to make an entire keyboard using microswitches. He describes his switch of choice as feeling somewhat similar to a very light weight buckling spring switch.
Here he discusses his switch choice http://mykeyboard.co.uk/microswitches/
Here is crazy looking home made microswitch based borad http://mykeyboard.co.uk/microswitch/
For some reason I am not surprised that a British person built this thing.
Wow, let me get in contact with this guy, to see if he helps me make one with standard mouse switches...
-
Just tested my reaction time to visual stimuli online, got a average of ~220ms, so this numbers seem to be correct.
But the reaction time to tactile stimuli could be faster, also one can probably lower it by training (see sports...)
-
Just tested my reaction time to visual stimuli online, got a average of ~220ms, so this numbers seem to be correct.
But the reaction time to tactile stimuli could be faster, also one can probably lower it by training (see sports...)
My basic understanding is this. Audio reaction time is the fastest because your hearing is mostly closely connected to the part of your brain that controls such reactions. Tactile sensations have to travel from your finger tips to your brain, and are thus slower. Visual stimulus is a little slower than audio because it has to go to the back of your brain for processing before it reaches the rest of your brain.
NPR's Radiolab program had a segment on this in their most recent program which is sort of what sparked my initial question.
-
Yes, mouse switches have very short travel and provide feedback. Anybody know why no keyboards with mouse like switches? (besides: "I don't like them" OR "they shouldn't work").
Would that be a little like ML switches?
Definitely not. ML in specs and my opinion feels like a lighter, and shorter travel, Red MX. I hate the clack of the bottom, and it's very light. It is not a membrane either... I still use and love my G84-4100 or ML-4100 (but I don't like the spacing and some of the key placements)
Not totally certain but I believe the Omron microswitches in my mouse in principle operate most closely to a buckling spring.
There is a guy out there who decided to make an entire keyboard using microswitches. He describes his switch of choice as feeling somewhat similar to a very light weight buckling spring switch.
Here he discusses his switch choice http://mykeyboard.co.uk/microswitches/
Here is crazy looking home made microswitch based borad http://mykeyboard.co.uk/microswitch/
For some reason I am not surprised that a British person built this thing.
Do not use Omrons and similar for a keyswitch, they are meant for a light touch and not designed for abuse. If you notice on mice, you aren't directly pressing on them, you are using a long, flexible lever, this reduces the amount of force being exerted on them. They use a very small leaf spring inside, and a hard slap, like you can do on the spacebar or enter key, will eventually deform it sooner, rather than later. When this happens you begin to get double clicks, it's a common issue with this sort of switch. Not to mention accidentally dropping something like a cell phone on your keyboard will likely trash the switch immediately. My last few Logitechs died because the Omrons wore out due to playing FPS and a bit too much force.
If you want a short travel, tactile feel and light spring rates, take a look at Jailhouse Blues.
Mine need 1mm of travel to actuate with only 50-55grams of force (which is tuneable). Travel after actuation (which is around 3mm) can be shortened with o-rings. It's a much cheaper, simpler, durable and less expensive option than rigging up Omrons.
-
To me this suggests that the tactile and audio que probably don't really do much to keep you from bottoming out, unless you are a slower and younger typist. For everyone else, it is just about keyboard feel.
I can agree that audio and tactile feedback does not keep you from bottoming out unless you are typing slowly, but they are invaluable assistance to help us learn the amount of force to not bottom out. Over time, I learn to press the keys with a slightly lower force, and so I don't bottom out as often.
-
Yes, mouse switches have very short travel and provide feedback. Anybody know why no keyboards with mouse like switches? (besides: "I don't like them" OR "they shouldn't work").
Would that be a little like ML switches?
Definitely not. ML in specs and my opinion feels like a lighter, and shorter travel, Red MX. I hate the clack of the bottom, and it's very light. It is not a membrane either... I still use and love my G84-4100 or ML-4100 (but I don't like the spacing and some of the key placements)
Not totally certain but I believe the Omron microswitches in my mouse in principle operate most closely to a buckling spring.
There is a guy out there who decided to make an entire keyboard using microswitches. He describes his switch of choice as feeling somewhat similar to a very light weight buckling spring switch.
Here he discusses his switch choice http://mykeyboard.co.uk/microswitches/
Here is crazy looking home made microswitch based borad http://mykeyboard.co.uk/microswitch/
For some reason I am not surprised that a British person built this thing.
Do not use Omrons and similar for a keyswitch, they are meant for a light touch and not designed for abuse. If you notice on mice, you aren't directly pressing on them, you are using a long, flexible lever, this reduces the amount of force being exerted on them. They use a very small leaf spring inside, and a hard slap, like you can do on the spacebar or enter key, will eventually deform it sooner, rather than later. When this happens you begin to get double clicks, it's a common issue with this sort of switch. Not to mention accidentally dropping something like a cell phone on your keyboard will likely trash the switch immediately. My last few Logitechs died because the Omrons wore out due to playing FPS and a bit too much force.
If you want a short travel, tactile feel and light spring rates, take a look at Jailhouse Blues.
Mine need 1mm of travel to actuate with only 50-55grams of force (which is tuneable). Travel after actuation (which is around 3mm) can be shortened with o-rings. It's a much cheaper, simpler, durable and less expensive option than rigging up Omrons.
Thanks, sounds good!
-
Another thing worth looking at:
Seimitsu or Sanwa microswitches (as used in arcade push-buttons)
Around 50g, short travel, probably pretty durable as arcade style button mashing doesn't break em.
-
Maybe its a learned behavior when it clicks? Maybe because the distance traveled doesn't change?
-
Another thing worth looking at:
Seimitsu or Sanwa microswitches (as used in arcade push-buttons)
Around 50g, short travel, probably pretty durable as arcade style button mashing doesn't break em.
I'll check em out, but thinking about it ... more money, time and know how needed.
What "Leslieann" did (with Jailhouse Blues) makes sense because we already have a keyboard and the switches, we just spend a little time/money modding them.
Regarding mouse micro switches, those would be my reference regarding; clicking sound volume, feedback AND travel.
I guess they can do the same with a new Cherry MX model/color, but they haven't got a model with the specs, I like (mouse micro switch characteristics).
-
From the description, it doesn't seam like the Jailhouse Blues would "click". The part that makes them "click" is fixed, so they should be more like browns.
-
My basic understanding is this. Audio reaction time is the fastest because your hearing is mostly closely connected to the part of your brain that controls such reactions. Tactile sensations have to travel from your finger tips to your brain, and are thus slower. Visual stimulus is a little slower than audio because it has to go to the back of your brain for processing before it reaches the rest of your brain.
When it comes to people who type on tactile switches most of the time, I do not believe their fingers are responding to the presence of the tactile bump so much as they are responding to the expectation of it. That is, the fingers know from experience where the bump is and when they will hit it and how much force it takes to push through it. It is all muscle memory. Someone who has typed on nothing but linear switches for a long time will be a bit slower typing on cherry browns, not being able to rely on their muscle memory and instead having to rely directly on tactile response. Once their fingers have learned the switch, they will be fast again.
So, I don't think the tactile bump is inherently better or more ergonomic, but it does provide a useful reference point around which the fingers can build their muscle memory, whereas no such clear defining point exists on linear switches. But some are still able to type well on them, so it is probably different for each person.
-
From the description, it doesn't seam like the Jailhouse Blues would "click". The part that makes them "click" is fixed, so they should be more like browns.
Browns don't really click as the bump is simply too small, Blues make a loud click because the bottom half slams into the bottom of the housing.
Jailhouse Blues and Clears do click, but it's a much softer click than a normal Blue, the sound is closer to the sound of a muted mouse click or retractable ballpoint pen makes.
-
From the description, it doesn't seam like the Jailhouse Blues would "click". The part that makes them "click" is fixed, so they should be more like browns.
Browns don't really click as the bump is simply too small, Blues make a loud click because the bottom half slams into the bottom of the housing.
Jailhouse Blues and Clears do click, but it's a much softer click than a normal Blue, the sound is closer to the sound of a muted mouse click or retractable ballpoint pen makes.
LeslieAnn: When/If you have time, could you post/attach the sound?
-
LeslieAnn: When/If you have time, could you post/attach the sound?
If I can figure out a way to do it.
Someone with clears might be able to do it better if they have some recording equipment.
-
LeslieAnn: When/If you have time, could you post/attach the sound?
If I can figure out a way to do it.
Someone with clears might be able to do it better if they have some recording equipment.
If you have a laptop and Windows you probably have a microphone at the top. In which case use "sound recorder" to record.
If no microphone, you can use the one from a headset or a stand alone mic. $1 from EBay: http://www.ebay.com/sch/Computers-Tablets-Networking-/58058/i.html?_from=R40&_nkw=microphone&rt=nc&LH_BIN=1
-
I use regular keyboards at work, and really don't think I am any faster or any slower depending on whether or not I use a mechanical keyboard. I am faster on those keyboards that I use regularly though.
I think it is all about feel, which in a way facilitates typing as well. The only thing I know is that I tend to type slower on Cherry MX blacks (even slower than rubber domes) just because the key is too heavy for me and I am naturally not a heavy typer.
I also don't think it is true that any fast typer bottoms out on every keystroke. The fastest typer I know is my sister, and she types very light. I'm always a little jealous of her ninja typing skills. I spend more time behind a keyboard than her, and for some inexplicable reason she can type like that while my typing is rather pedestrian in comparison.
-
I would say that the louder it is, the less likely you are to make any mistakes whilst typing.
I've recently got a BS keyboard and my WPM has improved dramatically in comparisons to the Grey ALPS i had on my AT102W.
-
LeslieAnn: When/If you have time, could you post/attach the sound?
This is with a cell phone held right next to it and this is just the actuation, not the key bottoming or topping out. Sound compared to the real thing is quite good.
Volume is about the same as a ballpoint pen or mouse button being clicked, maybe a little quieter. Keep in mind you still have the much louder topping and bottoming out noise, just as any other switch.
Edit: making a better copy
Update:
New copy attached. First few clicks are me just trying to isolate the actuation sound, the last ones are normal keystrokes. This is with no o-rings.
-
The "feedback" is bottoming out. No reason to have extra feedback in the middle of the keystroke.
-
The "feedback" is bottoming out. No reason to have extra feedback in the middle of the keystroke.
I completely disagree.
Like I said earlier, I can feather by just riding the tactile bump while gaming.
Personally, I feel like a chav slamming keys like an ape. I don't see the point of a mid-level actuation point if I plan on bottoming out anyhow. May as well go back to rubber dome which at least gives you a nice padded bottom.
-
I bottom out my Reds but I have been slowly adapting. I can tell I am bottoming out a lot more softly than when I got it and compare to my old rubber dome board. It takes some time to learn to trust that you don't have to be pressing down at all at the end of the stroke to get the key to register. I could see getting to the point of not bottoming out at all even without the tactal feedback.
-
Yes, mouse switches have very short travel and provide feedback. Anybody know why no keyboards with mouse like switches? (besides: "I don't like them" OR "they shouldn't work").
Would that be a little like ML switches?
Definitely not. ML in specs and my opinion feels like a lighter, and shorter travel, Red MX. I hate the clack of the bottom, and it's very light. It is not a membrane either... I still use and love my G84-4100 or ML-4100 (but I don't like the spacing and some of the key placements)
Red? More like browns/clears IMHO. Definitely not lighter than reds.
Just tested my reaction time to visual stimuli online, got a average of ~220ms, so this numbers seem to be correct.
But the reaction time to tactile stimuli could be faster, also one can probably lower it by training (see sports...)
I was surprised to get about 180 ms with a little bit of concentration, but hey! pro gamers probably attack 100 ms.
-
The "feedback" is bottoming out. No reason to have extra feedback in the middle of the keystroke.
You are partially correct. Bottoming out is extremely useful feedback. I realized this when I tried soft-landing pads, they were so soft that I couldn't tell when I was bottoming out, and I ended up mashing the keys more!
However mid-strock feedback is useful as well, since it tells you the point of actuation.
Summarizing:
mid-stroke feedback is useful because lack of it tells you that you need to use more force.
bottom-out feedback is useful because it tells you that you need to back off.
I am using cherry red switches, with an electronic click sound (kinesis advantage) and it works out beautifully.
Regarding response times:
It doesn't matter much, the click won't save your current keystroke that you might measure the miliseconds for, but as you type, it guides your overall force, so the feedback DOES help.
Also I find our audio memory very important; I can tell that I missed a key after typing a password for instance, if I didn't hear all the clicks I was expecting. I find myself backspacing and adding the missing keystroke without looking, pretty successfully when that happens.
-
LeslieAnn: When/If you have time, could you post/attach the sound?
This is with a cell phone held right next to it and this is just the actuation, not the key bottoming or topping out. Sound compared to the real thing is quite good.
Volume is about the same as a ballpoint pen or mouse button being clicked, maybe a little quieter. Keep in mind you still have the much louder topping and bottoming out noise, just as any other switch.
Edit: making a better copy
Update:
New copy attached. First few clicks are me just trying to isolate the actuation sound, the last ones are normal keystrokes. This is with no o-rings.
Thanks!
-
Honestly, I bottom out a WHOLE lot on all my boards. So, It doesn't matter too much which switch I use.