The iPad has a reputation for being difficult to type on. It’s generally accepted that the iPad is okay for short emails and notes, but is not suited to longer documents. The anecdotal consensus seems to be that an on-screen keyboard, with no tactile feedback, leads to more errors than a physical keyboard with real keys. Based on my research today, this simply isn’t the case. Instead, it’s the iPad’s software that’s causing the majority of typing errors.
I’ve never learnt to type properly – I use at most three fingers – but I can get around a physical keyboard pretty quickly. As an interface designer, I’m intrigued to see whether an on-screen keyboard really can be a valid alternative for a physical one. As a result, I’ve studied the iPad’s keyboard in quite a bit of detail.
The more I’ve looked into it, the more I’ve suspected that my typing isn’t actually at fault for the mistakes in my iPad documents. To test this theory, I positioned my iPhone 4S (the best video camera I own) over an iPad 2 (supported by a folded Smart Screen), and filmed myself typing some sample text into a Pages document. I decided that Pages, as the de facto (and Apple-created) word processor for iPad, would be the best test of using the device for long-form writing.
Here’s the video of my typing, played back at 12.5% of the original speed so that you can see the individual keys being pressed. I’ve used the on-screen keyboard’s highlighting of pressed keys (fading to a darker grey and back again) to ascertain which keys I have successfully pressed, and in which order. I deliberately used a piece of text I don’t know well, to avoid familiarity; the pauses in the video come from me reading each block of text from the screen of my Mac. (The video is 12 minutes long, and is quite dull to watch in its entirety; it’s included here primarily to illustrate the examples below.)
Here’s what I’m trying to type:
Of the causes which have induced me to print this volume I have little to say; my own opinion is, that it will ultimately do some service to science, and without that belief I would not have undertaken so thankless a task. That it is too true not to make enemies, is an opinion in which I concur with several of my friends, although I should hope that which I have written will not give just reason for the permanence of such feelings.
…and here’s what I end up with on the iPad (with differences from the original text highlighted in red):
Of the causes which have induced me to print this volume image little to say; my own opinions, that t will ultimately do some service to science, and without that belef I would not ave undertaken so tankless a task. That it’s tod rue nt to Mae enemies, is anopinionin which I concur with several of y friends, although I should le that what iave written will nt give just reason fr the permanence of such feelings.
Clearly this is an unacceptable number of errors. The first error occurs just after typing this volume. On the iPad keyboard, I correctly type i [space] have. There’s very little time between pressing h and a, but long enough for the framerate of the iPhone video camera to detect them as being pressed in the correct order.
After I type the letter i, the following space is ignored, as is the letter h. Is this because the iPad is expecting me to select or dismiss an autocorrect overlay for the lowercase i? Or is it because the keypresses are missed by Pages? In either case, Pages doesn’t start processing my keystrokes again until the letter a. It then correctly detects the v, the e and the subsequent space, causing it to display iave; this is autocorrected to image after the ending space is pressed. All of this despite the fact that I typed all of the letters correctly, apart from using a lowercase i.
The second error is a genuine typing mistake, as I fail to hit the space bar between opinion and is. The autocorrect suggestion of opinions is a sensible one, albeit not what I wanted.
The third error seems to be another ignored keypress. I’m typing that it will, and from the video recording, I press all of the keys in the correct order. However, the keypress of the letter i in it is ignored, despite the fact that the keyboard shows it turning grey, leaving me with that t. Likewise after the second l in will, I definitely press space, and the keyboard notes this by turning the spacebar grey, but doesn’t register a space in the document on screen, leaving me with willultimately.
To its credit, the iPad then autocorrects willultimately to will ultimately without showing a suggestion overlay. Nonetheless, it shouldn’t have needed correcting in the first place, as I typed the two words correctly.
Another error – during my typing of the word science, my correct keypress of the letter i (clearly visible on the keyboard) is ignored, resulting in scence being displayed on screen. Again it’s autocorrected, and ends up as science, but should never have been wrong in the first place.
More letters are missed; the i from belief is ignored, and the word isn’t autocorrected, leaving me with belef. The o from not is ignored too, leaving me with nt. (We’ll skip over my slightly embarrassing inability to type the word several – I get it right in the end.)
Yet more characters are ignored in the second half of the text. Here’s the final output text, showing my errors in green; missed-letter errors (not subsequently autocorrected) in red; and missed-letter errors (subsequently corrected by autocorrect) in blue.
Of the causes which have induced me to print this volume image little to say; my own opinions, that t will ultimately do some service to science, and without that belef I would not ave undertaken so tankless a task. That it’s tod rue nt to Mae enemies, is anopinionin which I concur with several of y friends, although I should le that what iave written will nt give just reason fr the permanence of such feelings.
Out of a 435-character text, 30 keypresses were ignored – nearly 7% of the characters in the text. And out of 17 errors in the final text (after autocorrect had helped out), only three were my fault – about my usual error rate on a physical keyboard. (I made five mistakes when typing the same text in Pages on an 11” MacBook Air.)
So it turns out I can type every bit as well on an iPad as I can on a Mac – it’s just that the iPad can’t keep up. If this was fixed, my iPad would be my default typing device on the move; until then, I’ll stick with the MacBook Air.
Update 1: following a few questions on Twitter, I’ve run the same test in other apps (Notes and Mail), and the missing characters problem seems to occur there too. This suggests there might be a lower-level text input issue, rather than a Pages-specific issue. I’ve also tried using a physical Bluetooth keyboard; this didn’t trigger missing characters, suggesting that the issue might be specific to text input from the on-screen keyboard only.
Update 2: a few people have asked if this problem might be specific to my iPad 2. I’ve tried the same test on two iPad 1 devices (albeit without the benefit of slowed-down video analysis), and seen very similar behaviour. Although this isn’t conclusive, it leads me to believe that the problem isn’t specific to my device.
If you’ve found this article interesting, please do feel free to follow me on Twitter, where I post regularly about iOS app design and development.
Fascinating analysis. The difference between iPad and keyboard is the touch screen. Rather than iPad ignore a key ‘press’ it is more likely that you didn’t touch the letter cleanly and with consistent pressure. The errors increase the longer you are typing as you become more adapted to the task. Many people use a Bluetooth keyboard for long typing on touchscreen devices to avoid this problem.
Hi Martyn,
You can tell whether or not a key was pressed by watching the keyboard closely. Pressed keys fade to dark grey and then back to light grey. The main reason for my slow-motion video research was to confirm that the keypresses took place; if you watch the video around the time that the errors occur, you can clearly see the keypresses are acknowledged by the on-screen keyboard, but do not appear in the document.
I stopped using auto correct about 10 minutes after buying my iPad. Just like nearly all users afaik. You might want to redo the experiment with a lower number of errors
Hi Stephan,
I haven’t performed the video recording experiment with autocorrect turned off, but I tried it without a recording, and I still ended up with a lot of missing characters, despite the keys seeming to light up on the keyboard when I pressed them (as far as I could tell without a slowed-down recording to check against, anyway). I’ll try and perform the video test again with autocorrect turned off.
At first I thought that autocorrect might be the culprit, but now I don’t think so; it looks more like there’s a disconnect between what keypresses the keyboard recognises, and what the iPad’s text handling code actually processes.
Finally! I have been searching the web for anyone else who has experienced this problem. I also videos myself to check that it wasn’t me making the mistakes, and I could clearly see the buttons being registered, but the letters not appearing. I also downloaded a third party keyboard (by Batsu), which was awkward to use, but didn’t miss a single keystroke. It is a real problem for me, and stops me from using the iPad for writing anything but a brief email. I can get it to happen every time by typing “Thi” with a capital “T”
Ti
Ti
Ti
It seems to be a caused by a delay when caps lock switches off. For example, it doesn’t happen if I continue to type “thi” after the start of the sentence, but if I force a capital T, it then happens again. Eg
Ti thi thi thi thi Ti thi thi thi thi Ti.
Words I often have a problem with are “the”, “that” and “this”, all very common words, which is very frustrating!
Could you file a radar? Bugreport.apple.com in case you don’t know. As they say, they’re in a bubble and need the pokes.
Hi Cyril,
RADAR already posted (#10636169). My only guess is that it hasn’t been spotted until now, although it does seem odd. That said, I’d always assumed something else was at play – perhaps I was missing the keys? – so it’s plausible it has been missed, if unfortunate.
If my research *is* correct, then my hope is that it can be fixed in software. I certainly don’t want to berate Apple for this if it is a bug; rather, I’d like it to be fixed, so I can stop carrying a laptop around!
As I use computers for input about 1/4 to 1/3 of the time and for viewing (mostly web) the rest of the time I have found it hard to see a significant attraction in a tablet computer: I would find it so frustrating not having access to a half-reasonable input device.
It’s completely predictable that tablets are very substandard in this respect by their very nature:
(1) a screen and keyboard on the same plane is really bad ergonomically;
(2) there is no tactile component to allow the constant, subconscious repositioning of fingers that happens with shaped (concave) keys and ‘marker’ keys (raised dots on ‘f’ and ‘j’);
(3) the haptic feedback is non existent on most current tablets (including iPads). There are one or two more or less in the R&D phase which use vibration to provide feedback, but even so, they probably won’t come close to a well-designed ‘real’ keyboard, which includes carefully engineered features such as the strength and performance of springs, distance of travel, acceleration / deceleration characteristics and the start and end of each depression and release (including the feel of the ‘click’), the audio feedback, etc., etc.
I can’t remember the model, but one of the IBM keyboards was a real classic of superb engineering and design for ‘serious’ typists. Nothing implemented using a tablet comes even close to even an average real keyboard. And, of course, much of the already tiny screen (compared with, say, a good quality 24″ monitor) is occupied by the keyboard when in use. I really can’t imagine anyone who makes serious, heavy use of keyboards, whether writing a novel or programming, or anything in between being satisfied for more than a short while with a tablet once the novelty has worn off.
Of course, for someone who mostly ‘consumes’ content, types the occasional couple-of-paragraph email or enters a few URIs in their browser, the attraction of tablet portability is high, as long as the small screen is bearable.
The points you list are precisely what I thought was at fault with my typing on iPad. However, if my research is correct, then (for me at least) they aren’t anywhere *near* as much of a problem as I first thought. In fact, without the missed keypresses, my iPad typing is every bit as good as when I use a physical keyboard.
It’s easy to assume that the physical characteristics of a new typing setup are to blame for errors and inaccuracies – I certainly did. After this research, I’m no longer sure that this is the case.
Actually it is made on purpose (i.e. a feature), and I’ve just tested it.
Here’s how:
Press any of the keys, it turns grey, now keep holding down your finger and slide over to a neighbouring key. Now release it. The whole key press is ignored.
You need to press *and release* the virtual key while your finger is still over it for it to be passed on the application.
That’s what’s happening in the video, you’re typing quickly and moving to the next key without raising the finger enough from the previous press.
This may be some feature to avoid spurious keypresses.
That’s certainly one possibility, and it’s something I’d considered during my testing. The behaviour you describe is the standard UIButton control behaviour on iOS, and it may be that the on-screen keyboard is made up of a set of UIButtons. However, this button-like behaviour – sliding off a key to negate a keypress – wouldn’t be appropriate for fast typing, which is much more about the initial impact on a key, rather than the subsequent movement.
If this *is* the cause of the missed keypresses, then it’s a mistake to use this kind of touch handling approach on a keyboard. Unlike a considered button press, typing is all about moving speedily around the keyboard. To put it another way: if every keypress registered on the keyboard in my video *had* been used, my document would have had fewer mistakes.
Unfortunately, without a high-framerate video camera, it’s very hard for me to find out if this is the likely cause. Personally I’m not convinced, but it’s certainly worth considering.
Your test seems reasonable and seems to show a problem, although I haven’t experienced that problem myself, and personally I find the iPad’s autocorrect feature to be very helpful.
The idea that Nerg suggested seems like a likely possibility. A big problem with onscreen keyboards is that accidental touches could be mistaken for keystrokes, and the iPad is programmed to ignore sliding keystrokes. The key will turn gray upon touch, but sliding off the key before release will cancel the keystroke. It may be that you just need to adjust your typing style a bit.
Hi James,
I agree about auto-correct being helpful. In fact, it was only when I started using Lion (which introduced global auto-correct for Mac OS X) that I wondered whether something might be up on the iPad. I find auto-correct invaluable on Lion – it saves me a bunch of time correcting mistakes, and has improved my typing. However, on iPad I found the opposite, which led me to wonder if it might be masking an underlying missing-character problem, as described here.
I’m going to run some further tests to see if I can ascertain my finger position when each missing-key touch event ends. It’s hard to capture without a high-framerate video camera, but I’m keen to prove (or disprove) the sliding-keystroke theory, so I’ll give it a try.
Ludicrous article. If you can’t touch-type, why do you write about something out of your sphere of experience. Any *real* typist can tell you typing on a physical keyboard is the *only* way to touch-type accurately. Your fingers actually rest on 8 home keys. So-called haptic feedback has nothing to do with it. Write about things you know *something* about, puh-leese!