Robotic testing finds touchscreen inaccuracies at edge of iPhone display
The touchscreens on Apple's flagship iPhone 5s and mid-range iPhone 5c show significant degradation in touch detection accuracy along their edges, which can affect hitting outside keys on the iOS virtual keyboard, new test data shows.

Green points indicate measurements within ? 1 mm. Red points indicate measurements in excess of 1mm. Black circles show the robot's touch point. | Source: OptoFidelity
Using a robotic finger and a specialized suite of test software, Finnish automated testing company OptoFidelity found that Apple's latest handsets accurately detect touch inputs only across a small swath of their displays, roughly equating to the location of the on-screen keyboard.
The iPhone 5s and 5c, according to the company, suffer from "extremely bad" touch performance near the edges of the display. This makes it difficult for users to hit outlying keys on the virtual keyboard, like Q, O, and P, the report says.

Touch accuracy measurements overlaid on an iPhone 5s keyboard, top, and a Galaxy S3 keyboard, bottom | Source: OptoFidelity
OptoFidelity compared the iPhones' displays to that of Samsung's Galaxy S3 and found that the Korean phone offered significantly more accurate touch detection across its entire touch surface.
"Because the touch accuracy is more even in Galaxy S3 than in iPhone 5s and iPhone 5c, you get a lot less typing errors, and letters which are close to the edge are working better," the report said.
An example of OptoFidelity's testing system used on Windows 8 tablets
The test methodology considers a difference of less than one millimeter between the robot's reported position and the touch screen's reported position to be "accurate."
The analysis marks a change from 2010, when testing showed that the iPhone display was by far the most accurate among smartphones then available on the market. That test from three years ago was conducted by drawing lines on the screen, and measuring how straight they would be.
While the latest test suggests the outside of Apple's iPhone displays may experience inaccuracies, the iPhone still leads the way when it comes to responsiveness. One study published last month by Agawi found the iPhone 5 display to be twice as fast as competing Android-based handsets.

Green points indicate measurements within ? 1 mm. Red points indicate measurements in excess of 1mm. Black circles show the robot's touch point. | Source: OptoFidelity
Using a robotic finger and a specialized suite of test software, Finnish automated testing company OptoFidelity found that Apple's latest handsets accurately detect touch inputs only across a small swath of their displays, roughly equating to the location of the on-screen keyboard.
The iPhone 5s and 5c, according to the company, suffer from "extremely bad" touch performance near the edges of the display. This makes it difficult for users to hit outlying keys on the virtual keyboard, like Q, O, and P, the report says.

Touch accuracy measurements overlaid on an iPhone 5s keyboard, top, and a Galaxy S3 keyboard, bottom | Source: OptoFidelity
OptoFidelity compared the iPhones' displays to that of Samsung's Galaxy S3 and found that the Korean phone offered significantly more accurate touch detection across its entire touch surface.
"Because the touch accuracy is more even in Galaxy S3 than in iPhone 5s and iPhone 5c, you get a lot less typing errors, and letters which are close to the edge are working better," the report said.
An example of OptoFidelity's testing system used on Windows 8 tablets
The test methodology considers a difference of less than one millimeter between the robot's reported position and the touch screen's reported position to be "accurate."
The analysis marks a change from 2010, when testing showed that the iPhone display was by far the most accurate among smartphones then available on the market. That test from three years ago was conducted by drawing lines on the screen, and measuring how straight they would be.
While the latest test suggests the outside of Apple's iPhone displays may experience inaccuracies, the iPhone still leads the way when it comes to responsiveness. One study published last month by Agawi found the iPhone 5 display to be twice as fast as competing Android-based handsets.
Comments
Well. Good thing I'm not a robot!
Hmm. . .
So faster mistakes then according to AI. What an odd way to attach a positive to a negative finding (if true). :err:
What I've seen in iPhone 5 and 5s (but not in 4S and previous versions) is a greater probability of trying to type M, L or N and ending up deleting the previous character instead.
It's about time this type of testing is done and reported.
Touchgate!
Also, are the edges 'numbed' on purpose, to prevent accidental touch from the bezel?
The apps I used the most have placed their buttons (text) in bad locations.
The Music apps "Now Playing" and "Back to List."
The Nike Running app's "Run" button.
Instacast's "Resume playing full screen" and "Play episode" buttons.
Very annoying.
I'm not sure how the problem could be with typing text, given the only letters close to the edge are Q and P. I'm sure there are apps where buttons could be in problematic areas, but I've seen no issues on my 5s. Even swiping on web pages to go back and forth show no issues for me, where you have to start at the very edge of the screen.
"Everybody can do this same test with a real device. For example: you can try to tap the letter P
(English keyboard) when your finger is close to the right side of the screen. In many cases smartphone
does not react to your tap because touch result is reported outside of the button (as seen on Image
6). Also letters I, O and P are not responding as well as letters E,R,T,Y and U because the touch
accuracy is not constant. This will cause extra typing errors that are not actually caused by the end
user but the smartphone. "
http://www.optofidelity.com/wp-content/uploads/2013/10/OF_iPhone5C_vs_iPhone5C.pdf
Human fingers press on touchscreens at different angles, depending on how far you are reaching and how you are holding a device (try it yourself). It's not inconceivable that this is by design, because our fingers do not operate like a thin pin coming straight down onto the screen. With different angles, we have a different "feel" for where our fingers are supposed to be pressing.
That said, I have no information that would suggest Apple has actually designed for this, and it seems to me that if that were the case I would expect a more symmetric pattern than what this test is showing.
Just wanted to make people think a little bit. There are a lot of explanations for the result, including the possibility of a bad batch of sensors that doesn't represent the entirety of 5s and 5c phones.
The nike+ app! I thought I was losing my mind when I thought that run button was unresponsive.
Compared to my prior 4s, on the 5s I have a noticable increase in difficulty selecting things that have small-ish selection areas.
Example, the new swipe left to delete in email, constantly must redo. Ususally end up accidently opening the mail. Im also seem to be tapping multiple times to open something. That was not the case on the 4s(ios6).
Whether this is 'just me', new to iphone 5 screen or is it an iOS7 issue, dunno.
Is it overwhelming enough to take the phone back etc-no; is it noticable and can be a bit irksome-for me yes.
Worked good so far. I call BS on their testing methods. Apple applies a lot of heuristics to the touch input to weed out false touches from what the user actually wants. Thus would make it practically impossible to do this type of automated testing because you have no way of knowing if and under what circumstances the heuristics are doing anything.
Less than 1mm is "accurate"? Where did they get this figure from? Is it a standardized published specification? Who published it? And how come the previous test drew straight lines and this one used a keyboard? How can they compare results from 2010 to now when the testing methodologies are completely different?
Completely bogus study by someone again looking to generate traffic by bashing Apple over some useless metric.
From the report:
"Everybody can do this same test with a real device. For example: you can try to tap the letter P
(English keyboard) when your finger is close to the right side of the screen. In many cases smartphone
does not react to your tap because touch result is reported outside of the button (as seen on Image
6). Also letters I, O and P are not responding as well as letters E,R,T,Y and U because the touch
accuracy is not constant. This will cause extra typing errors that are not actually caused by the end
user but the smartphone. "
http://www.optofidelity.com/wp-content/uploads/2013/10/OF_iPhone5C_vs_iPhone5C.pdf
Just tried this on my iPhone 5 without incident. Every tap was correctly registered as intended.
"analysis marks a change from 2010" -- no, not really, it's not even remotely the same test
Also: users are not reporting inaccurate touches. And this is measuring dots. Multitouch doesn't return dot accuracy by millimeters to apps; it registers fat finger presses in oval areas, and does lots of math to determine the intent of the press or swipe or tap.