@tomer I think if you look at the two extremes. Having about 1600 pixels, it would be very easy (have a large margin for error) to hit the right pixels and to show up perfectly straight. More pixels, means a more precise alert to a screen's or a robot's impreciseness. So an perfectly precise screen approaching an infinite amount of pixels, would be able to show any errors a robot made.
For the same reason though, I would assume if Apple's screen (or the robot) were imprecise, it's errors would be exaggerated more than the another screens.
I'm not sure if it becomes easier or harder to fabricate a more accurate screen at higher or lower resolutions and I think that has some relevance to the conversation since we want large, beautiful screens with loads of information but it'd be good to know if there is that inherent trade-off.
@tomer Note that this has VERY little to do with the screen itself, and more to do with the heuristic functions that the OS uses to interpolate the input. So you can look at it two ways. Either the iPhone has the best heuristics and can "figure out what you meant" the best. Or the Droid does the least manipulation and doesn't try to figure out what you meant and just takes the raw (or almost raw) data for face value.
Unless you're trying to draw straight lines, it doesn't really matter one bit. What matters is can you click on a link. What matters is if your touches register in a reasonable amount of time. What matters is that you have a good experience.
I'm starting to get tired of all the false conclusions people are drawing from their experiments/research. Just because a phone doesn't draw a perfectly straight line doesn't mean the screen is "bad". It doesn't mean that the screen is any less usable. All it means, is that the screen/os combo is not as good at drawing straight lines. Without a LOT of further research (possibly mating a Droid screen to the iPhone OS, doing other point tests, etc) you cannot draw the conclusion that one screen is better than the other (in an absolute sense). If you're going to portray what you're saying as some kind of fact, at least learn statistics and proper research methods...
"Note that this has VERY little to do with the screen itself, and more to do with the heuristic functions that the OS uses to interpolate the input. So you can look at it two ways. Either the iPhone has the best heuristics and can "figure out what you meant" the best. Or the Droid does the least manipulation and doesn't try to figure out what you meant and just takes the raw (or almost raw) data for face value."
Would you then like to explain how the other devices using the android OS performed better than the Droid?
So the droids screen is crap compared to the competition. All that matters is whether you are happy with it or not.
@ircmaxell Assuming Android uses the same methods on all the device to 'interpolate' your input (I don't know enough about the inner workings of how this would work to say for certain) then surely it can be said that the only differentiating factors are the screen resolution and the quality of the touch sensors (obviously the same can't be said about the iPhone).
However, does it even matter if it the OS, the screen or the sensors that are at fault? What matters is the end experience and whilst you may be right in saying individual touches (i.e. clicking buttons) would not be affected, anything more than that that involves handwriting or drawing of any kind obviously would. Deciding whether those features are important would be up to the individual.
I personally would be willing to sacrifice some accuracy in the touchscreen in return for a (arguably) better device like the nexus one (I would like to see how the Desire does though). Having said that, the Droid did perform really poorly so if I were to consider it I would have to try it out myself.
@GmanC "Would you then like to explain how the other devices using the android OS performed better than the Droid?"
Sure. First, the human test was performed inconsistently. I pulled up simplydraw on my Droid, and was able to make almost perfectly straight lines with no effort. Nothing like what they produced on the Droid (I did have a slight waviness to it, but no erratic jumps). Second, each screen gets its own drivers. Drivers interpolate the electrical signals and convert it to information the OS can use. So drivers could be an issue. But the important one is that the robot tests performance of a point contact. None of these phones were designed for input from a point contact (they are designed for fingers after all). The fact that the Droid's screen is worse at registering point contacts says nothing about it's ability to register fingers (which is why the conclusion that was drawn from these tests is invalid).
This comparison is akin to saying that Verizon service is crappy because it doesn't work in *Insert country here*. You're comparing an edge case that it was never designed for, and using that to draw a conclusion about the product itself. It's not valid reasoning. That's all I'm saying...
@GmanC I'll gladly explain why the other devices running the Android OS show different results, assuming that we aren't ruling out that this is mostly (or purely) a software issue.
The Motorola Droid, Droid Eris, and Nexus One run Android, but have completely different touch screen drivers and kernel branches. Different drivers equals different results. It really has nothing to do with the Android specific parts of Android (if that makes any sense) - the Android framework just takes the input from the driver and feeds it to applications.
Of course, the Droid does have different touch screen hardware, and you really can't rule that out as a factor as well.
@tomer The handsets have different panels to begin with, so there's part of the problem. Different panels get different drivers. When performing this "test" on my Nexus One, I got no waviness at all. In the first article, you'll see that they used a very "light" touch to generate the extreme waviness. I don't apply much pressure to my phone at all, and it works great. So I don't see this "issue" as being a problem for me. I get 800x480 pixels on a gorgeous AMOLED screen, and when I need to draw a straight line, it registers it for me. Period.
I would also like to see if all panels that use Gorilla Glass have similar issues; that would be interesting.
@ircmaxell I disagree with you. It has EVERYTHING to do with the touch screen. You say that it doesn't matter unless you're drawing straight lines and what really matters is clicking on a link. The test is a representation of the touchscreen's (and associated drivers) accuracy; if your screen isn't very accurate, when you tap that link, it might not register because it thought your finger was slightly off from where it really was, hence the phone not thinking you tapped the link.
You sound like a little kid that just got his feelings hurt and most likely a Droid owner too. If you think you can do better and come up with a better testing methodology, than please do so.
@tomer Not surprising at all. Using my friend's Droid is a painful experience compared to my iPhone, precisely for the reason of all the vagueness of the key input (so many mistaken entries; frustration). This result doesn't surprise me at all. NOW, it makes sense why I find the Droid such a trial to use: because the screen sucks!! Curiously, it also explains why I like using the Droid Eris so much more than the Droid (he has both). I knew it wasn't just "me".
Honestly while this does show the iPhone is better, it isn't really indicative of performance as a whole.
If this was the complete answer, do you really think the Droid would have A) sold as many units, and B) not have had an outrageously high return rate?
If you're buying a phone to play tic-tac-toe, then most definitely, you want the iPhone, but honestly, can you (using your fingers, and only your fingers) draw a perfectly straight line?
@interconnect It's not about coming up with a better testing methodology. It's about figuring out what the testing methodology can tell you. Figuring out its limitations and assumptions. Considering that the test includes the underlying OS (drivers, kernel modules and such), it isn't so much a test of the screen, but a test of the screen and the software. And since smart software behaves differently in different "modes" (single touch vs touch and move vs multiple touches vs multiple touch and move), comparing drawing a line to clicking is not a valid argument. Now sure, it's possible that what you say is true (that one is inherently inaccurate, and the other is much more accurate), but the test provided here didn't test that. It tested drawing lines. Period. If he had it press the screen and create a matrix of dots, then you'd have an argument. But he didn't. He drew lines. It's a common theme lately to misinterpret results, and to me that's what's happening here. All that you can conclusively say is that the Droid performed the worst in this particular test. It's flawed logic to draw any kind of generic conclusion from this test alone...
And yes, I do own a Droid. And I have NEVER experienced any issues with the touch screen. It's always been 100% accurate for me. To the point that even clicking on pagination links (1 2 3 4 5) is simple at the widest zoom. That's why I am so adamant to protest the conclusions that people are drawing from this test. Because I know from my own experience that those conclusions are wrong. And combine that with the fact that they are not logically derivable from the research that was done, and you understand my objections.
@tomer yet the Droid apologists will explain away the results:
"iphone has less resolution" - well, that would only make the screen less accurate no? and yet we see the results.
"test is meaningless" - i beg to differ... if the screen won't register a tap where i actually tapped, then i'm going to have to tap again, which is inefficient.
"it's not the Droid itself, it's the drivers..." - a distinction without a difference. bottom line is the experience is inferior.
Over all these things are not a big deal to anyone but the apologists twisting the results and explaining them away are so damn annoying.
@Student Driver: It's not light pressure that makes the test harder to pass, it's the size of the contact point. Compared to that dry, metal tip used in the test, a nice, moist, meaty human finger finger is easily detectable, no matter how lightly you press. Your anecdotes don't change the test results.
@everyone talking about resolution: The display pixels are not the touch sensors! The capacitive sensor grid is completely independent of the display. You could use a display of any resolution in any of these devices, and the test results would be identical.
I'd really like to see a single point touch comparison to see where contact actually was compared to where it registers on screen. The movement pattern could be the same, but creating dots instead of smooth lines.
Stop making excuses for the Droid's horrible tracking. It has nothing to do with the screen resolution. It has everything to do with how well the software tracks your finger on the screen. If the tracking on the Droid were worth a shit, a robot would be able to draw straight lines on it. The end.
I would like to point out that the guy in the video states that the robot pointer is more difficult to detect than a human finger. That means this test gives little insight into actual usability, unless you are a cyborg. It's entirely possible the drivers/OS for the Droid are optimized towards real-world situations (like actual fingers), while the iPhone is optimized for anything you throw at it.
@kyphem "Sure doesn't matter software of hardware, Droid perfroms terrible you can't deny that you can't."
I don't deny that the Droid performs terribly in this particular test. What I do deny is that the results of this particular test are applicable anywhere but within the confines of this test (Basically it's useless to the average person who reads it, since they won't draw lines on the screen with a robot finger)... That's what I'm so adamant about. Not refuting the results of the test (Although I do think they are suspect, and would like to be able to perform them myself), but refuting the conclusions that were drawn from them...
Now that we've thrown 'em off the trail, use the form below to get in touch with the people at Engadget. Please fill in all of the required fields because they're required.
Wow the Motorola Droid did really badly - quite surprising.
Also I guess the iPhone does have the best touchscreen after all.
@tomer
No it has the smaller resolution (4 times smaller the Droid and NExus one) and that is why the points are more precise ;)!
@sliderem Surely a higher resolution would mean more precise points?
@tomer I think with a higher res you would notice the inaccuracy's more so then on a lower-res screen.
@tomer I think if you look at the two extremes. Having about 1600 pixels, it would be very easy (have a large margin for error) to hit the right pixels and to show up perfectly straight. More pixels, means a more precise alert to a screen's or a robot's impreciseness. So an perfectly precise screen approaching an infinite amount of pixels, would be able to show any errors a robot made.
For the same reason though, I would assume if Apple's screen (or the robot) were imprecise, it's errors would be exaggerated more than the another screens.
I'm not sure if it becomes easier or harder to fabricate a more accurate screen at higher or lower resolutions and I think that has some relevance to the conversation since we want large, beautiful screens with loads of information but it'd be good to know if there is that inherent trade-off.
@juanvaldez Fair points.
@tomer Note that this has VERY little to do with the screen itself, and more to do with the heuristic functions that the OS uses to interpolate the input. So you can look at it two ways. Either the iPhone has the best heuristics and can "figure out what you meant" the best. Or the Droid does the least manipulation and doesn't try to figure out what you meant and just takes the raw (or almost raw) data for face value.
Unless you're trying to draw straight lines, it doesn't really matter one bit. What matters is can you click on a link. What matters is if your touches register in a reasonable amount of time. What matters is that you have a good experience.
I'm starting to get tired of all the false conclusions people are drawing from their experiments/research. Just because a phone doesn't draw a perfectly straight line doesn't mean the screen is "bad". It doesn't mean that the screen is any less usable. All it means, is that the screen/os combo is not as good at drawing straight lines. Without a LOT of further research (possibly mating a Droid screen to the iPhone OS, doing other point tests, etc) you cannot draw the conclusion that one screen is better than the other (in an absolute sense). If you're going to portray what you're saying as some kind of fact, at least learn statistics and proper research methods...
@tomer And yet you're still highly ranked for inaccurate information.
I love engadget commenters...
@ircmaxell
"Note that this has VERY little to do with the screen itself, and more to do with the heuristic functions that the OS uses to interpolate the input. So you can look at it two ways. Either the iPhone has the best heuristics and can "figure out what you meant" the best. Or the Droid does the least manipulation and doesn't try to figure out what you meant and just takes the raw (or almost raw) data for face value."
Would you then like to explain how the other devices using the android OS performed better than the Droid?
So the droids screen is crap compared to the competition. All that matters is whether you are happy with it or not.
@ircmaxell
Assuming Android uses the same methods on all the device to 'interpolate' your input (I don't know enough about the inner workings of how this would work to say for certain) then surely it can be said that the only differentiating factors are the screen resolution and the quality of the touch sensors (obviously the same can't be said about the iPhone).
However, does it even matter if it the OS, the screen or the sensors that are at fault? What matters is the end experience and whilst you may be right in saying individual touches (i.e. clicking buttons) would not be affected, anything more than that that involves handwriting or drawing of any kind obviously would. Deciding whether those features are important would be up to the individual.
I personally would be willing to sacrifice some accuracy in the touchscreen in return for a (arguably) better device like the nexus one (I would like to see how the Desire does though). Having said that, the Droid did perform really poorly so if I were to consider it I would have to try it out myself.
@GmanC "Would you then like to explain how the other devices using the android OS performed better than the Droid?"
Sure. First, the human test was performed inconsistently. I pulled up simplydraw on my Droid, and was able to make almost perfectly straight lines with no effort. Nothing like what they produced on the Droid (I did have a slight waviness to it, but no erratic jumps). Second, each screen gets its own drivers. Drivers interpolate the electrical signals and convert it to information the OS can use. So drivers could be an issue. But the important one is that the robot tests performance of a point contact. None of these phones were designed for input from a point contact (they are designed for fingers after all). The fact that the Droid's screen is worse at registering point contacts says nothing about it's ability to register fingers (which is why the conclusion that was drawn from these tests is invalid).
This comparison is akin to saying that Verizon service is crappy because it doesn't work in *Insert country here*. You're comparing an edge case that it was never designed for, and using that to draw a conclusion about the product itself. It's not valid reasoning. That's all I'm saying...
@tomer No surprise here.
@GmanC I'll gladly explain why the other devices running the Android OS show different results, assuming that we aren't ruling out that this is mostly (or purely) a software issue.
The Motorola Droid, Droid Eris, and Nexus One run Android, but have completely different touch screen drivers and kernel branches. Different drivers equals different results. It really has nothing to do with the Android specific parts of Android (if that makes any sense) - the Android framework just takes the input from the driver and feeds it to applications.
Of course, the Droid does have different touch screen hardware, and you really can't rule that out as a factor as well.
@tomer The handsets have different panels to begin with, so there's part of the problem. Different panels get different drivers. When performing this "test" on my Nexus One, I got no waviness at all. In the first article, you'll see that they used a very "light" touch to generate the extreme waviness. I don't apply much pressure to my phone at all, and it works great. So I don't see this "issue" as being a problem for me. I get 800x480 pixels on a gorgeous AMOLED screen, and when I need to draw a straight line, it registers it for me. Period.
I would also like to see if all panels that use Gorilla Glass have similar issues; that would be interesting.
@coolbho3000 & ircmaxell
Thanks for the explanation. It's certainly plausible.
Well if that's the reason then Motorola need to get their finger out* and improve their software.
Still doesn't explain the poor performance on the light press test though.
@Federaly That tah means hi rez screens are bad. Gotcha.
@ircmaxell I disagree with you. It has EVERYTHING to do with the touch screen. You say that it doesn't matter unless you're drawing straight lines and what really matters is clicking on a link. The test is a representation of the touchscreen's (and associated drivers) accuracy; if your screen isn't very accurate, when you tap that link, it might not register because it thought your finger was slightly off from where it really was, hence the phone not thinking you tapped the link.
You sound like a little kid that just got his feelings hurt and most likely a Droid owner too. If you think you can do better and come up with a better testing methodology, than please do so.
@tomer Not surprising at all. Using my friend's Droid is a painful experience compared to my iPhone, precisely for the reason of all the vagueness of the key input (so many mistaken entries; frustration). This result doesn't surprise me at all. NOW, it makes sense why I find the Droid such a trial to use: because the screen sucks!! Curiously, it also explains why I like using the Droid Eris so much more than the Droid (he has both). I knew it wasn't just "me".
@interconnect
Honestly while this does show the iPhone is better, it isn't really indicative of performance as a whole.
If this was the complete answer, do you really think the Droid would have A) sold as many units, and B) not have had an outrageously high return rate?
If you're buying a phone to play tic-tac-toe, then most definitely, you want the iPhone, but honestly, can you (using your fingers, and only your fingers) draw a perfectly straight line?
@interconnect It's not about coming up with a better testing methodology. It's about figuring out what the testing methodology can tell you. Figuring out its limitations and assumptions. Considering that the test includes the underlying OS (drivers, kernel modules and such), it isn't so much a test of the screen, but a test of the screen and the software. And since smart software behaves differently in different "modes" (single touch vs touch and move vs multiple touches vs multiple touch and move), comparing drawing a line to clicking is not a valid argument. Now sure, it's possible that what you say is true (that one is inherently inaccurate, and the other is much more accurate), but the test provided here didn't test that. It tested drawing lines. Period. If he had it press the screen and create a matrix of dots, then you'd have an argument. But he didn't. He drew lines. It's a common theme lately to misinterpret results, and to me that's what's happening here. All that you can conclusively say is that the Droid performed the worst in this particular test. It's flawed logic to draw any kind of generic conclusion from this test alone...
And yes, I do own a Droid. And I have NEVER experienced any issues with the touch screen. It's always been 100% accurate for me. To the point that even clicking on pagination links (1 2 3 4 5) is simple at the widest zoom. That's why I am so adamant to protest the conclusions that people are drawing from this test. Because I know from my own experience that those conclusions are wrong. And combine that with the fact that they are not logically derivable from the research that was done, and you understand my objections.
@tomer yet the Droid apologists will explain away the results:
"iphone has less resolution" - well, that would only make the screen less accurate no? and yet we see the results.
"test is meaningless" - i beg to differ... if the screen won't register a tap where i actually tapped, then i'm going to have to tap again, which is inefficient.
"it's not the Droid itself, it's the drivers..." - a distinction without a difference. bottom line is the experience is inferior.
Over all these things are not a big deal to anyone but the apologists twisting the results and explaining them away are so damn annoying.
@Student Driver: It's not light pressure that makes the test harder to pass, it's the size of the contact point. Compared to that dry, metal tip used in the test, a nice, moist, meaty human finger finger is easily detectable, no matter how lightly you press. Your anecdotes don't change the test results.
@everyone talking about resolution: The display pixels are not the touch sensors! The capacitive sensor grid is completely independent of the display. You could use a display of any resolution in any of these devices, and the test results would be identical.
@vqro Exactly! But you know, robots are Apple fanboys too right?
I'd really like to see a single point touch comparison to see where contact actually was compared to where it registers on screen. The movement pattern could be the same, but creating dots instead of smooth lines.
@sliderem
Stop making excuses for the Droid's horrible tracking. It has nothing to do with the screen resolution. It has everything to do with how well the software tracks your finger on the screen. If the tracking on the Droid were worth a shit, a robot would be able to draw straight lines on it. The end.
@tomer
It depends.
Draw a thicker line and the iPhone will look like the droid and the droid will look like the iPhone.
@ircmaxell Sure doesn't matter software of hardware, Droid perfroms terrible you can't deny that you can't.
I would like to point out that the guy in the video states that the robot pointer is more difficult to detect than a human finger. That means this test gives little insight into actual usability, unless you are a cyborg. It's entirely possible the drivers/OS for the Droid are optimized towards real-world situations (like actual fingers), while the iPhone is optimized for anything you throw at it.
Please note also that I do not own a Droid.
@kyphem "Sure doesn't matter software of hardware, Droid perfroms terrible you can't deny that you can't."
I don't deny that the Droid performs terribly in this particular test. What I do deny is that the results of this particular test are applicable anywhere but within the confines of this test (Basically it's useless to the average person who reads it, since they won't draw lines on the screen with a robot finger)... That's what I'm so adamant about. Not refuting the results of the test (Although I do think they are suspect, and would like to be able to perform them myself), but refuting the conclusions that were drawn from them...