5" screen (on a P2)


with a good LCD controller, a 18bpp LCD screen can be space-and-time dithered to display 24bpp just fine.

The quality of spatial dithering is dependent on resolution. The quality of temporal dithering is dependent on refresh rate and response time (ideally a poorer response time in conjunction with a higher refresh rate is what you want, since you want a low pass filter on top of the oversampling.. but that throws out high frequency data and results in ghosting).


So we can't really talk about how effective dithering is without knowing these parameters. But I'd prefer to not have to worry about it altogether.


While there are lots of 6bpp displays using dithering today there are also plenty of real 8bpp and higher displays out there. You'd think if it had no impact they wouldn't bother.

black and white resolution for browsing ... TEXT :)

You don't like anti-aliased text..? Your browsing only consists of text?? o_O

most of the emulated consoles only generates a 15bpp picture anyway, because its good enough.

True, although the color space may not match up that well. byuu for instance likes pushing 10bpp displays to best accommodate whatever gamma corrections he's doing. If you're outputting with some kind of NTSC and/or TV phosphor/scanline simulation having more depth helps..


Anyway, PS1, N64, PSP, Saturn, 3DO, Jaguar, and Dreamcast at least have 24bpp capability. DS has 18bpp capability.

what does make a huge difference is the LCD signal generator's capability to correct the separate color curves and apply dithering.

we had to do that in software on GBA, color-correct all the source images on PC and it did a huge difference on the picture quality.

Yeah, but GBA, especially the original, had a really messed up color space. Since most stuff wasn't using direct color (mode 4) I doubt you were actually dithering the images rather than just transforming the palette data.


Do a lot of mobile 18bpp LCDs even have 8-bit wide component interfaces? And to do dithering they'd need enough memory to store at least a few lines before they display them.. You'd probably be relying on the SoC's display controller to do dithering instead.
 
Last edited by a moderator:
That's hot :wub:


logo%20full%20hd%201080.jpg
 
As noted by frz in this thread:


http://boards.openpandora.org/index.php?/topic/9007-the-ideal-screen-for-the-pandora-2/


LG has shown a prototype of a 5" 1080p screen.


http://www.engadget.com/2012/05/27/lg-display-five-inch-1080p/


1920x1080 pixels in 5" comes out to 440ppi.


There are some real solid brag stats to saying that your handheld device can play 1080p videos ON the handheld device in full resolution.


When you cross over past 300ppi you no longer see the pixels. BUT you can still tell the difference. Look at an old iPad next to a new retina iPad, both displaying the same high resolution picture. You can't see the pixels on either one, but the difference is startling in the quality of the image.


So - contact LG and see if they'd like to sell you batches of five thousand of those at $50 each. :blink:
 
Last edited by a moderator:
I think you should focus more on the behaviour with direct daylight than resolution.


The resolution won't make the device more useful it will be just a "oh, look we have got it bigger" factor.
 
I'm reasonably sure the 1080 screen will be one of the best in most, if not all areas, including viewing.


Resolution strategy seems to work for Apple and Samsung.


Besides, if we can have the 1080 screen, then we should.


The P2 has to be the best handheld in the world, ever.
 
Last edited by a moderator:
Don't forget that more pixels to draw also means you need a faster GPU and CPU and more video memory. If you really want a screen with such an insane resolution, I would probably want to use it in a downscaled resolution: for me, one quarter of 1920x1080, which would be 960x540 (still more than the current Pandora!) would be more than enough, and I'd rather have more RAM and processing power than wasting those resources on rendering pixels I cannot see.
 
Last edited by a moderator:
d I'd rather have more RAM and processing power than wasting those resources on rendering pixels I cannot see.

Exactly. Most people seem to think pixels are "free" to render. They are not.

Personally i would also prefer a 720p screen in the p2.


Lets keep 1080p for the pandora 3 :p


If i could pick between Ps2 emulation on a P2 with 720P screen, or having a 1080P screen but no ps2 emulation, i would go for 720p.


Perhaps we could somehow test what would be possible emulation wise when having a 1080P screen, as opposed to a 720P screen.


If a 1080P screen would eat away too much resources, to do stuff like ps2, psp and windows xp, then it would not be worth it.


Ofcourse all the above is hypothetical, i have no idea how well ps2 and xp would do on the pandora 2.
 
Last edited by a moderator:
PS Vita has a big (in my opinion) flaw: unless I'm indoors, it's REALLY hard to see what's on the screen. Do you guys think this could happen with P2? Why does this happens? Is that because of the OLED screen?
 
A 1080p 5" screen would also consume a lot more power and generate a lot more heat. iPad 3 is a perfect example of this phenomenon. The temperature impact would be even worse given the increased density for all those pixels.

I'm reasonably sure the 1080 screen will be one of the best in most, if not all areas, including viewing.

Why? You don't think optimizing for resolution could come at the expense of anything else? You don't think going with a screen that will ostensibly be provided by very few sources won't limit your ability to select based on other parameters like response time, ones that absolutely cannot be compromised?

Resolution strategy seems to work for Apple and Samsung.

Everything works for Apple. Before they touted Retina displays they were already selling phones like wildfire, phones that were nominally much lower resolution than everyone else. Think about what makes Apple products sell, and think about the consumer reaction of OPT actually tried following the same path (I'm quite sure you wouldn't get very far this way). Samsung has their own particular strengths that are also difficult to imitate.

Besides, if we can have the 1080 screen, then we should.

You were saying that Pandora 2 must be $699 in order to be financially sustainable. Surely a 5" 1080p screen in 2013 will cost significantly more than the alternatives, if even purchasable in low quantity. Do you consider that it needs the 1080p screen in order to be financially sustainable? Because I don't think many people will agree with this. Or are you saying you'll push the price even higher to get this screen?

The P2 has to be the best handheld in the world, ever.

Well it probably won't be, particularly not in every way possible.


Samsung and now Apple (and Sony, if you want to compare with PSV, which would be an inevitable comparison) are making their own SoCs. Samsung's may in theory be available for use in other products but in practice they rarely are, and are certainly available for Samsung products first. The others are of course exclusive to that vendor. Other SoC vendors will often not sell their best stuff in low quantity. nVidia for instance, who has been consistently winning in time to market, set minimum orders of 100,000.


Let's say you get an OMAP5 in Pandora 2. Its GPU will be behind some devices already out now. Who is to say that an Exynos 5 device won't beat you to market, or that Apple A6 won't also have a decent dual or quad A15? Then there's Snapdragon S4 Pro with quad Kraits and Adreno 320..


A small player is going to be disadvantaged here, no matter how high you set the unit asking price. The "world's most powerful" tagline was ultimately not the factor in the device's feasibility and survival. And even if it met its original 2008 release it would have still been beaten to market by OMAP3530 Archos PMPs already available (that also had a comparable screen, RAM, flash/hdd, etc). Back then TI was willing to push its high end SoC into low volume channels, since then things haven't been so nice (there's the DaVinci line but there's no OMAP4 level DaVinci, and while there are quad Cortex-A15 + high end DSP devices announced they're far from suitable nor intended for handhelds).. and now other companies seem to be playing spec wars a lot more heavily in the handheld space.
 
Back
Top