Latest in the petty touchscreen superiority dispute

[mavere]

The touch skew is a "feature" to compensate for finger angles, and apps can turn it off if they need to.

It was also part of the iPhone 1's marketing pitch in that Apple tuned the input response for real people and real world usage. At the very least, this has been a non-issue for 5+ years...

I wonder if Android apps can "turn on" faster response time?

 

[RampantAndroid]

Having worked on touch screens and having some patents there...this isn't surprising. It's not uncommon for touch screens to be calibrated at the factory, and Apple may have done this intentionally.

The real test would be pulling the phone apart and accessing the screen directly to see what it's getting. Who makes the touch controller anyway?

 

[Mopetar]

Kind of interesting. I remember when Apple used to have the best touch-screens:
http://www.cnet.com/8301-31045_1-10431635-269.html
https://vimeo.com/10367683#

 

[Suspicious-Teach8788]

You really didn't contribute anything with this either. And just fueled his claim.

Uh. I posted a response regarding this issue. What I'm saying is what matters is real world performance, and the real world testimony regarding the iPhone's touchscreen performance has been great.

As other users have said this is just a meaningless jab at Apple, and would not have come up if it weren't for the touch input lag discussion. It's clear that one has been a problem and is at the heart of why many users complain of lag on Android phones, whereas this inaccuracy is a pure red herring.

The OP doesn't do us any favors by saying this is a petty dispute. It was never about superior touchscreens. It was about revealing issues with the touchscreens. It's as if he can singlehandedly dismiss the input lag issue like it doesn't matter at all.

 

[s44]

It's as if he can singlehandedly dismiss the input lag issue like it doesn't matter at all.

LOL

You stopped reading that thread, didn't you? Samsung beats Apple on input lag.

 

[Belegost]

The touch skew is a "feature" to compensate for finger angles, and apps can turn it off if they need to.

I don't see anything in that link which confirms what you state, maybe you linked the wrong thing?

But, as much as I hate having to be on Apple's side, I think you're essentially right. Let's look at the image of the locations overlaid on the keyboard

Note how near the center of the keyboard the dots are centered, but as you move away they are increasingly putting the tap farther toward the edge than was actually touched. This works with visual feedback, when a person is typing the phone will generally be held such that it puts the keyboard in the center of the FOV. As the digit is moved off the center line the visual cue will increasingly be perceived as the digit being farther towards the edge of the device that it actually will be because the brain will tend to interpret motion as being along the line connecting center of vision and the digit.

This will be further compounded by the natural tendency to "stay within the lines" which will have an untrained user trying to keep the perception of the digit within the bounds of the phone (this is also mitigated by the use of the reviled bezel), but due to the width of the digit and the angle of the visual cue that will tend to cause the touch to occur closer to the center than the user expected it to be.

I have a feeling that Apple did work on this with human factors specialists and purposely designed their touch feedback accordingly to make up for the way humans perceive the world without the user needing to go through a learning period.


Ugh, writing that makes me feel dirty...

 

[Suspicious-Teach8788]

LOL

You stopped reading that thread, didn't you? Samsung beats Apple on input lag.

The latest Note 3 wins, yes I do read it. The point of the thread wasn't that the Note 3 is better and therefore there are no input lag issues. The point of the lag was to discuss input lag, and maybe why it makes sense that some Android phones do lag quite a bit, which fueled the fire for the perception that Android = laggy. Older Android phones suffer from the issue more, and it also depends app to app. I mentioned in that thread that certain apps like Maps will lag like hell and I suspect the input lag is well over 100 ms there.

Regarding the Note 3, I agree that it's a great device. And maybe Android has improved greatly on the input lag issue, but it's still been a complaint in the past. Heck, if you see this video I posted (http://www.youtube.com/watch?v=xlvki2Imfaw&t=30s), you can see the Nexus 7 2013 lagging badly in Maps.

The thing about this thread is that we've uncovered something, and to me this isn't a huge issue or something that has been bugging users on the iPhone, which is why I say that real world effects need to be investigated. The only people quick to jump in this thread are the ones who are going "HAHA APPLE"

With that said I'm interested in these implications of inaccurate touch screen. I haven't noticed it yet personally, but maybe there's something I'm missing? I guess why I'm skeptical is that for years now the keyboard of iOS has been heralded as a great keyboard and one of the best. Sure it's missing features like gesture typing, etc, but in terms of accuracy and and typability, it's always been good. Even though I find the Nexus 4's screen incredibly nice to type on, and the iPhone 5 feels small, I hit the right buttons surprisingly well.

I don't see anything in that link which confirms what you state, maybe you linked the wrong thing?

But, as much as I hate having to be on Apple's side, I think you're essentially right. Let's look at the image of the locations overlaid on the keyboard

Note how near the center of the keyboard the dots are centered, but as you move away they are increasingly putting the tap farther toward the edge than was actually touched. This works with visual feedback, when a person is typing the phone will generally be held such that it puts the keyboard in the center of the FOV. As the digit is moved off the center line the visual cue will increasingly be perceived as the digit being farther towards the edge of the device that it actually will be because the brain will tend to interpret motion as being along the line connecting center of vision and the digit.

This will be further compounded by the natural tendency to "stay within the lines" which will have an untrained user trying to keep the perception of the digit within the bounds of the phone (this is also mitigated by the use of the reviled bezel), but due to the width of the digit and the angle of the visual cue that will tend to cause the touch to occur closer to the center than the user expected it to be.

I have a feeling that Apple did work on this with human factors specialists and purposely designed their touch feedback accordingly to make up for the way humans perceive the world without the user needing to go through a learning period.


Ugh, writing that makes me feel dirty...

Interesting that this is the case. Given our thumbs work differently at the edges of the screen versus the center, you may have a point. It sounds like Apple did some testing here.

 

[mavere]

I don't see anything in that link which confirms what you state, maybe you linked the wrong thing?

The relevant part was a posted remark from an Autodesk engineer. I wasn't able to directly link to it:

this offset issue is caused by iOS when it compensates for angle of refraction, so the event positions have an offset, bigger the closer you go to the top of the screen.
This becomes evident with pens that have a clear disk.
There is a technical reason why other apps behave correctly and SketchBook doesn't (for now) but I can't discuss the details.

We will look into this but cannot promise when this fix will be released.

Ivan Pinzon
Principal Engineer
SketchBook Dev Team

Adonit pens (with their clear disk points) have been used on iOS devices through note taking, handwriting apps for a while now. Clearly, this "issue" is a non-issue when developers want it to be so and a UX feature otherwise.

 

[dguy6789]

Meh. Every single touch screen issue that has been "unmasked" in the past three+ years has been a non issue that affects virtually no one.

 

[Cerb]

So it's not just me. I was ready to curse the 5S when I was trying to set it up for a non-techie. Just entering the local wifi password was a PITA, despite not having any problems with most phones, including older iPhones. It was mainly the keys towards the edges. No thumbs.

Gotta love the silliness, though:

iOS is designed to compensate for the angle that it expects a phone to be held at, which suggests that some of the discrepancies between screen accuracy between the iPhone 5s/5c and the Galaxy S3 may be intentional on Apple's part, in order to provide a better experience for users that are not holding their phones in static positions.

Meanwhile, if you do hold the phone in a static position, so as to keep a decent grip on it, or try to use a stylus (see other link), you're clearly using it wrong .

If it truly is an intentional usability tweak, that is not a net positive for all users, there should be a user option to disable it, for users whom it gets in the way of. Just like the too-fluid motion and parallax effect problems, they really need to put their UI/UX decisions through some usability filters that involve more than just an average or median archetype. Jobs had a good sense for these things. Without his tyranny, they should really look into processes that can help find such problems before release, and either nix them, or offer ways to turn them off.

Also, direct link for the jot pen issue:
https://getsatisfaction.com/sketchb...stylus_tracks_at_the_top_of_the_stylus_center

 

[magomago]

I'm surprised no one has brought this up yet, but....Thinking about the results critically, all I know is the following:
Accurate within 0.5 mm (Could be from 0 to 0.5 mm)
Accuracy is > 1mm (could be 1.1mm to infinitty)


This kind of binomial answer doesn't really tell us much. Are all the red points within 1.2mm? 1.2mm is still greater than 1 mm! Hell 1.05 is still greater than 1mm, but functionally I'd argue it means little. What about the error of the measuring device; was that considered as well?

Similar logic applies for the 0.5 mm threshold; are all at that boundary condition? We really don't know.

If you want to provide more useful data, looking at the distribution of the errors shows a much more valuable story (especially around where the user hits the key) than simply knowing a binary answer that is designed to create confusion. Dare I say publicity for a technology might be the main aim?

 

[WelshBloke]

I don't see anything in that link which confirms what you state, maybe you linked the wrong thing?

But, as much as I hate having to be on Apple's side, I think you're essentially right. Let's look at the image of the locations overlaid on the keyboard

Note how near the center of the keyboard the dots are centered, but as you move away they are increasingly putting the tap farther toward the edge than was actually touched. This works with visual feedback, when a person is typing the phone will generally be held such that it puts the keyboard in the center of the FOV. As the digit is moved off the center line the visual cue will increasingly be perceived as the digit being farther towards the edge of the device that it actually will be because the brain will tend to interpret motion as being along the line connecting center of vision and the digit.

This will be further compounded by the natural tendency to "stay within the lines" which will have an untrained user trying to keep the perception of the digit within the bounds of the phone (this is also mitigated by the use of the reviled bezel), but due to the width of the digit and the angle of the visual cue that will tend to cause the touch to occur closer to the center than the user expected it to be.

I have a feeling that Apple did work on this with human factors specialists and purposely designed their touch feedback accordingly to make up for the way humans perceive the world without the user needing to go through a learning period.


Ugh, writing that makes me feel dirty...

Do they turn that "tuning" off when the keyboard isn't being used?
I'd believe that it was deliberate then, otherwise they are sacrificing accuracy in all other uses when they could have just tuned the keyboard software.

 

[ttott]

S3, is it possible that it's all green?

 

[AznAnarchy99]

It is pretty interesting that these screens aren't that accurate. Honestly, it's a big surprising considering how many people own this device and don't really complain about it.

To be honest, I do notice it coming from a S4 to a 5s.

Not the typing but rather random spots like the "phone" button after pressing "contact" in messaging. There's been a few points when I touched it and it wouldn't respond.

 

[rcpratt]

I think this is actually a rather intriguing issue, but the Apple fanboys are sure obnoxious. Too bad.

 

[phucheneh]

I am the furthest thing from an Apple fan, and this still seems like a dumb 'issue.' I've never noticed accuracy issues on devices I've used. Maybe they sufficiently corrected for an inaccurate screen with programming. Maybe they intended the offset based on what they found to be the most typical viewing angle. Maybe...lots of things.

But what is more definite is that it sure doesn't seem to make a damn bit of difference when it comes to functionality. You'd have to be using a stylus to have any chance of this mattering. The coding for how your comparatively fat thumb/finger press is interpreted is going to be far more important than these neglible differences in the hardware.

 

[lopri]

Isn't this more of a hardware thing than the OS thing? My old Nexus 7 was way more sensitive to touch than my first Nexus 4 which had an issue on the bottom left corner.

 

[lopri]

BTW, is there a technological barrier to making more accurate styli than the ones out there? Out of curiosity. I know a stylus will not match the accuracy and versatility of digitizers, but it is curious why no one even seems to try to make one with more durable and skinny tips.

 

[RampantAndroid]

BTW, is there a technological barrier to making more accurate styli than the ones out there? Out of curiosity. I know a stylus will not match the accuracy and versatility of digitizers, but it is curious why no one even seems to try to make one with more durable and skinny tips.

You can't make it skinny - it has to be a conductive mass that's around 3-4mm in diameter at least. Touch screens are designed to detect things 3-4mm and up (6.5mm back when I was doing stuff with them) - they cannot detect small pointy items. If you need that precision, you need a digitizer.

 

[MrX8503]

Edit: Although, looking at the pattern on the iPhone with how the touch points begin to spread out, I can't help but wonder a little bit if this touch pattern is intentional on Apple's part.

I believe it is intentional by Apple and if you look at it in that sense, Apple's implementation is smarter.

This would also explain why its hard to hit keys on an iOS device when its upside down.

I think this test is misleading.

 

[dawheat]

I believe it is intentional by Apple and if you look at it in that sense, Apple's implementation is smarter.

This would also explain why its hard to hit keys on an iOS device when its upside down.

I think this test is misleading.

The test is valid, the conclusions people are drawing may be wrong.

 

[MrX8503]

The test is valid, the conclusions people are drawing may be wrong.

Yeah I think the conclusions are wrong.

Looking at the pattern again you can see that it becomes less accurate outside the easy tap area. This would decrease incorrect taps.

What determines easy tap areas? Possibly viewing angle of the phone itself.

 

[Suspicious-Teach8788]

The test is valid, the conclusions people are drawing may be wrong.

Yup this was my point to begin with. This is far different than the response time test.