Results matching “ipad retina” from Bill's Words

I might have to back off my stance that Apple will bring out an iPad “retina display” that won’t be twice the resolution of the current iPad.

MacRumors reports that Samsung has developed just the LCD required for that (give or take a few pixels).

Sigh. And I thought I had excellent reasoning.

John Gruber and I have briefly conversed about what the resolution of an upgraded-resolution iPad will be. He firmly believes it will be 2x the current resolution because it’s just easier. (True, it is.) But I believe that some future revision of the iPad will have a higher-resolution display which is not a fixed integer multiple of the current resolution for reasons I’ve outlined previously. Two more things make me think I’m right.

First, Apple itself. With 10.4 and 10.5, resolution independence was introduced in phases and was “developer only.” (Via 10.6 was supposed to be more resolution-independent than the previous two releases, but… alas, not so much. That’s bad.

Or… is it? Wouldn’t it be a convenient alignment of things if iOS 5 and Mac OS X 10.7 introduced resolution independence about the same time? They share the same rendering engines, after all, and many other parts of their architecture. If both OSs introduced the last bits of making their interfaces resolution independent, it would make good use of resources in the company. Not that Apple has to think frugally, but if they’re trying to converge the two platforms somehow…

I’m just sayin’.

Second, WebOS Enyo apparently does some resolution independent jiggery-pokery according to this Engadget note. Not that I understand exactly what this means, because I am not about to plunge into the WebOS SDK documentation, but they are making a concerted effort to attract developers by allowing devs to write once for multiple target resolutions. That’s powerful stuff.

Now, Apple does not promote that devs should write once for iPad and iPhone/iPod (because the interfaces themselves should really be tailored to the different screen sizes), though it does work. Instead, Apple would likely try the same approach as HP and promote resolution independence as the bridge between future higher-resolution devices and the past’s lower-resolution devices. WebOS is ahead of both Mac OS and iOS here—there’s some real competition, finally.

So that’s it. I further stick my neck out on the subject of the iPad X’s resolution.

I sure hope I’m right…

There’s been a lot of guessing about the resolution of the iPad 2. Everybody who has been speculating (except me, it seems) has been saying that in order for Apple to be happy with the product, they’ll have to call it a “Retina Display,” and that in order to do so, they’ll have to double the iPad’s 1024x768 resolution. That is, after all, what they did to the iPhone 3 to achieve Retina Display status for the iPhone 4.

Yes, this makes good sense for lots of good reasons, not the least of which is that graphics work better when the scaling factor is a nice, round two. And with the more-limited hardware of the iPhone, it makes a lot of sense. But as others have noted, that resolution would be a whole helluva’ lot of pixels to push, and that ain’t happening in the iPad. Not yet, anyway.

I, on the other hand, am guessing that Apple will choose something else and that there’s nothing special about “times two.” They’ll choose manufacturable, cost-effective, and, above all, marketable over “times two” any day. Technically, I believe “times x” is completely feasible as I outline in a previous entry, especially if the iPad 2 has more horsepower than an iPhone 4.

So I’ll guess that Apple will define “Retina Display” as it properly should be, i.e., they’ll base it on the resolving resolution of the human eye, and not on an arbitrary number of pixels “times two.”

The thing that most people miss in their guessing is that the average user doesn’t hold an iPad at one foot as she might with an iPhone 4. No, the typical iPhone user holds it at about 1.5’ to 2’—lap distance or so. If we start with Bryan Jones’ assertion that the eye can resolve about 287ppi at one foot and scale that triangle up to 1.5’, then the resolution required is only about 190ppi. Since the current resolution of the iPad as stated by Apple is 132ppi, then an iPad Retina Display would have a minimum of 1475x1106 pixels, which is pretty close to the SXGA+ standard of 1400x1050.

If Apple uses a 2’ figure for average use distance, then 142ppi will do, or 1101x826. But that’s too close to make any kind of marketable difference, and it’s at the hairy edge of acceptable Retina Display resolution. If anything, they’ll shoot for over, not under.

Anyway, at 1400x1050 pixels, that’s not even twice as many pixels to push, and is a much more reasonable expectation of the iPad 2’s hardware.

Bottom line? Here’s my bet: if—and that’s a big if—if Apple changes the iPad resolution for the iPad 2, I’m going to bet on somewhere between 175ppi and 200ppi, favoring the upper end of that range.

Speaking of bets, John Gruber hasn’t taken me up on my bet. I don’t expect him to—heck, he’s a celebrity! But I think my wager requires clarification: I’m not betting whether or not Apple will change the iPad resolution with the iPad 2, but rather that if they do, it’ll be to something other than (and less than) “x2.” So that’s the wager: if they do, it’s less than two.

In this article, Gruber makes some good arguments against a double-resolution iPad 2 display, chief among them that it’s one helluva’ lot of pixels to push. I have to admit, I’m with him on that.

But I side with Engadget which thinks that the display resolution will increase, i.e., it won’t remain at 1024x768, but won’t double, either. Gruber thinks that’s unlikely, too, because that would require graphics to scale up from original iPad resolution to some new iPad resolution, a non-integer (i.e., “not easy”) multiple. And that would require developers to implement three different graphic designs for their programs in order to make it look pixel perfect.

Here’s where he and I disagree.

First, I think the iPad display looks great, but not that great. I’m comparing it to an iPhone 4 which looks absolutely frickin’ fantastic. Today’s iPad looks positively dowdy in comparison, and I think Apple knows it. A resolution bump is a good fix.

Second, I think that this change might result in a push for developers to use Core Graphics and Quartz 2D to create resolution-independent graphics. (App icons still remain a problem.) In general, I think Apple is gently pushing developers in this direction, so it kinda’ fits.

Third, where bitmap graphics are required (icons, for example), my guess is that we’ll see developers encouraged to create 2x graphics and let the hardware do scaling down, not up. That’s considerably easier to do and have decent results.

Finally, all of Quartz 2D and Core Graphics is implemented as floating point. There’s nothing special about the number “2” except that it is easier for graphic designers to deal with in the land of integral pixels. Each UIView, the iPhone container for any kind of on-screen graphics, goes through a massive amount of floating-point computation (the “transform,” if I understand things correctly) before bits are moved to the screen buffer. Because we’re not dealing with integer math, there really is nothing inherent in the system which makes dealing with one screen size and/or resolution any easier than any other.

So while I am completely on board with the “no double resolution,” I’m willing to bet John a Dogfish Head World Wide Stout (which costs considerably more than my usual nickel bet) that we’ll see a non-integer bump in resolution with the iPad 2. It’s feasible and—I believe—likely.