You mean you don't see how evidence that the droid phones have the worst finger tracking is actually proof that they have the best finger tracking?
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Touchscreen Accuracy: iPhone Compares Favorably to Other Smartphones
- Thread starter MacRumors
- Start date
- Sort by reaction score
You mean you don't see how evidence that the droid phones have the worst finger tracking is actually proof that they have the best finger tracking?
Could it be the capacitance's grid array sensors(nodes) are much to wide apart(dot pitch) than the ones on the iPhone? The Touchscreen's resolution perhaps is too low to interpolate between contact nodes thus it generates aliased xy data points. The drawing software has to spline interpolate between those received touch points that's why you see wavy lines as the average area of the fingertip approach the proximity of the contact nodes. Imagine walking diagonaly on a 1' floor tiles vs 1". There are more points to interpolate
from thus wavy lines seem non appearant. So I don't think it's the drawing app as it uses the OS touch controller API. It could be hardware/chip, driver related.
from thus wavy lines seem non appearant. So I don't think it's the drawing app as it uses the OS touch controller API. It could be hardware/chip, driver related.
I dunno - to me it looks like something in the software stack (probably at the driver level) where it is trying to manhattanize movement for some reason.
If you actually watch the video rather than trying to make excuses, you can see that the tester is clearly making a straight line along the touchscreen of each device. Its a bit of a stretch to claim that his finger is making very regular waveforms and with some of the heights of those waves, you would be able to tell from the video if he was trying to rig the tests.
In addition, the fact that the oscillations are exacerbated when a lighter touch (and thus a weaker signal with which to produce a good signal-to-noise ratio) only helps to confirm the conclusions drawn.
And whether or not the problem is with hardware alone, software alone, or a combination of both is wholly irrelevant if we're only considering which devices and OSes can most accurately reproduce the input from the user.
It is certainly possible that the drawing programs are doing some sort of filtering of their own but unlikely, in my opinion anyway, given that the lines aren't snapping to horizontal/vertical or something. The test would definitely be enhanced by the group producing their own testing software for each OS which directly used the touch events and touch coordinates provided to the application from the OS.
Sorry. Is that a joke or a serious attempt to explain the poor performance of the majority of phones tested? I'd really like to know.
If it is a joke, then you ought to consider adding a smiley. If it isn't, then I feel truly sorry for anyone close to you...
(and no, I'm not joking)
I just want to get the facts. And, as adults are wont to do, I bring up possibilities to encourage thinking and discussion. Kneejerk responses and blind belief don't count. My daily job entails going down multiple theoretical and empirical paths to determine the truth.
In addtion, I don't like publicity ploys. They're all too common these days. The source company themselves say they actually use more scientific methods to check touchscreens, and yet they published the roughest method results, because it makes for better headlines.
It's like last month when Gizmodo checked just two locations in all of Manhattan for a couple of minutes, and declared ATT to have the highest speeds in NYC. Their "report" was all about getting publicity and website traffic, not about good analysis, no matter how much people try to defend it.
I've worked with capacitive screens for about twenty years now. That's probably two decades more than anyone else in this thread. There are lots of factors to consider.
As I read Apple's patent, I see that they use lots of filtering to determine the midpoint of touches. That alone could make their screen _less_ sensitive. However, first I want to do some testing myself to see.
I was not declaring anything at this time... only giving out thoughts to persue.
In addtion, I don't like publicity ploys. They're all too common these days. The source company themselves say they actually use more scientific methods to check touchscreens, and yet they published the roughest method results, because it makes for better headlines.
It's like last month when Gizmodo checked just two locations in all of Manhattan for a couple of minutes, and declared ATT to have the highest speeds in NYC. Their "report" was all about getting publicity and website traffic, not about good analysis, no matter how much people try to defend it.
I've worked with capacitive screens for about twenty years now. That's probably two decades more than anyone else in this thread. There are lots of factors to consider.
As I read Apple's patent, I see that they use lots of filtering to determine the midpoint of touches. That alone could make their screen _less_ sensitive. However, first I want to do some testing myself to see.
I was not declaring anything at this time... only giving out thoughts to persue.
Pay no mind to kdarling. That user has an "anything but apple" policy. I would not be surprised if kdarling is somehow connected with one of the other handset makers and view the iPhone as a threat to their livelihood.
If you look at their tag line: "Somewhere designing devices and UIs for over thirty years". That would point to that conclusion.
You can't be more wrong.
Artifacts are not due to the knockoffs being more sensitive.
Exactly.
Right, that is the type of response that is helping the scientific community. If you do not like a point of view, instead of finding a reason to counter it intelligently, you dismiss it and then attack the person. Classy.
He was simply stating that if your device were too sensitive, it could be picking up too much vibration and movement. Like the slightest changes in the human body and trying to interpret all of it. This is why a mechanical finger would produce much more satisfying results as there would be less variables (screen coatings, temperature settings, vibrations, ergonomics, etc).
The entire idea that a touch screen might be interpolating based on less sensitive data is entirely plausible. Actually having a test that finds the real facts, not just guesses them based off of a quick test (further testing is needed) could really help both handset makers and developers.
Dismissing someone's credibility because their opinions don't match yours is petty and unhelpful. Not exactly "Think Different" is it?
I agree.
However, dismissing someone's argument because they are palpably talking out of their arse is entirely appropriate!
No, the person is not joking.
Too bad the forum doesn't have any tosser alerts.
Most people are up for intellectual discussions. Saying a cat is a rock isn't a discussion, it's insanity.
MOTO has mentioned the scientific reason for the test.
Right ? The iPhone fails horribly on the edge of the screen in this test. The lines curve and there's nothing showing up a few mm off all 4 edges.
Read the test results, the curve in the iPhone lines is not done by the finger, they moved in a straight line. The curves are put there by the iPhone.
So the verdict is more along the lines of : No one has gotten touch technology right yet.
I'd have to say the argument is valid. The curving lines along the edges of the iPhone screen could be caused by such an interpolation of sensors.
This technology has much room to grow, and it will be interesting how the balance of touch input and sensitivity is reached in future touch devices with SPECIFIC tasks.
I was thinking that the curviness is there by intention near the edges because there isn't a reason normally for you to slide your finger off screen to select something. One handed use of the screen is obvious use case for this scenario.
From the video I would say the Droid Eris had the most accurate response according to the "tests" but the tester did not agree.
Try drawing with a pencil on paper those same lines. Maybe then you can see the variety in which the hand shakes and lacks precise control.
A test with a mechanical finger that uses a control such as a pencil on a drawing surface to show it's exact tracking to the viewer would give a better demonstration.
Also having the sensors mapped out and actual hardware comparisons made with this information, versus MOTO's opinion would draw a much more precise conclusion about screen sensitivity.
It seems MOTO made this test to try and claim that the iPhone's keyboard was best. This seems faulty for any keys located near the edge of the screen because of the curving.
I don't think there are clear winners or losers in this test because of the faulty scientific logic. If MOTO wanted to test to see which keyboard tracked best, why weren't the onscreen keyboards actually tested with a mechanical "control" finger. Then sample data from input errors using a group of randomly selected users on all devices. That would provide real world results and actually test the hypothesis MOTO seems to have made.
Try drawing with a pencil on paper those same lines. Maybe then you can see the variety in which the hand shakes and lacks precise control.
A test with a mechanical finger that uses a control such as a pencil on a drawing surface to show it's exact tracking to the viewer would give a better demonstration.
Also having the sensors mapped out and actual hardware comparisons made with this information, versus MOTO's opinion would draw a much more precise conclusion about screen sensitivity.
It seems MOTO made this test to try and claim that the iPhone's keyboard was best. This seems faulty for any keys located near the edge of the screen because of the curving.
I don't think there are clear winners or losers in this test because of the faulty scientific logic. If MOTO wanted to test to see which keyboard tracked best, why weren't the onscreen keyboards actually tested with a mechanical "control" finger. Then sample data from input errors using a group of randomly selected users on all devices. That would provide real world results and actually test the hypothesis MOTO seems to have made.
I have noticed the loss of sensitivity on the edge of the iPhones screen quite often. It's a real pain surfing the web and texting sometimes. No one has perfect touch technology on cell phones yet, and I suspect it will be quite a few years before they do.
Wouldn't a screen with a higher resolution show a straighter line because of a lower margin of error due to denser pixels? While, on the other hand, a lower resolution would show curved lines.
Texting? Texting seems fine. I can go right to the very outer edge on any letter or number sitting on the edge, like, Q and P and 1 and 0, etc. Or did you mean something else?
If you look closely at the video demo, the guy's finger really seems especially unsteady while doing the Droid. I'm not accusing of any malfeasance, but it crossed my mind that the difference was so pronounced against the other two phones that I wondered if MOTO had an agenda to prove.
I'd be much more satisfied with the legitimacy of this comparison if they had used some kind of template to draw the lines, and better yet, some kind of mechanical or robotic finger to do the drawing that absolutely guaranteed no human shakey-finger influenced the test. Yes, in real world the human finger is there, but it would be interesting to see if it's the human finger, or the software, or the hardware that causes the apparent problem.
That said, I also wondered if the higher resolution, or higher sampling of the faster processor made the difference here, and if Apple will face the same issue. Who knows, maybe Apple chose to forego higher screen resolutions EXACTLY for this reason...?
I'd be much more satisfied with the legitimacy of this comparison if they had used some kind of template to draw the lines, and better yet, some kind of mechanical or robotic finger to do the drawing that absolutely guaranteed no human shakey-finger influenced the test. Yes, in real world the human finger is there, but it would be interesting to see if it's the human finger, or the software, or the hardware that causes the apparent problem.
That said, I also wondered if the higher resolution, or higher sampling of the faster processor made the difference here, and if Apple will face the same issue. Who knows, maybe Apple chose to forego higher screen resolutions EXACTLY for this reason...?
Malfeasance. I love that word. I don't see it used as often as I'd like. Looks good in red, too.
Finally a quantatitve touch screen test!
This is one of those test that not a single hardware maker is interested in doing. In many cases, touch screen accuracy comes down to if it work or not and does the customer complain. One issue that I had with these no stylus touch screens is accuracy and the fingertip hiding precise actions on the screen.
The "wavy lines" on the one phone is probably from a tracking algorithm not adjusting for the speed of the fingertip moving across the screen. Thus creating a graphical beat frequency akin to moving your hand into a smoke cloud at just the right speed to create a ripple effect.
As usual in this business, I have been told that Xerox PARC has some very good fingertip tracking code that filters out these ripples that may have never left their research labs. I'm sure there is a white paper out there with lots of math in it that shows the relationship between finger movement rate and sample rate of the touch surface that nips this problem in the butt. Again, these days, that would be just a few kilobytes of code and a labor if anyone is wiling to implement it.
This is one of those test that not a single hardware maker is interested in doing. In many cases, touch screen accuracy comes down to if it work or not and does the customer complain. One issue that I had with these no stylus touch screens is accuracy and the fingertip hiding precise actions on the screen.
The "wavy lines" on the one phone is probably from a tracking algorithm not adjusting for the speed of the fingertip moving across the screen. Thus creating a graphical beat frequency akin to moving your hand into a smoke cloud at just the right speed to create a ripple effect.
As usual in this business, I have been told that Xerox PARC has some very good fingertip tracking code that filters out these ripples that may have never left their research labs. I'm sure there is a white paper out there with lots of math in it that shows the relationship between finger movement rate and sample rate of the touch surface that nips this problem in the butt. Again, these days, that would be just a few kilobytes of code and a labor if anyone is wiling to implement it.
nip it in the butt?
Xerox PARC? That's an interesting theory. To what degree do you think they would have been involved in such research?
MOTO vs Motorola
Could this be a way for MOTO Development Group to get back at Motorola? Motorola uses the term Moto for some of its phones and the "Hello Moto" slogan. MOTO Development Group might resent this. This "comparison" is a great way to bring attention to MOTO and let people know that they are not associated with Motorola.
Could this be a way for MOTO Development Group to get back at Motorola? Motorola uses the term Moto for some of its phones and the "Hello Moto" slogan. MOTO Development Group might resent this. This "comparison" is a great way to bring attention to MOTO and let people know that they are not associated with Motorola.
I have a motorola droid and every time I do the test in a drawing app, the lines are as straight as I can make my finger move.
So really, I can't say this test is anything but some sort of publicity stunt/etc., especially considering that different phones might have different software interpretation of touchscreen input that is completely independent of the actual touchscreen accuracy.
I am not anti-apple/anti-iphone, but when I use my droid and see completely different results than what the MOTO group did, it makes me question it.
So really, I can't say this test is anything but some sort of publicity stunt/etc., especially considering that different phones might have different software interpretation of touchscreen input that is completely independent of the actual touchscreen accuracy.
I am not anti-apple/anti-iphone, but when I use my droid and see completely different results than what the MOTO group did, it makes me question it.