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.
OS X Lion's HiDPI Modes Lay Groundwork for Retina Monitors
- Thread starter MacRumors
- Start date
- Sort by reaction score
What do you mean you have 50 of them? I don't get what you're saying here.
Regarding file sizes, my challenge remains. If anyone can recreate one of Apple's traffic light window controls in Illustrator and output a file that is smaller in size than the equivalent bitmap at twice the current dimensions (anything bigger would be pointless), I'll take back every word I've said on the topic!
Actually, this made me think right away of my mom, who uses a much-lower-than-native resolution on her Mac (something like 1024x540 on a 19" monitor!) because that's the only way she can read most of the screen's contents. Now if we just had the money to switch out her six-and-a-half-year-old PowerBook G4 for something Lion will actually run on...
They are likely done in Photoshop using vector masks. This way you have pixel-level accuracy but still have the ability to adjust vectors. Illustrator is pretty terrible when it comes to pixel-level accuracy.
What's hard about that ? Sierra games used to be all made of vector graphics (the PIC format) back in the 80s... :
Yes, that's a bunch of "Line from X to Y, fill with color ZZ".
Vector graphic based UIs are probably where we should be heading. KDE/Gnome have supported SVG for a icon format since the late 90s. That is true resolution independence and something that should be worked on, not just double sized bitmaps.
Yes. Loading a bitmap into texture memory based on a context is far faster than loading a vector and then rendering it out to the correct size in real-time. Perhaps this will be a negligible performance hit in 10 years, but right now, 50 bitmaps vs 1 vector that needs to be re-rendered is always going to be better.
I'm a pixel counting snob but I'm unconvinced this is required
I'll preface my statement by saying I was wrong about how dramatic the difference the retina display makes in iPhones. In hate 3gs's now, and love the 4. Given the screen proximity of my iPad, it could do with doubled (really 4x the total count) pixels, although I think it would be less dramaticly better than the iPhone.
I have a 27" iMac running a 30" and at roughly 2.8' away, I really think the improvement would be a diminishing return, hardly noticeable. I haven't crunched the DPI:average viewing distance numbers, but I think for normal computing (ie. graphics/photo editing not withstanding) it's not worth the GPU strain.
I'll preface my statement by saying I was wrong about how dramatic the difference the retina display makes in iPhones. In hate 3gs's now, and love the 4. Given the screen proximity of my iPad, it could do with doubled (really 4x the total count) pixels, although I think it would be less dramaticly better than the iPhone.
I have a 27" iMac running a 30" and at roughly 2.8' away, I really think the improvement would be a diminishing return, hardly noticeable. I haven't crunched the DPI:average viewing distance numbers, but I think for normal computing (ie. graphics/photo editing not withstanding) it's not worth the GPU strain.
But we're not talking just about UI stuff. We're talking about all graphics, including photorealistic stuff.
arn
You are missing my point!
As I understand it for each resolution there is a certain sized red traffic light right? So say there are 10 different resolutions that means 10 different bitmaps for the red traffic light. With a vector you only need one file which in file-size is less than the combined amount of bitmaps.
I know a tiny bitmap will be smaller in filesize than a vector. I do know how they work!
Its the backend stuff I dont fully understand which I said earlier and so proposed why not use vectors? I'm merely asking questions to better understand the reason why they are not using vectors. I'm not here to accept stupid challenges!
Yeh, I dont think we are 'there' yet, I was just asking the question. As it seems the logical thing to do.
It's a lot easier when you've got 2-bit (4 color), 4-bit (16 color) or 8-bit (256 color) color values. When you're talking about 24-bit (millions) color values it's an entirely different animal. Just that simple aspect has implications across RAM, CPU Cache, GPU memory, the memory bus, etc.
That's because Apple uses a couple of different size icons. They just have 1 big bitmap that they scale. See icon composer and an actual icon file. See attached screenshot for an example of how it works.
What photo realistic UI elements exactly ? All I see are icons and controls in my UI, I don't quite see this "photo realistic" stuff...
Bitmaps also use 24 bits of color (32 bits if you count the alpha channel), I don't quite follow your logic... Heck, bitmaps require 32 bits of data per pixel in RGBA format, whereas with patterns and fills in vector graphics, you can do with much less.
Also, using a SVGZ, a compressed XML format, you can reduce those graphics down quite a bit...
Attachments
are you saying that i will not have a new monitor to upgrade in near future from my dell 30 inch monitor.
it is so sad news.
i thought when osx lion have 3200x2000 desktop picture, apple will sure produce a new monitor, i guess i am wrong.
i am so sad. I have to live with my 30 inch dell for many years to come. what a tragedy.
it is so sad news.
i thought when osx lion have 3200x2000 desktop picture, apple will sure produce a new monitor, i guess i am wrong.
i am so sad. I have to live with my 30 inch dell for many years to come. what a tragedy.
No more complaints about dead pixels!
While adding more pixels increases the likelihood that any one of them might malfunction, it also makes malfunctioning pixels much harder to see. My iPhone 3G has a dead pixel, but I never noticed until I tried different colors on a flashlight app and noticed that one of the red sub pixels was missing. When I upgrade to the iPhone 5, it'll be impossible for me to tell without a microscope. One of my biggest pet peeves is dead pixels, but when they're this small, I won't be able to notice. (Especially since I'll be well over 40 by the time we see retinal monitors on Macs.) Yay for my OCD!
While adding more pixels increases the likelihood that any one of them might malfunction, it also makes malfunctioning pixels much harder to see. My iPhone 3G has a dead pixel, but I never noticed until I tried different colors on a flashlight app and noticed that one of the red sub pixels was missing. When I upgrade to the iPhone 5, it'll be impossible for me to tell without a microscope. One of my biggest pet peeves is dead pixels, but when they're this small, I won't be able to notice. (Especially since I'll be well over 40 by the time we see retinal monitors on Macs.) Yay for my OCD!
This thread is hilarious...
Everything in vector? I'm guessing anybody who suggests that doesn't have much design or development experience.
Everything in vector? I'm guessing anybody who suggests that doesn't have much design or development experience.
Why are you limiting this to UI? It's everything that displays on your computer.
arn
Attachments
So, maybe you'd care to enlighten us ? Maybe I'm just nostalgic for the 80s and 90s, back when vector graphics were a possibility. Now that we have much more powerful CPUs and GPUs, it seems that vector graphics have become way too hard to handle...
Only stuff that's required to scale. IE : the UI.
And games have been using vector graphics since the 80s and 90s, at least, games made with a little known technology called Flash. Or the above Sierra game, Conquest of Camelot
In fact, isn't that screenshot from a game that uses vector graphics ? (Plants vs Zombies ?).Obviously, not all UI elements are needed to scale. The desktop wallpaper can remain a bitmap image, but things like icons, buttons, fonts and all UI controls would need to be scaled.
Think man, think.
If vectors are so simple, why is no one doing it? Reasonable answer is that it isn't that simple to have everything be vectors. It also doesn't deal with legacy software.
If HiDPI can actually work decently for legacy software to give it big high res fonts, while only upscaling some interface elements that are bitmaps then it may be enough to jump-start high DPI monitors.
I expect a 27" 3840x2160 166 dpi iMac within 18 months. Using the 166 dpi 27" panel LG is already talking about(third from left)
If HiDPI can actually work decently for legacy software to give it big high res fonts, while only upscaling some interface elements that are bitmaps then it may be enough to jump-start high DPI monitors.
I expect a 27" 3840x2160 166 dpi iMac within 18 months. Using the 166 dpi 27" panel LG is already talking about(third from left)
Last edited by a moderator:
We've had the ability to do everything in vector for a long time. If it was a good idea, we wouldn't use bitmaps in programs. Its not like Apple is falling behind here, no one does it. Displaying everything as a vector would be way too taxing on our computers and would result in a non-pixel perfect product. And that's why nobody does it. End of story.
Guys... might want to remove that foot from your mouths :
http://www.svgopen.org/2008/papers/104-SVG_in_KDE/
Read. Learn. Again, the open source world isn't afraid of going that extra mile.
Nope. Only the stuff that needs to scale dynamically. Static images can remain bitmap based graphics with just more pixels. And even some vector based static images might make sense. I know KDE supports SVG for a wallpaper format.
I have presented quite a few examples of why vector graphics make sense, and of their use. What is your proof that vector graphics are too hard ? The fact that "no one" uses them ? Well, except for those that do use them it seems.
The plain fact is, for a commercial entity like Apple, it's just cheaper to use bigger and bigger bitmap images. Why try to innovate and reinvent the wheel ?
You do understand that vectors don't exist as vectors once they are rendered right? They become bitmaps and exist in texture memory? At some point you have to convert the mathematical construct into pixels. Doing this in real-time vs loading the exact bitmap is where the performance issues are. Especially when you have very complex renderings with lots of layers. Just look at Adobe Flash, basic vectors no problems, complex illustrations, lots of problems.
The reason why Sierra chose to do their games that way (at least through King's Quest 4) is because of a lack of storage space. In Kings Quest 5, they switched to EGA-based bitmaps/sprites.
Huh? I think your definition of "scale" is different than mine.
bitmap graphics using more pixels is pretty much my definition of scaling something larger.
arn
And games have been using vector graphics since the 80s and 90s, at least, games made with a little known technology called Flash. Or the above Sierra game, Conquest of Camelot
Have you ever heard your CPU fans go into overdrive when you open a Flash page? There are reasons for that.
Also the SCIV games were 16 colors and were obviously vector. By the time they moved on to VGA graphics they no longer used vectors.