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.
Photos of 4:3 Display from Apple's 'iPad Mini' Surface
- Thread starter MacRumors
- Start date
- Sort by reaction score
Actually, no. iOS doesn't use the @2x values for touch points. That's a big part of why they use the @2x factor for retina scaling - it doesn't require separate touch point calculations. So on a 960x640 resolution, from a touch point perspective, they still use 480x320.
Yes, that's exactly what the poster meant. He just didn't use the proper jargon (the points aren't scaled, only the mapping between points and pixels are changed from a 1:1 system to a 1:4 system). Of course, we all understood what he meant.
1.5 scaling factor breaks this badly as I've stated.
How so? As long as it continues to virtually use 768x1024 from a touch point perspective, how does it matter how the images are scaled so long as they are uniform?
Doubling from 16GB to 32GB costs Apple much less than the doubling from 32GB to 64GB. From that standpoint the price differences should be different.
But flash is only one component of the device and Apple doesn't apply a markup to the flash alone. It marks up the entire device and then rounds off each price as determined by the marketing department.
16GB iPod touch (4th generation) - C$199
32GB iPod touch (4th generation) - C$249
16GB iPhone 5 (unlocked) - C$699
32GB iPhone 5 (unlocked) - C$799
iPhone customers pay $50 more for the same flash, but the increase in price is a smaller percentage of the base cost.
Regardless of how Apple/Asus/Google/HTC/Samsung choose to markup their devices it's still frustrating when you realize how little the flash really costs them. Moving from 16GB to 32GB costs a manufacturer less than $10.
What KnightWRX is saying is that you cannot split the virtual pixel into smaller ones for the purpose of the touch point, well I guess you could but that'll make things messy. For this and the ease of scaling reasons, I agree with Knight and I don't think Apple will go with anything other than 2X for iPad. It'll either be 2048 x 1536 or 1024 x 768.
Well, you could, Quartz CGPoints use floating point representations. But for a developer, that would be a nightmare to figure out bounding box around pixel based objects, especially considering all the code is probably written in a way that touchpoints fall right on pixel boundaries, not at half pixels...
It's just not a feasible, "easy no updates required" solution.
Although touch sensitivity is increased in retina mode, so that even though the touch range in 'points' is still (0, 0) to (480, 320), you can get half point touch coordinates, so that, for example, the touch point (7.5, 9.5) would map onto the retina pixel coordinate (15, 19).
From this, it would theoretically be possible to have any pixel resolution map onto the existing touch resolution, it's just that the touch points may have fractional values.
Is that really a problem? I mean, the level of touch accuracy with fingers isn't that precise.
source
Couldn't coordinates of touch points on a new HxW 4:3 resolution screen be evaluated and translated to a 1024x768 touch point value with enough accuracy and no real loss in precision (0.5/163 inches, seriously?).
It's anyway an approximation. And that's anyway what will happen with a 1024x768 7.85" screen: fingers won't shrink, and area of touch contact neither.
Last edited:
It's a problem in that developers have not had to do fractional touch point analysis, so existing code does not use the fact that CGPoints are represented using floating point values (I already stated they are) :
Code:
struct CGPoint {
CGFloat x;
CGFloat y;
};
typedef struct CGPoint CGPoint;ok
but again:
I'm not speaking here about fractional touch point analysis by the dev, but by iOS. Couldn't it be completely transparent, as if the app was running on a 1024x768 iPad, with the underlying Multi-Touch system doing the necessary work.
Last edited:
At the time of the iPhone 4 rumors, when the display parts leaked, the source pointed a microscope at the screen and discovered the unprecedented retina.
Today we all know what retina is, so why didn't these guys examine their parts more deeply to know if it's retina or not?
Today we all know what retina is, so why didn't these guys examine their parts more deeply to know if it's retina or not?
But technology changes. I just held a 3GS in my hand last week (sold my old phone on eBay) and the display looks like crap. But that's because of the display type and all that. I'm not tech savvy enough for the technical terms, but a display made back then isn't nearly as quality as they are now, even at the same res.
Unfortunately 1024x768 makes the most sense from compatibility and price point perspective. I really hope it is higher though. I want to buy one for reading e-books and portability, but having tried both Kindle Fire and the original Samsung Tab 7" at 1024x600, that dpi just doesn't cut it, especially seeing it next to iPad 3 or any new iPhone.
I'm sure it will sell well unless it is way overpriced, but it seems that Apple and a small but growing chunk of Apple users are diverging. First with the iPhone 5 and its lack of any width increase that I wanted the most, to the mini iPad with sub-Retina resolution.
I'm sure it will sell well unless it is way overpriced, but it seems that Apple and a small but growing chunk of Apple users are diverging. First with the iPhone 5 and its lack of any width increase that I wanted the most, to the mini iPad with sub-Retina resolution.
What work ? In the end, you'll be getting a X.Y value, floating point, since the proposed scaling factor would be 1024x768 points over 1536x1152 pixels. Otherwise, if iOS does the work for you and converts the 1024x768 to a 1536x1152 point system... then well... you're not using a 1024x768 point system anymore, hence a lot of other code will break (code that assumes a touch can't be outside of 1024 or 768 depending on the orientation.
That's the whole point. 1.5 scaling factor breaks touch input. Then there's the matter that upscaling or scaling by a fractional value will result in the system having to compensate and do some kind of filtering (because Nearest neighbor scaling ain't so pretty at things other than integer scaling values) which will blur the image or result in loss of details. Letting the system choose the details isn't quite Apple's way. It won't "Just work!".
Now, not saying Apple wouldn't go this route, but obviously, if they do, it's going to require quite a bit of work on the part of developers.
If anything, I would have thought they would have gone iPhone 5 on the iPad Mini, make it 1366x768, 16:9. That way, original iPad apps can run letter boxed like iPhone apps do on iPhone 5, and you still have something close to the specs of the competition running 720p HD resolutions. This 4:3 parts leak however tells me it's either going to be full on Retina like The new iPad or just plain iPad 2 1024x768.
Yeah, if my math is right, that's 204mm, or 8.03" diagonally. Only think I can think of is that it;s the measurement of the entire assembly, which may be a bit larger than the visible screen area. I assume it needs to be somewhat larger to be able to secure it in the assembly.
not sure if I make sense ...
All I meant is it appeared to me it could be possible for devs in this 1.5x resolution case to not change anything to their code:
The Multi-Touch system give X,Y approximated value of the touch point.
Devs could still have to consider the screen as a 1024x768 pts canvas in their app.
The OS also deals with the 1.5x scaling for display, as it already does at 2x with the retina iPad.
Devs not only supply 1x and 2x bitmap elements for their apps but also a 1.5x version, for a better rendering on screen.
Well, I'm not especially fond of this idea anyway. But your initial statement I quoted seemed to me too definitive.
If the rumoured battery specs are true, i.e. 16.7 Wh for the iPad mini, compared to 25 Wh for the iPad 2 and 42.5 Wh for the 3rd gen iPad, I'd say 1024x768 would be more likely.
So much for doubling down on secrecy.
I'm sorry the system has reached its limit for this comment input. Please type your comment again. Please add additional data.
That's already the case. But like I said, in a 1.5 scaling mode, you'd have fractions. Now is that fraction a touch to the left or right ? Is either really what the dev wants anyhow ? We don't know, since he's never had to deal with this case.
Like I said, it's not ideal because of this. I wouldn't put my chips on this solution, Apple is much more likely to go 1024x768 or full retina before they go half way at x1.5.
I'm not saying they wouldn't, but just that it is going to mess with touch targets in apps that assume integer point systems.
You sound like someone who talks very "matter of fact" about tech rumors until the actual device comes out with different specs...and then you talk about how you knew it all along.