Value of 3D display in a handheld *spawn

[Deleted member 13524]

Yes, the brain joins left and right images together, those images have a horizontal perspective shift not a vertical one, so you get extra resolution horizontally, not vertically.

RealD cinema glasses use circular polarisation to block out sequentially displayed left/right eye images, this is not the same as a parallax barrier.

Of course it's not the same system as a parallax barrier, geez.. I was just saying the increased resolution "perception" is the same.
The RealD movies are also projected at half the resolution (compared to the standard digital projection ones), but the optics experts say the perception is the same as if it was a full resolution movie.


In the 3DS, the brain sees one 3.53" screen with a 5:3 aspect ratio and 192k pixels in it. Assuming an equal spacing between the pixels (the eyes could never distinguish those, nonetheless), do the math yourself and see how many columns and lines of pixels the brain will see.
I don't know how else I'm going to explain this, sorry.

I'm not sure what "today's general opinion" on current autostereoscopic screens really is.

I meant general opinion about the immersion in a movie using stereoscopic screens, from the cinema to hometheater setups.
There's no general opinion on autostereoscopic screens because those screens are hardly available to the mass market right now.

 

[Arwin]

I think it is hard to quantify what 3D depth information brings. It adds a whole new dimension, and that could de-emphasize a lot of the regular 2D information, depending on what you're looking at, while your brain is able to construct depth information which leads to who knows how much (or little) additionally inferrable information. I think to understand how far that can go 3D displays need to evolve further and people get used to it, but simple tests like putting my head to the side of a wall and then look at it with one or two eyes makes a big difference.

 

[JohnH]

Of course it's not the same system as a parallax barrier, geez.. I was just saying the increased resolution "perception" is the same.
The RealD movies are also projected at half the resolution (compared to the standard digital projection ones), but the optics experts say the perception is the same as if it was a full resolution movie.


In the 3DS, the brain sees one 3.53" screen with a 5:3 aspect ratio and 192k pixels in it. Assuming an equal spacing between the pixels (the eyes could never distinguish those, nonetheless), do the math yourself and see how many columns and lines of pixels the brain will see.
I don't know how else I'm going to explain this, sorry.

The brain can't generate additional vertical resolution if there is no vertical difference between the two images i.e. the images are NOT offset vertically, only horizontally.

RealD results in a percieved increase in resolution in both directions due the time sequential nature of the images afaik.

By the way, 3DS's use of a autosteroscopic display isn't particularly innovative, it's done been done before,

"These systems are referred to as Autostereoscopic displays. They were initially developed by Sharp. The first Autostereoscopic LCD displays first appeared on the Sharp Actius RD3D notebook[15] and the first LCD monitor was shipped by Sharp in 2004 for the professional market.[16] Both have since been discontinued. The first Autostereoscopic mobile phone was launched by Hitachi in 2009 in Japan and in 2010 China mobile is to launch its version."

 

[Deleted member 13524]

The brain can't generate additional vertical resolution if there is no vertical difference between the two images i.e. the images are NOT offset vertically, only horizontally.

RealD results in a percieved increase in resolution in both directions due the time sequential nature of the images afaik.

For the left eye, the pixel in position X,Y has a different color than the pixel in position X,Y for the right eye. It's a different viewing angle alltogether. The same applies for any other position.

IMO, you're weirdly assuming the games will be a tile constantly scrolling horizontally. They're not.

 

[Simon F]

The 400*240 images are not side-by-side, they're in the same place (at least that's how the brain interprets it).

I think John H knows how they work

 

[JohnH]

For the left eye, the pixel in position X,Y has a different color than the pixel in position X,Y for the right eye. It's a different viewing angle alltogether. The same applies for any other position.

The image has horizontal perspective shift not a vertical shift, the parallax barrier splits the image into left and right nothing else. If you move the image vertically it could look very odd gien the normal arrangement of human eyes.

That said, the app _could_ render left and right time separated, however I'm not convinced that isn't going to look strange given that the images are both visible at the same time (for realD they're displayed sequentially not at the same time), I'm also not convinced that object wouldn't need to be moving in order for you to precieve higher resolution.

Incedentaly higher resolution is the wrong term for these effects, it's more like a form of antialiasing.

IMO, you're weirdly assuming the games will be a tile constantly scrolling horizontally. They're not.

You seem very sure of yourself, you're getting this information from where exactly?

 

[Exophase]

Cool, let's play that numbers game again.

Now go to the same page and look at the 20 best-selling DS titles. How many of those do NOT depend heavily on the touchscreen?
- New Super Mario Bros
- Mario Kart DS
- Dragon Quest IX

3, out of the 20 best-selling games, could be made without the secondary screen. Looking down the rest of the titles and except for half-a-dozen of J-RPGs, all the titles depend on the touchscreen.

Of course the games library was responsible for its success, but it was the touchscreen (by 2004, a "gimmick") who allowed the brain-trainings, cooking-mamas and wario-wares that launched the console into stardom back in 2005.

Nonsense. Many games do use dual screens + touch for something meaningful - I'm sure Nintendo has a policy to at least try - but that doesn't mean that their gameplay utterly relies on it. This is certainly not the case for the Pokemon games that permeate the list, or the games like Yoshi's Island that worked fine in their incarnations without it. Some other games use it as a poor substitute for proper analog controls, like Mario 64 or (IMO) the Zelda games. Nintendo obviously got the picture since they're including analog on 3DS, which if you ask me is a more important change than the 3D display even though it's a far less innovative one. Very few of the games are really designed around touch input, and a second screen by itself is more of a hindrance than anything, given the number the try splitting a single display between the two. It brings little to the table outside of form-factor consolidation.

You speak as if the DS was Nintendo's strategy to success. In reality Nintendo has owned the handheld market since its inception, and the wildly successful GBA was the antithesis of innovation. You'd have a difficult time showing any evidence that a Nintendo handheld wouldn't be successful regardless of the gimmicks.

 

[Deleted member 13524]

You speak as if the DS was Nintendo's strategy to success. In reality Nintendo has owned the handheld market since its inception, and the wildly successful GBA was the antithesis of innovation. You'd have a difficult time showing any evidence that a Nintendo handheld wouldn't be successful regardless of the gimmicks.

 

[Deleted member 13524]

The image has horizontal perspective shift not a vertical shift, the parallax barrier splits the image into left and right nothing else. If you move the image vertically it could look very odd gien the normal arrangement of human eyes.

The horizontal displacement between left and right eyes gives you a different value for each pixel "place" seen in the screen. There's different information being sent to each eye. -> except for objects that are rendered very "far away" (i.e. stars and the moon in a background), in which case, the value for each pixel position should be the same for each eye.
It doesn't matter if the hardware renders it differently, I'm talking perception.

The same way as super-sampling anti-aliasing gives you the perception of looking at a scene rendered at a higher resolution.

Incedentaly higher resolution is the wrong term for these effects, it's more like a form of antialiasing.

Yes, SSAA or simply rendering at a higher resolution and then downscaling.
In both cases, unless you're close enough to the screen to discern the pixel placement, you get the perception of a higher rendering resolution.

You seem very sure of yourself, you're getting this information from where exactly?

I wrote here all the logic I followed, all the steps I took. Which part didn't you understand?

 

[Exophase]

On the contrary ToTTenTranz, Gameboy Micro seemed pretty gimmicky to me. It was a revised form factor that no one really wanted, backed by questionable marketing (I remember Reggie Fils-Amie actually called it "hip") and at a price point that was actually well beyond GBA SP. Add to that problems with dead pixels and it all adds up to a winning combination.

Kind of hard to compare a one-off model to an entire platform. If anything I'd say something like GBM earning money even is more than successful enough for Nintendo, which I'm sure it did, having sold a couple million IIRC.

But you seem to be under the impression that I said Nintendo will always succeed, which is hardly the case. What I said is that Nintendo doesn't need gimmicks to succeed, something that has been demonstrated several times over. If anything one of Nintendo's more gimmicky endeavors was also their biggest failure (Virtual Boy), I mean, if you want to continue the specious arguments.

 

[JohnH]

The horizontal displacement between left and right eyes gives you a different value for each pixel "place" seen in the screen. There's different information being sent to each eye. -> except for objects that are rendered very "far away" (i.e. stars and the moon in a background), in which case, the value for each pixel position should be the same for each eye.
It doesn't matter if the hardware renders it differently, I'm talking perception.

*Sigh* I have no idea why you can't grasp the simple fact that you're not generating any extra vertical resolution so there is nothing for the brain to interpolate between, "perception" doesn't change this.

The same way as super-sampling anti-aliasing gives you the perception of looking at a scene rendered at a higher resolution.

It isn't the same way super sampling works, super sampling is a HW process not somethign done by the brain. They are completely different processes.

However, if you truely beleive it's the same as SSAA, what resolution do you think a down sampled 800x240 would equate to?

Yes, SSAA or simply rendering at a higher resolution and then downscaling.
In both cases, unless you're close enough to the screen to discern the pixel placement, you get the perception of a higher rendering resolution.

Again this process has nothing to do with super sampling, but is effectively an anti-alias process.

I wrote here all the logic I followed, all the steps I took. Which part didn't you understand?

Eh? You're making baseless claims, none of your "logic" is explaining those claims, it's becoming rather like talking to a 5 year old child who is stomping their feet because they can't get their own way!

 

[Deleted member 13524]

*Sigh* I have no idea why you can't grasp the simple fact that you're not generating any extra vertical resolution so there is nothing for the brain to interpolate between, "perception" doesn't change this.

It isn't the same way super sampling works, super sampling is a HW process not somethign done by the brain. They are completely different processes.

However, if you truely beleive it's the same as SSAA, what resolution do you think a down sampled 800x240 would equate to?

Again this process has nothing to do with super sampling, but is effectively an anti-alias process.

I use the terms "as if", "the same perception as", I write the main words in bold.. And geez.. never once I said something as stupid as SSAA being done in the brain.. and yet you continue to distort my words in order to have your way.

And in the end, you preach some stuff about reacting childishly (which is a bit ironic, in my opinion).

Guess what? Keep the "Guru of the Year" belt, I won't start an insult war just for trying to have a reasonable argument.
I'm off.

 

[cjo]

ToTTenTranz,

The point is, the different X,Y values don't give you any more information vertically than a 2D image. If you take a 2D "image" a single pixel tall and then add the equivalent right eye image, there isn't any extra information about what is above or below the left eye image to allow your brain to infer anything about what is going on above or below the original left-eye image.

The left-right eye images are a horizontal offset, so the brain may be able to reconstruct more information than you would see in one image horizontally (only for objects that are visible in both eyes) by getting slightly different sample points for each object but as properly done 3D images are *only* offset horizontally, you don't get the same overlap of information vertically, which means there isn't a perceived increase in image quality vertically.

 

[Deleted member 13524]

The left-right eye images are a horizontal offset, so you may well get the effect of anti-aliasing (only for objects that are visible in both eyes) by getting slightly different sample points for each object but as properly done 3D images are *only* offset horizontally, you don't get the same overlap of information vertically, which means there isn't a perceived increase in image quality vertically.

And how exactly does the human brain discern a difference between horizontal and vertical image quality, in a ~280 ppi screen?

 

[cjo]

Just one more thing: we cannot really say it's 400*240. It's 400*240 for each eye and they're looking at different pixels, so the brain should interpret the acquired image as something like ~565*340 (twice the pixels at the same screen ratio).

And how exactly does the human brain discern a difference between horizontal and vertical image quality, in a ~280 ppi screen?

That's a different question entirely. The point is, it won't magically go from 240 lines to 340 (or any other random number), because there just isn't the information there for your brain to interpret.

 

[JohnH]

I use the terms "as if", "the same perception as", I write the main words in bold.. And geez.. never once I said something as stupid as SSAA being done in the brain.. and yet you continue to distort my words in order to have your way.

And in the end, you preach some stuff about reacting childishly (which is a bit ironic, in my opinion).

Guess what? Keep the "Guru of the Year" belt, I won't start an insult war just for trying to have a reasonable argument.
I'm off.

*Yawns*

 

[Xmas]

And how exactly does the human brain discern a difference between horizontal and vertical image quality, in a ~280 ppi screen?

How do you get 280 ppi? It's 264 ppi horizontally and 132 ppi vertically.

 

[Eteric]

Hrm.

I think the problem here is that you guys aren't thinking from the perspective of the average customer. The average customer doesn't really care what the screen resolution on their handheld is, so long as it looks nice and has games that they'd like to play.

The 3D is there to separate the 3DS from the DS. It's there to convince people to buy it, and show it off to their friends (a sort of viral effect). It's also there to show people something they've never seen before.

Nintendo was a toy company for a long time. And in a lot of ways, they still think like a toy company. A device with glasses-less 3D, that takes 3D pictures, etc, is fun and interesting to people.

 

[Deleted member 13524]

http://www.ubergizmo.com/2011/02/rightwares-stereoscopic-3d-android-homescreen-picture/

Pretty much every new high-end SoC is claiming 3D-stereoscopic capable display support.
It seems to me that 2012\13 will be the year of stereo-3D standardization in handhelds.

 

[JohnH]

This is not unexpected. If the handset producers demand it the SoC vendors have no choice but provide. If it can re-vitalise TV sales, the handheld marketing bods will be looking for the same in handheld. It doesn't mean that it still isn't a silly gimmick.