Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
The reason it can't scale is because if it isn't anything but a multiple of the original, you will get "half pixels" with controls that have uneven dimensions.

Videos and images work fine because they don't have controls being drawn on them, they are just stretched to fit the screen. Unlike an app, which has buttons, lists, labels, etc., which have specific dimensions within that app.

I don't believe Apple will adjust the resolution unless they do what they did with the iPhone 4, as it would cause fragmentation -- needing to support multiple resolutions. If the iPad 2 is to incorporate a "retina display", it can only be exactly 2x what it is now, which is 2048x1536 (264 PPI).

Well, you can get around this to a point if you base your system of vector as opposed to bitmapped graphics, then you are free to pick any screen res you like.

Of course, you do need a reasonably high screen res to make a vector graphics based system look nice when things get on the small size on screen.

I know some of Windows and Mac icons now are vector based rather than bitmapped. I'm not sure how much though.

And I don't know if any of the graphics systems in the iPad are vector based.

Perhaps there is a dev who would be kind enough to shed some light on this point?
 
Show me this images that looked a lot worse.

The math is simple. If you had a 1px x px image on the previous iPhones they would be 2px x 2px on the iPhone 4. The image size would be the same but it would use 4x as many pixels to render it, but those pixels are still taking up the same physical area, hence no change to the user’s perception.

It would appear that way yes, but that is not how I and many more people experienced older apps on our iPhone 4 displays
 
The reason it can't scale is because if it isn't anything but a multiple of the original, you will get "half pixels" with controls that have uneven dimensions.

Videos and images work fine because they don't have controls being drawn on them, they are just stretched to fit the screen. Unlike an app, which has buttons, lists, labels, etc., which have specific dimensions within that app.

I don't believe Apple will adjust the resolution unless they do what they did with the iPhone 4, as it would cause fragmentation -- needing to support multiple resolutions. If the iPad 2 is to incorporate a "retina display", it can only be exactly 2x what it is now, which is 2048x1536 (264 PPI).

Hearing you guys talk, it's a wonder PC gaming ever took off or that we can use Macs at all.

Screen resolutions have never been an absolute. Developers since the early days of bitmap displays have had to contend with the fact that the screen on which their app was displayed was not garanteed to have a specific resolution.

Visual Basic programming, of all things simple and easy to pick up by newbies, had facilities and very detailed examples in its documentation on how to properly make apps resize and keep their controls properly sized/positionned, no matter the scaling factor (be it 1.135789734973 or 1.745874587 or 0.75).

Yet all of you are in, in 2010, telling us how hard this going to be on iOS ? Let's face it, it is as hard as many of you layman claim, then it's because Apple failed at designing iOS, not because the concept is inherently hard at all.


It would appear that way yes, but that is not how I and many more people experienced older apps on our iPhone 4 displays

Perception. The older apps that assumed 480x320 looked more pixelated (not blurry) because you were comparing them side by side to 960x640 apps. Put up a 480x320 image on an iPhone 4, put it side by side with a 3GS showing the same image and both will look exactly the same.

Unless... Apple screwed up and is applying aliasing, making the "Edges" of the pixel blurrier on the iPhone 4. Which would be a big face palm moment.
 
You do realize that A-GPS is a regular GPS chip? It only utilizes a cell network to find the initial position faster.

It's also how all cell phones get their location.

In other words, Assisted GPS is actually BETTER than "GPS."

To all you demanding that apple put dual core CPUs in the next iPad, this is an example of why Steve doesn't care what you think. The customer is seldom right, especially when the customer is demanding a particular technical solution instead of demanding a particular product capability.
 
The reason it can't scale is because if it isn't anything but a multiple of the original, you will get "half pixels" with controls that have uneven dimensions.

Videos and images work fine because they don't have controls being drawn on them, they are just stretched to fit the screen. Unlike an app, which has buttons, lists, labels, etc., which have specific dimensions within that app.

I don't think that "half pixels" are a problem - try it yourself: take a screenshot of some text at 1024x768 and size it up to, say, 1280x960 in Photoshop - even with a simple bilinear filter (it's safe to assume that when running an app designed for the then-old iPad, there will be more than enough power left for a simple bilinear filter) you get a perfectly usable resized version; sure, it's not as pretty or crisp as the original, but that would just as well hold for an integer resize.

Especially coming from a 1024x768 resolution, controls are going to be more than just a few pixels wide, so it hardly matters if you're sizing something from 30 pixels to 60 pixels (at double size) or just to 45 pixels.
 
If the only features mentioned are the display and cameras, I wonder if that means we will not see a dual core iPad. Seems silly to me not to offer a dual core version.

It is only silly if you expect the iPad to be a full computer same as a notebook. But it is not and won't be for several more generations if ever

Right now the only things folks should be expecting are things to make what the iPad currently does work better. A camera so your chatting can have video, more ram so your apps and games run smoother and with less crashes, more storage for media, more connectivity options, better battery life, mildly coated glass that reduces glare with destroying colors completely and perhaps a screen that can handle blu-ray level video to go with some as yet new format for video. But even these things won't come all at once due to component costs, licensing issues etc. Even my list is at least two more generations. Right now the camera and storage is perhaps all we can get. Although adding a CDMA connection would be nice. Especially if it is in the same unit. Then I can ATT here where I gave a great signal and Verizon off in the places where it is stronger.


At that resolution, the GPU's framebuffer would be 150MB. (2560x1920x32 bits per byte) / 2048. You'd have to bump up the shared RAM at least to 1GB and have a bus line of 70Gbit/s to achieve reasonal framerate.

Actually, PowerVR SGX 543MP4 can achieve 4Gpixels/s in fillrate so it could be possible.

So if I understand your math, just having enough to handle 1080 is definitely in the cards. Once the sound and file size is covered.
 
Last edited:
As a developer who has been working on game development for iOs for both the iPad and iPhone 4 platform, I thought I'd chip in with my two cents - sorry if some of this probably makes sense only to people on the developer side, but to me it seems that the question of the increased resolution is very clear in the sense that it's not going to happen right now, for several reasons:

...

In fact, I'd be willing to bet a good amount of money on the fact that the iPad 2 will NOT feature anything that can even remotely be called retina (say, at least 1600x1200) - anyone here willing to take the other side of that bet? :)

Thank you for that very insightful post. I'd suggest that 1600x1200 isn't the lowest retina resolution for the iPad, though. A screen resolution of 1440x1080 would meet the demand for indistinguishable pixels at a viewing distance of about 14", which Apple could easily argue is a nominal viewing distance for the iPad. That is, clearly, still a significant increase in the total pixel count, only saving about 19% from the 1699x1200. So, I will not, however, take you up on that bet.

Personally, I have never seen much to complain about with my iPad's display as it is. Do I think it would be nice to have a higher res display? Sure, but I hardly think that it'll be a deal breaker for me getting an iPad 2.

The reason it can't scale is because if it isn't anything but a multiple of the original, you will get "half pixels" with controls that have uneven dimensions.

Videos and images work fine because they don't have controls being drawn on them, they are just stretched to fit the screen. Unlike an app, which has buttons, lists, labels, etc., which have specific dimensions within that app.

...

This argument makes n sense to me, whatsoever. If you're arguing that the iPad would have trouble scaling the image of the controls, while it wouldn't have trouble with full frame video, that seems absurd to me. If your arguing that it would have trouble interpreting where on the touch screen the edge of the control is because it's scaled, you are aware that a single pixel right now is about .007" and the required size of a touch for iOS to respond is significantly larger, right? So, unless you're meaning something else entirely, I just don't see your argument at all.
 
You rally want to use the iPad to shoot HD video??? I think the form factor would be horrible. You probably couldn't hold it steady enough, wired form to hold up and shoot videos, probably have to hold it with two hands to shoot decent video, .... just not practicable. Just imagine how funny it looks people holding up the ipads with two hands shooting video.

I was thinking the same thing.

Now what I could go with is making an iPad version of the whole iMovie app so that you could use your phone or touch as your camera, connect it to your iPad (perhaps with a built in USB) and edit it together before tossing it up to MobileMe, vimeo, YouTube etc.


In fact I would love that. I tried to use the phone version and found it a bit of a pain. Too small for my tastes

(and one official mention in a year of 10.7 with focus on the iOS based Mac App Store doesn't make me confident that Apple has their eye on desktop systems as much as iOS systems)

That is hardly a fair summary of what Apple has officially mentioned for either OS. That is more a summary of macrumors

And Apple can hardly be held at fault for a websites uneven coverage. MR etc will go with makes them the most money and right now that is likely iOS related articles. Only the real geeks care to debate and discuss MacOS and computer hardware. And they are outnumbered 20 to 1 by the hits on the other side.
 
Last edited:
Rumor has it that it will be an ultra-retina display, which means 7680x4320 (33mp) pixels.
The screen will remain a 9.7", which results in a 900ppi.


:D
 
That is hardly a fair summary of what Apple has officially mentioned for either OS. That is more a summary of macrumors

And Apple can hardly be held at fault for a websites uneven coverage. MR etc will go with makes them the most money and right now that is likely iOS related articles. Only the real geeks care to debate and discuss MacOS and computer hardware. And they are outnumbered 20 to 1 by the hits on the other side.

I think that it's fair to say that people, in general, are more excited right now about the iOS and the iDevices (as well as the competition, i.e. Android) than about MacOS and Macs (or the competition, i.e. Windows and Linux). The fact is MacOS is a very mature operating system. Sure, there are still great ways it could be improved, but most of those aren't at the forefront and don't effect most people directly. Thus, they're a bit harder to get really excited about. On the other hand, iOS is still a nascent operating system with plenty of room to grow in ways that most people can directly feel with each update.

And the hardware front is similar. The difference between the generations of hardware are much more pronounced in the iDevices than in the Macs. Thus, it's much easier for most people to get really excited about the next iDevice than it it for them to get really excited about the next Mac.

None of this means that Apple isn't working with all due diligence on the next Macs. It just means that that isn't where the attention of the public at large, and therefore the media, is going to be.
 
To all you demanding that apple put dual core CPUs in the next iPad, this is an example of why Steve doesn't care what you think. The customer is seldom right, especially when the customer is demanding a particular technical solution instead of demanding a particular product capability.

Thank you.

Case in point:

.. but the iPad can't?

That is just marketing, so in a year they can offer the iPad 3 "with retina display". Still the iPod and iPhone hast it even they are smaller devices. Pure BS.
 
Honeycomb looks awesome but its probably gonna be released around the same time as iOS 5. I'm sure as much Android users are looking towards Honeycomb, iOS 5 is going to be a much more significant leap in features for iPhone users. The new features that Google has shown off so far GMaps tricks and Widget tools are improvements of Android features already built in, while an Apple built Map app (speculated) and Widgets are probably going to be added to iOS. Two features that were unavailable before. Other Honeycomb optimizations include new UI and redesigned Google Apps like GMail and Youtube. Hopefully Apple also allows certain conditions in which apps/tweaks can run/modify code on the iPhone and also live backgrounds and notification pull down bar.

So we have iOS5 in March?

Sorry to burst your bubble, but I guess we may have a iOS5 preview at WWDC. Yes, exactly, WWDC, because I lost faith and don't think Apple has to show significant stuff for Lion. Therefore WWDC will be iOS-centric again.

Expect an iOS5 preview at best - maybe some touting of how great multicore-awareness in iOS5 will be.

Honeycomb is slated for March and from what I've read that doesn't sound far-fetched.
 
So we have iOS5 in March?

Sorry to burst your bubble, but I guess we may have a iOS5 preview at WWDC. Yes, exactly, WWDC, because I lost faith and don't think Apple has to show significant stuff for Lion. Therefore WWDC will be iOS-centric again.

Expect an iOS5 preview at best - maybe some touting of how great multicore-awareness in iOS5 will be.

Honeycomb is slated for March and from what I've read that doesn't sound far-fetched.

I was expecting some type of delay but i just read an article on engadget saying Honeycomb was going to be tablet only initially. OK so google obviously wants to get in on the tablet market this year ASAP so i would say support for phones would probably come two months later at the earliest. That does NOT include wait time for phone manufacturers to install their skins on the OS. So while Honeycomb will be certainly sweet for new tablet owners, everybody waiting for cool phones like the Motorola Atrix to come out with Honeycomb will have to wait longer. I myself am interested in getting a Motorola XOOM possibly versus iPad 2 but i'll wait until iPad 2 gets here.
 
I don't know, but they could well enough host an event and talk about the iPad 2, which would be available for retail later in February. Since Verizon has an event unveiling the Verizon iPhone next week, they won't be stealing any thunder by Apple hosting an event regarding the new iPads a couple days before February 3rd (for argument sake)

Apple has done media events right after the other a few times. It creates a roller coaster effect in the hype.

There's nothing exciting about an iPhone launch on Verizon except that it puts the device in more people's hands. It's not a new product. It's an old product without a sim card and a CDMA chip instead that can't access the net while in a call because of CDMA.

I also agree with another poster. The first iPad saw a late January announcement because developers needed those few months to develop new apps. Apple generally announces refreshes a few weeks before they ship, not months unless it's totally something new. I vote late Feb/early march before we see a glimpse of iPad 2 from Apple.
 
Surely, an incremental increase would only mean leaving a border around older apps, developers would release updated versions quickly enough after release anyway. I really don't see them leaving the resolution as it is, the competition is catching up, they'll want to move ahead again.
 
As a developer who has been working on game development for iOs for both the iPad and iPhone 4 platform, I thought I'd chip in with my two cents - sorry if some of this probably makes sense only to people on the developer side, but to me it seems that the question of the increased resolution is very clear in the sense that it's not going to happen right now, for several reasons:
First, thanks for your well-reasoned post. I'd like to present a few counter-arguments, but please note I'm not arguing that the next iPad will have a higher resolution, just that it's not as impossible as many seem to believe.

1. Display power consumption:
The power draw of an LCD panel is dominated by the backlight. That's why you typically see a huge difference in battery life between highest and lowest brightness settings, and why e.g. PixelQi can claim large power savings for their technology when used in reflective mode. Also, LCD technology still has significant potential for improving power consumption, it's far from maxed out. Progress in this area means that next year's screens will consume less power at the same brightness level, or consume the same amount of power at a significantly increased resolution.
The iPhone 4 is a good example of this: 4x as many pixels than its predecessor, yet better battery life (using a 16% larger battery, admittedly).

2. GPU performance:
It seems reasonable to expect a GPU upgrade from Apple this year. SGX535 has been in two generations of iOS devices, just like the MBX Lite GPU in two earlier generations. In addition, between the last two generations Apple have increased the resolution 4-5x. At the same time competition in this area is growing.
It probably makes sense to describe the performance jump from MBX Lite to SGX535 as something like 5-10x, depending on the metric used. SGX535 has been available in silicon since 2008 (as Intel GMA500). If we assume that Apple keeps using PowerVR GPUs they have a wide range of options for an upgrade. Realistically they could pick anything between 3x and 10x the current performance, though in the interest of battery life the lower end of that range seems more plausible.

Developers certainly won't ignore the existing market of first gen iPad users for a long while. If iPad 2 comes with both a faster GPU and higher resolution, they'd have the quick and easy option of letting iPad 2 owners benefit from a higher resolution while keeping geometry and shader workload the same. Of course at some point developers would want to start offering a path that uses more polygons and better shaders, but see point 4 below.

3. Fragmentation:
I think Apple handled the resolution transition somewhat poorly. They really should have let app developers provide bitmap graphics for the highest dpi target only (say, 326 dpi for iPhone 4) and perform high quality downscaling from there for lower resolution devices. That should work well for the vast majority of bitmap content. Then devs would just provide specific assets for the few cases where downscaling looks poor (e.g. pixel art).

4. Most importantly, though, there's no requirement that games render at the native resolution. Apps that show a lot of text or diagrams benefit by far the most from a higher pixel density. Games on the other hand, especially fast-paced 3D games which are most demanding on the hardware, still look ok when scaled up from a lower resolution.
Apple has made it extremely easy for developers to output graphics at essentially arbitrary resolutions and let the hardware scale to the native resolution. Scaling is commonly used on consoles, and it's not unusual on iOS devices, either: A lot of iPhone games still don't have "Retina graphics", and there are a number of iPad games which don't render at the native resolution for performance reasons (Infinity Blade probably being the most famous example).
 
So if I understand your math, just having enough to handle 1080 is definitely in the cards. Once the sound and file size is covered.

His math is wrong, don't put too much effort in understanding it. He's basically measuring 32 bytes of data per pixel instead of 4 (32 bits only require 4 bytes of storage).

Width * Height * (Bits per pixel / 8 (8 bits in a byte) = Bytes of memory needed per buffer / (1024*1024) = Megabytes of memory needed per buffer * 2 = Megabytes of memory needed for double buffered display.

2560 * 1920 * (32 / 8) = 19660800 / (1024 * 1024) = 18.75 * 2 = 37.5 MB of memory required for such a res at 32 bit color depth with a double buffer.

I don't know where he got the idea that 150 MB was required.
 
I suggest anyone who replied below reads all of these as well as the final note at the bottom, I'm not going to write the same thing to all of you.
Well, you can get around this to a point if you base your system of vector as opposed to bitmapped graphics, then you are free to pick any screen res you like.

Of course, you do need a reasonably high screen res to make a vector graphics based system look nice when things get on the small size on screen.

I know some of Windows and Mac icons now are vector based rather than bitmapped. I'm not sure how much though.

And I don't know if any of the graphics systems in the iPad are vector based.

Perhaps there is a dev who would be kind enough to shed some light on this point?
No, you can't. The graphics don't matter, I'm talking about the sizing of controls. Read below.
Hearing you guys talk, it's a wonder PC gaming ever took off or that we can use Macs at all.

Screen resolutions have never been an absolute. Developers since the early days of bitmap displays have had to contend with the fact that the screen on which their app was displayed was not garanteed to have a specific resolution.

Visual Basic programming, of all things simple and easy to pick up by newbies, had facilities and very detailed examples in its documentation on how to properly make apps resize and keep their controls properly sized/positionned, no matter the scaling factor (be it 1.135789734973 or 1.745874587 or 0.75).

Yet all of you are in, in 2010, telling us how hard this going to be on iOS ? Let's face it, it is as hard as many of you layman claim, then it's because Apple failed at designing iOS, not because the concept is inherently hard at all.
For starters, there's no need to be rude. You're missing one important point in your logic, when an iPhone app upscales it isn't keeping the controls the same size and just repositioning them like you said with a VB app, it's increasing the size of each control as well and keeping it in the same position it was originally.

When you have a button which is 101x31 pixels, and you scale it up by 50% you get 151.5 pixels and 46.5 pixels, a control can't be drawn at that, it has to rounded up or down and either way it'll not look right -- if it was a multiple, 2, 3, 4, etc., this would NEVER happen. And, games (like images and videos) aren't comparable to apps either.

There's a reason when you're designing an app for Mac, Windows, iPad, iPhone, that you can't set a control to have half a pixel in its size, I.E, 101.5 pixels wide, so why won't they let you do that, but they'll let that happen when they scale it? They won't.

As I said before, scaling can't happen unless it's 2x, 3x, etc. (not talking about images, or videos)

And, I'm certainly not a layman, whilst I may not be an expert I still know plenty.

I don't think that "half pixels" are a problem - try it yourself: take a screenshot of some text at 1024x768 and size it up to, say, 1280x960 in Photoshop - even with a simple bilinear filter (it's safe to assume that when running an app designed for the then-old iPad, there will be more than enough power left for a simple bilinear filter) you get a perfectly usable resized version; sure, it's not as pretty or crisp as the original, but that would just as well hold for an integer resize.

Especially coming from a 1024x768 resolution, controls are going to be more than just a few pixels wide, so it hardly matters if you're sizing something from 30 pixels to 60 pixels (at double size) or just to 45 pixels.
Like I said, images or videos don't matter -- they're different, they stretch, they don't have individual controls being drawn inside of them. Read above.
This argument makes n sense to me, whatsoever. If you're arguing that the iPad would have trouble scaling the image of the controls, while it wouldn't have trouble with full frame video, that seems absurd to me. If your arguing that it would have trouble interpreting where on the touch screen the edge of the control is because it's scaled, you are aware that a single pixel right now is about .007" and the required size of a touch for iOS to respond is significantly larger, right? So, unless you're meaning something else entirely, I just don't see your argument at all.
I'm arguing neither of what you said. Read above.


>>> Summary

If you increase the resolution of the iPad by 1.5x to 1536x1152 and you've got an iPad app which has a button which is 101x30 pixels, it'll be scaled up to 151.5x45 pixels, and as you can see that just won't work.

An app isn't the same as a video, or an image, which have no controls drawn inside of them that are a certain size and have a certain location, so when you scale it, it works fine.

Do you all get it now? It has nothing to do with images or video or even games.
 
2. GPU performance:
It seems reasonable to expect a GPU upgrade from Apple this year. SGX535 has been in two generations of iOS devices, just like the MBX Lite GPU in two earlier generations. In addition, between the last two generations Apple have increased the resolution 4-5x. At the same time competition in this area is growing.
It probably makes sense to describe the performance jump from MBX Lite to SGX535 as something like 5-10x, depending on the metric used. SGX535 has been available in silicon since 2008 (as Intel GMA500). If we assume that Apple keeps using PowerVR GPUs they have a wide range of options for an upgrade. Realistically they could pick anything between 3x and 10x the current performance, though in the interest of battery life the lower end of that range seems more plausible.

powervr SGX543MP2 will be in the next ipad and iphone
 
I don't think the iPad will get dual-core... because the iphone doesn't have dual-core. I hope I am wrong but I guess this means we won't have dual-core ipad til 4th gen... as iphone 6 in 2012 will probably get dual-core.... then the ipad the winter after.

nah.. Im probably wrong...

But I want to know when this will launch... so I can time when I sell my ipad 1. Will there be an apple media event?
 
I think the big question is how much better you want to get. You do want/need to get better... but you also don´t want to anger people that spend 500-800$ on the first Model.
Well, with a lot of Apps it will only be about speed and thats ok... but on the iPhone you already got games that will not run on the older processor, you don´t want that to happen with the first iPad right away. So a processor that it 5x better than the old on is quiet risky... or Apple could just disalow Apps that do not work on the old iPad untill the 3. Gen is released.... hmmm...
 
There's a reason when you're designing an app for Mac, Windows, iPad, iPhone, that you can't set a control to have half a pixel in its size, I.E, 101.5 pixels wide, so why won't they let you do that, but they'll let that happen when they scale it? They won't.
At least Windows Presentation Foundation lets you do that. It also allows arbitrary scaling.

The basic concept of vector graphics is that you create a geometrical representation in an abstract, device independent coordinate space. To show this representation on a raster display you need to sample, or rasterize, it. If you do it the cheap way, taking a single sample per pixel, you might get aliasing artifacts. If you want good quality, you take several samples per pixel and filter them (aka anti-aliasing).
 
...
>>> Summary

If you increase the resolution of the iPad by 1.5x to 1536x1152 and you've got an iPad app which has a button which is 101x30 pixels, it'll be scaled up to 151.5x45 pixels, and as you can see that just won't work.

An app isn't the same as a video, or an image, which have no controls drawn inside of them that are a certain size and have a certain location, so when you scale it, it works fine.

Do you all get it now? It has nothing to do with images or video or even games.

Sorry, no, I still don't get it. So, you're not talking about the pixel area of the control interface, which I'd agree can't in any way be fractional pixels. You say that you're not talking about images, but you are talking about buttons. You say that scaling up by 1.5x can't be done because "that just won't work." But, correct me if I'm wrong, aren't we talking about image graphics when we're talking about buttons? Therefore, why aren't we talking about scaling images?

And you can't say that this sort of thing is never done under MacOS. Have you ever used screen zooming? When using screen zooming, you scroll through scaling factors of 1.05x, 1.25x, 1.5x and so on. You can't tell me that buttons never show up as technically needing to be drawn with fractional pixel dimensions under those circumstances. Of course, the computer zooms the display and handles the fractional pixels through antialiasing essentially invisibly to the user. And, of course, when things are displayed that way they look a little fuzzy and not as good, just as a scaled up video doesn't look as clean and crisp as when it's displayed at native resolution.

So, I'm not saying the result would be perfect and gorgeous. Nor am I saying that I think this is what Apple will do. All I'm saying is that it is possible, which is the opposite of what you seem to be arguing.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.