Seems that the important question has not been answered. You may want the iPhone to have the same quality as print, but why? Is iPhone competing with magazines now? Do Apple customers want that, and would they be happy to pay for it?
Probably... the brand and marketing is so strong Apple could probably reduce the resolution of the screen and charge more for it. The fact that its the new iPhone would be enough to make up for it.
if they're going to push the iBook platform on the phones, they'll want as high quality of text as possible to get people who have the phone but not the iPad to buy books still, giving them their %, instead of buying through Kindle (though obviously you can also use the Kindle app... hm)
Agree. I wouldn't have entertained switching from my previous or current laptops (1400x1050 and 1680x1050 pixels respectively) to a 15" MacBook Pro with a lower resolution screen. The recent refresh suddenly makes them more tempting though.
Probably because it's far easier to scale up production on a smaller, 4th generation screen than it would be to scale up production on a larger, less established device.
I would happily pay more for an iPad HD (or whatever). I was (and still am) excited about it as an e-reader++, but print on the iPad looks merely adequate (ie, just like a laptop), not stunning, as I would have expected.
The larger sized your display and the denser your pixels, you're correspondingly fighting a sq cm pixel defect rate and drawing lower yields (^2) for each increase, with the obvious effect on price. This is why, at trade shows, you see huge screens with low resolution and tiny screens of stunning resolution.
Further, every time a new technology comes along (e.g. AMOLED), you throw out many of the manufacturing techniques you've developed to increase pixel yields.
The graphics hardware can't push enough pixels to render 60 FPS on a high dpi display of that size. Also, more RAM would be required. More RAM and faster graphics => more expensive with shorter battery life.
When I realized we're only a resolution doubling or two away from being beyond the limits of human visual perception, I realized why the much-discussed Resolution Independence feature keeps not making it into each release of Mac OS X.
I think it's funny how no one has noticed that the iPhone/iPad SDK essentially has resolution independence built-in. All the Interface Builder default controls are vectorized and can scale beautifully.
Except that if you want to use an image (say, on a button), you're still stuck with rasterized types. The only way to have resolution independence for buttons would be to use custom Quartz drawing.
There's no SVG or no PDF support in UIImage; the only part of the SDK that supports SVG is UIWebView. And Quartz PDF support is more suited for displaying pages of PDF's, not small images in a button.
There's no SVG or no PDF support in UIImage; the only
part of the SDK that supports SVG is UIWebView. And
Quartz PDF support is more suited for displaying pages
of PDF's, not small images in a button.
This.
I'm currently working on an iP* app, and I had to roll my own UIView class to handle PDF pages to provide vector imaging without resorting to a heavyweight web view.
This discussion is about a 162 dpi display, so quadrupling that to 648 or quadrupling on two dimensions gets us to 2592 dpi. That's the right ballpark [1], but only on iPhone-sized devices. The iPad is 132 dpi [2], about 10% better than a typical CRT from a decade ago [3]. MacBook Pros recently got the first resolution increase (36%, to about 115 dpi) in about 5 years. At that rate it will be more than 20 years before we reach the limits of human visual perception on large displays.
tl;dr Because 960x640 will give a picture quality (in particular for texts) comparable to high-quality prints.
But that's not much of an answer. Why higher resolution? Because it will look better. Why 960x640 precisely? Because it's exactly 4 times the current resolution. But beyond that…
I guess I'm missing a more insightful conclusion like: "since Apple is putting iBooks on the iPhone, they want to make it a decent e-reader".
Because competing devices have higher density displays. And high density allows for prettier, crisper, more natural graphics. iBooks are just the icing on the cake.
Why 960x640 precisely?
Because it makes things simple. Whether you have a bitmap or some procedural drawing (vector if you will) you can easily translate it into coordinates on the new screen - just multiply by 2. No need to worry about sub-pixel drawing and artifacts occurring as result of rounding errors.
These questions were actually rhetorical. (I actually answered them, though in less detail)
The thing is that I was kind of surprised by the length of the post compared to the relative lack of insight. Usually John Gruber brings interesting points and reflection to his posts, going in depth on a lot of his arguments; but with "Why 960x640" as a title, I was hoping to read something I didn't know or didn't think about. Here he mostly stated facts without, in my opinion, really answering the question "why does an iPhone need a higher resolution/pixel density?".
Well, there's this: "The next-gen iPhone is shooting for that caliber of resolution — not merely to exceed the resolution of competing devices, but to rival the optical quality of print." He doesn't belabour it, because it's obvious: high quality text is really pleasant to look at, and that will sell more iPhones.
I recently ported my company's iPhone app to Android (they both have the same look, rather than using each platform's look) and the difference between the two images side by side when testing on the Nexus One was striking. Quite simply, Apple could not have gone into the 2010 holiday season against Android phones with spectacular, dense 4" displays.
I really do think Gruber's going to have to eat his own chowder on this one.
There's just no precedent for Apple doing this. They've never played the resolution game. Remember 15" laptops a few years ago? While Sony and other higher-end makers started cramming more pixels in each year, Apple slowly went from 1280x854 to 1440x900 and stopped there. Only in 2010 have they started to offer (and only as a BTO option!) the 1680x1050 res that you've been able to get on Vaios for years. Apple just prefers to concentrate on their antialiasing algorithms than to pump pixels.
I think a much more likely resolution bump would be 1.5x each dimension for 720x480. It'll be a little short of HTC and Motorola's 852x480, but if Apple doesn't cut the color resolution with a Bayer pattern, their display will still look better. And most importantly, it'll use less memory and run faster than a full 960x480.
Your reasoning is solid, but Gruber has repeated the 960x480 resolution a few times in reporting on the future of the iPhone, and typically when he gets that specific about something it's because he knows.
Your reasoning, however, isn't very solid. "It's right because he wouldn't say it if it wasn't right"; it's circular and possibly an appeal to authority, and doesn't respond to the issue addressed.
The parent post is not questioning whether Gruber believes his sources, but arguing that the information seems implausible. Your reply is that Gruber believes his sources are correct - yes he might state it because he believes it, but it could be wrong.
>Gruber is right. His specific answer comes from his correct knowledge. I know he is correct because he's specific.
Erm, that's a very odd way of putting it. I think the OP was just saying "Gruber's specific claims have been reliable in the past, and hence are quite likely to be correct now." That is perfectly reasonable and it's not circular. (It is, of course, an inductive form of appeal to authority, but there is nothing wrong with that.)
As I first read it, I thought "he's right because he's specific and he's specific because he's right", which would be circular; but while the he's all refer to Gruber, the rights are distinct.
He thinks he is right therefore he is specific, because he is specific I think he is right.
That's what I was trying to say in my awkward earlier post. That does seem appealing to authority - thinking Gruber is right because of what he says and how he says it, rather than considering why he says it.
Actually, what riled me more was the interpretation
"He thinks he is right therefore he is specific, because he is specific I think he thinks he is right."
Which conveys absolutely no information and adds nothing to the discussion, just repeating that G thinks he is right, which is a given, and was still getting upvotes; so I tried to attack the post to feel better, banking on pedanticism being good for a few karma points around here anyway.
Apple _is_ playing the resolution game, but they're not in a hurry. That's why there was no resolution bump on the 3G or the 3GS, even though better displays were available at the time.
By doubling the resolution, they avoid a whole lot of problems by allowing old apps to run at their current resolution without any visible quality loss, something which a 1.5 bump would not achieve.
From the prototype, you'll notice that the battery is much bigger. You can expect that Apple will put a faster chip (similar to the iPad) in the new phone, and that memory and CPU speed will not be an issue.
Apple likes 960 width x 540 height (see AppleTV), because it's easy to get to 1/4 1080p by dropping every other line and every other pixel, or to line double back up to 1080p from iFrame.
A 1.5x scale factor would make existing apps look terrible. They will look fine scaled 2x. If they increase the resolution (and they will; the recently leaked prototype is irrefutable proof) they must increase it 2x for this reason alone.
Exactly. Doubling each dimension of an image is a cheap operation. Multiplying by 1.5 is not, due to the antialiasing that would be required. Even with aa, the scaled image would look worse than the original.
1.5x is a nightmare on Android requiring me to get artwork done twice for every single asset. 2x would be a typical Apple brilliancy--make things easier by using quality hardware. On the other hand, I can see them telling devs to update or get left behind.
1.5x would indeed look bad for "legacy" apps, but I think even this fits in to Apple's current developer strategy: Update or get left behind.
I mean, remember, Apple was willing just recently here to risk the incredible backlash of 3.3.1 for the sake of platform direction. I think they'll gladly risk developers complaining about redrawing UI bitmaps.
I think 3.3.1 is primarily a developer issue, whereas resolution is very much a user-facing issue as well.
With the iPad, Apple didn't force developers to recode their iPhone apps, instead it allowed the older apps to run at native or double resolution. Granted, the double resolution doesn't look pretty, and Apple is strongly encouraging developers to produce iPad native apps, but they definitely seem to be taking advantage of their 185,000 app head start, and 1.5x would presumably look even worse on an iPad than 2x.
> I think even this fits in to Apple's current developer strategy: Update or get left behind
That seems like the opposite of their developer strategy to me. They've resisted making any substantial changes to the basic hardware profile of the iPod Touch / iPhone lineup ever since it was released, presumably to save developers from having to target multiple devices (and this is still a point raised as a complaint against android all the time).
I'd say the reason that Apple hasn't gone for more pixels per inch on their computers has to do with the lack of resolution independence in their operating system (though it's been on the horizon for several years now). Making everything denser makes for harder-to-read menus and controls, unless the software (including third-party software!) is smart enough to compensate for it.
With the iPhone, they're presumably addressing the software aspects of the resolution change in addition to just bumping the pixel count.
Wondering about what will be Apple's next big product ...
Internet TV?? OMG!
iPad will rule the world because of the Internet TV!!
iPad will be the perfect interface for controlling a TFT on steroids, and the target markets fits so well it makes a lot of sense. Does anybody agree to me?
I see the iPad as an enabling device, it's ease of use and hipness will create a big market on the less technology savy.
It doesn't matter what the resolution is; until they get some typographical taste, it's not going to rival print. iBooks looks like [insert rude word] and it's not because of the resolution.
I feel like they're eventually going to try to sell me a laptop with a 2048x100 pixel screen where I have to read one very long line of text at a time. What is wrong with laptop screen manufacturers? I haven't bought a notebook in years from anyone but Dell (upgraded screen each time) and Apple (generally saner than other manufacturers).
Gruber ignores the Droid, which has sold far more units than the N1, does not have the Pentile subpixel reduction, and has an 854x480 display. A 960x640 display would rock but given that the LCD is smaller than current iPhones, I think it will be a 1.5x increase which will look fine with antialiasing and font/UI element scaling. So 480x720.
Everyone who has been following Apple closely for more than 10 years knows that 300+ dpi on every device has always been Apple's goal. They haven't pushed it in OSX due to the massive legacy effects. The Mac typographic layout system was designed for a 300 dpi future.
Whatever the actual resolution is, the iP* line definitely needs the upgrade. It's amazing how "fuzzy" everything looks when I switch from my N1 back to my iPod Touch. It makes a very noticeable difference when displaying and reading web sites.
I'm sure if it wasn't for Apple's amazing design skills this would have made their devices look outclassed long ago, but because everything is so beautifully done they have gotten away with 320x480 well past its use by date.
56 comments
[ 2.4 ms ] story [ 77.1 ms ] thread... except that high-quality print doesn't use Marker Felt...
Probably... the brand and marketing is so strong Apple could probably reduce the resolution of the screen and charge more for it. The fact that its the new iPhone would be enough to make up for it.
Is the 2nd generation iPad going to have a significantly better screen?
Further, every time a new technology comes along (e.g. AMOLED), you throw out many of the manufacturing techniques you've developed to increase pixel yields.
There's no SVG or no PDF support in UIImage; the only part of the SDK that supports SVG is UIWebView. And Quartz PDF support is more suited for displaying pages of PDF's, not small images in a button.
I'm currently working on an iP* app, and I had to roll my own UIView class to handle PDF pages to provide vector imaging without resorting to a heavyweight web view.
This discussion is about a 162 dpi display, so quadrupling that to 648 or quadrupling on two dimensions gets us to 2592 dpi. That's the right ballpark [1], but only on iPhone-sized devices. The iPad is 132 dpi [2], about 10% better than a typical CRT from a decade ago [3]. MacBook Pros recently got the first resolution increase (36%, to about 115 dpi) in about 5 years. At that rate it will be more than 20 years before we reach the limits of human visual perception on large displays.
[1] http://www.edwardtufte.com/bboard/q-and-a-fetch-msg?msg_id=0... [2] http://www.apple.com/ipad/specs/ [3] http://viewsonic.com.pk/pdf/us_eng/products/pf77.pdf
But that's not much of an answer. Why higher resolution? Because it will look better. Why 960x640 precisely? Because it's exactly 4 times the current resolution. But beyond that…
I guess I'm missing a more insightful conclusion like: "since Apple is putting iBooks on the iPhone, they want to make it a decent e-reader".
Because competing devices have higher density displays. And high density allows for prettier, crisper, more natural graphics. iBooks are just the icing on the cake.
Why 960x640 precisely?
Because it makes things simple. Whether you have a bitmap or some procedural drawing (vector if you will) you can easily translate it into coordinates on the new screen - just multiply by 2. No need to worry about sub-pixel drawing and artifacts occurring as result of rounding errors.
The thing is that I was kind of surprised by the length of the post compared to the relative lack of insight. Usually John Gruber brings interesting points and reflection to his posts, going in depth on a lot of his arguments; but with "Why 960x640" as a title, I was hoping to read something I didn't know or didn't think about. Here he mostly stated facts without, in my opinion, really answering the question "why does an iPhone need a higher resolution/pixel density?".
I recently ported my company's iPhone app to Android (they both have the same look, rather than using each platform's look) and the difference between the two images side by side when testing on the Nexus One was striking. Quite simply, Apple could not have gone into the 2010 holiday season against Android phones with spectacular, dense 4" displays.
There's just no precedent for Apple doing this. They've never played the resolution game. Remember 15" laptops a few years ago? While Sony and other higher-end makers started cramming more pixels in each year, Apple slowly went from 1280x854 to 1440x900 and stopped there. Only in 2010 have they started to offer (and only as a BTO option!) the 1680x1050 res that you've been able to get on Vaios for years. Apple just prefers to concentrate on their antialiasing algorithms than to pump pixels.
I think a much more likely resolution bump would be 1.5x each dimension for 720x480. It'll be a little short of HTC and Motorola's 852x480, but if Apple doesn't cut the color resolution with a Bayer pattern, their display will still look better. And most importantly, it'll use less memory and run faster than a full 960x480.
The parent post is not questioning whether Gruber believes his sources, but arguing that the information seems implausible. Your reply is that Gruber believes his sources are correct - yes he might state it because he believes it, but it could be wrong.
It might be an appeal to authority, but there isn't any circularity. If you're going to play fallacy bingo, at least get it right...
It isn't circular, is it? I'm mixing G's belief in his correctness and this poster's belief in G's correctness.
Erm, that's a very odd way of putting it. I think the OP was just saying "Gruber's specific claims have been reliable in the past, and hence are quite likely to be correct now." That is perfectly reasonable and it's not circular. (It is, of course, an inductive form of appeal to authority, but there is nothing wrong with that.)
He thinks he is right therefore he is specific, because he is specific I think he is right.
That's what I was trying to say in my awkward earlier post. That does seem appealing to authority - thinking Gruber is right because of what he says and how he says it, rather than considering why he says it.
Actually, what riled me more was the interpretation
"He thinks he is right therefore he is specific, because he is specific I think he thinks he is right."
Which conveys absolutely no information and adds nothing to the discussion, just repeating that G thinks he is right, which is a given, and was still getting upvotes; so I tried to attack the post to feel better, banking on pedanticism being good for a few karma points around here anyway.
:/
By doubling the resolution, they avoid a whole lot of problems by allowing old apps to run at their current resolution without any visible quality loss, something which a 1.5 bump would not achieve.
From the prototype, you'll notice that the battery is much bigger. You can expect that Apple will put a faster chip (similar to the iPad) in the new phone, and that memory and CPU speed will not be an issue.
http://support.apple.com/kb/ht3905
Apple likes 960 width x 540 height (see AppleTV), because it's easy to get to 1/4 1080p by dropping every other line and every other pixel, or to line double back up to 1080p from iFrame.
They have also never made a music player. Right until the moment when they made one.
I mean, remember, Apple was willing just recently here to risk the incredible backlash of 3.3.1 for the sake of platform direction. I think they'll gladly risk developers complaining about redrawing UI bitmaps.
With the iPad, Apple didn't force developers to recode their iPhone apps, instead it allowed the older apps to run at native or double resolution. Granted, the double resolution doesn't look pretty, and Apple is strongly encouraging developers to produce iPad native apps, but they definitely seem to be taking advantage of their 185,000 app head start, and 1.5x would presumably look even worse on an iPad than 2x.
That seems like the opposite of their developer strategy to me. They've resisted making any substantial changes to the basic hardware profile of the iPod Touch / iPhone lineup ever since it was released, presumably to save developers from having to target multiple devices (and this is still a point raised as a complaint against android all the time).
With the iPhone, they're presumably addressing the software aspects of the resolution change in addition to just bumping the pixel count.
Internet TV?? OMG!
iPad will rule the world because of the Internet TV!! iPad will be the perfect interface for controlling a TFT on steroids, and the target markets fits so well it makes a lot of sense. Does anybody agree to me?
I see the iPad as an enabling device, it's ease of use and hipness will create a big market on the less technology savy.
The style is provocative but just to get some attention to the idea.
I'm sure if it wasn't for Apple's amazing design skills this would have made their devices look outclassed long ago, but because everything is so beautifully done they have gotten away with 320x480 well past its use by date.