Randomoneh

Randomoneh

Joined on Aug 1, 2012

Comments

Total: 12, showing: 1 – 12
On Hands-on with the HTC One post (30 comments in total)
In reply to:

solarsky: 4 Ultra-pixels, oh my goodness ;-)...
It's just a normal 2µm 4MP Bayer-Sensor combined with some "propaganda".
So it only has an effective 1 MP true RGB-resolution.
Totally lame. My Nokia 808 has a 1.4µm 41.48 MP Bayer-Sensor which gives me effective 10.37 MP of true RGB-resolution.
HTC will just be trying to make the best out of additional well-capacity offered by larger sensor diodes and BSI-design. The rest is anything but "PureView", as clearly shown by the test photos from the HTC One which are circulating on the web.
The whole concept of this phone's camera could have been a lot more interesting, if they had managed to integrate a 4MP global shutter sensor, instead of THIS old sham...

Solarsky is right. Take a look at sample images full-screen on a regular 1080p or 2560*xxxx display. Blurry.

Direct link | Posted on Mar 12, 2013 at 00:13 UTC
On HTC One: Is it the the ultimate camera phone? post (78 comments in total)
In reply to:

kff: Senzor similar Sigma Faveon .... ?

Are you sure? I see a lot of "every element captures RGB" talk online.

Direct link | Posted on Feb 20, 2013 at 02:21 UTC

Can anyone give final answer - is it Bayer? I'm finding contradictory statements.

Direct link | Posted on Aug 2, 2012 at 23:43 UTC as 5th comment
In reply to:

AndyGM: This is a good summary of the work NHK have done:

http://www.nhk.or.jp/strl/english/aboutstrl1/r1-1-1.htm

Figure 2 is very interesting, it is demonstrating the results of 2 "perception tests" they have carried out using 8K displays.

The 2 perceptions they have tested are "sense of realness" and "sense of being there". If you sit closer to the screen, it fills more of your field of view so your sense of being there goes up, but you might start to be able to see the pixels in the image, so your sense of realness goes down.

The sense of being there stops scaling linearly at about 70deg FOV, this is a point of diminishing returns. This is also roughly where the 2 lines cross (where the sum of both perceptions is at a maximum). This corresponds with a viewing position somewhere between 1 and 1.5 times the picture height. I will leave it to Randomoneh to work out how many Pixels/deg this is, but its still more than enough for most people to not see individual pixels.

@AndyGM
Absolutely.

Direct link | Posted on Aug 2, 2012 at 19:08 UTC
In reply to:

malcolm82: From Engadget:
"Ultimately, there's plenty of reason to believe the BBC's project head, Tim Plyming, when he says that "8K is the maximum the human eye can understand" and that "it's the end of the resolution story.""

That's only true for a field of view equivalent to about 40mm (or about 55-60° diagonal). For ultra wide angle viewing we need much higher.

@Cy Chaze

But you'll move your eyes around thus every part of display has to match the limitation of your eyes - from nearest (center) to farthest (edges).

Direct link | Posted on Aug 2, 2012 at 14:59 UTC
In reply to:

AndyGM: This is a good summary of the work NHK have done:

http://www.nhk.or.jp/strl/english/aboutstrl1/r1-1-1.htm

Figure 2 is very interesting, it is demonstrating the results of 2 "perception tests" they have carried out using 8K displays.

The 2 perceptions they have tested are "sense of realness" and "sense of being there". If you sit closer to the screen, it fills more of your field of view so your sense of being there goes up, but you might start to be able to see the pixels in the image, so your sense of realness goes down.

The sense of being there stops scaling linearly at about 70deg FOV, this is a point of diminishing returns. This is also roughly where the 2 lines cross (where the sum of both perceptions is at a maximum). This corresponds with a viewing position somewhere between 1 and 1.5 times the picture height. I will leave it to Randomoneh to work out how many Pixels/deg this is, but its still more than enough for most people to not see individual pixels.

You're confusing cycles per degree with pixels per degree. 1 cycle per degree = 2 pixels per degree. 156 cycles per degree = 312 pixels per degree.

But if you read their research papers, it seems limit is ~200 pixels per degree (100 cycles per degree). I can link you to pdf if you want.

Direct link | Posted on Aug 2, 2012 at 14:51 UTC
In reply to:

malcolm82: From Engadget:
"Ultimately, there's plenty of reason to believe the BBC's project head, Tim Plyming, when he says that "8K is the maximum the human eye can understand" and that "it's the end of the resolution story.""

That's only true for a field of view equivalent to about 40mm (or about 55-60° diagonal). For ultra wide angle viewing we need much higher.

@beholder1
That is pretty weird thing to say (human central vision is 20/20). First of all, you're talking about measurement of visual acuity done with Snellen chart. Second, 20/20 is nominal, not average value. Third, do you think you can simply translate Snellen chart value to ability to benefit from x angular resolution? It's not as simple as that.

@malcolm82
I wanted you to check out their (NHK) research on subject.

Direct link | Posted on Aug 2, 2012 at 01:42 UTC
In reply to:

malcolm82: From Engadget:
"Ultimately, there's plenty of reason to believe the BBC's project head, Tim Plyming, when he says that "8K is the maximum the human eye can understand" and that "it's the end of the resolution story.""

That's only true for a field of view equivalent to about 40mm (or about 55-60° diagonal). For ultra wide angle viewing we need much higher.

@malcolm82
Some of my info on that Macrumors thread is not correct and I don't feel like editing it :D
Oh well...

Direct link | Posted on Aug 1, 2012 at 22:40 UTC
In reply to:

malcolm82: From Engadget:
"Ultimately, there's plenty of reason to believe the BBC's project head, Tim Plyming, when he says that "8K is the maximum the human eye can understand" and that "it's the end of the resolution story.""

That's only true for a field of view equivalent to about 40mm (or about 55-60° diagonal). For ultra wide angle viewing we need much higher.

@malcolm82

Google for "NHK realness".

Direct link | Posted on Aug 1, 2012 at 21:40 UTC
On NHK working on 8k video sensor capable of 120fps article (101 comments in total)
In reply to:

pitou250: If a human has a field of view of ~155 degrees horizontally by 105 degrees vertically - and the angular resolution of the eye is approximately 1 arcminute (1/60th of a degree), covering the entire view of a human with maximum distinguishable resolution would require 155x60 = 9300 pixels horizontally by 6300 pixels - or ~58MP.
(sorry if my figures are off - I don't claim accuracy)
Screens do take up less than the entire fov in normal viewing situations though - let's say half for a very big screen, that's 4650x3150pixels or 14,5 MP.
Except for immersive content and displays, I can't see the use of 8k for straight 'human consumption' and by the time it arrives in my living room, my eyesight will not be up to it anyway.

That's not how it works. If you calculate it by simply multiplying then center of the display won't match that limit.

Direct link | Posted on Aug 1, 2012 at 21:20 UTC
In reply to:

malcolm82: From Engadget:
"Ultimately, there's plenty of reason to believe the BBC's project head, Tim Plyming, when he says that "8K is the maximum the human eye can understand" and that "it's the end of the resolution story.""

That's only true for a field of view equivalent to about 40mm (or about 55-60° diagonal). For ultra wide angle viewing we need much higher.

Actually, it depends on what we say is a limit.

If limit is 0.3 arcminutes per pixel (200 pixels per degree), than 8K is enough for 37 degrees of person's field of view horizontally. Remember, you can't simply divide horizontal resolution with limit (number of pixels per degree) to get number of degrees of field of view which display can occupy while matching limit of vision. Why? Well, display is straight, isn't it?

Direct link | Posted on Aug 1, 2012 at 20:46 UTC

It's about angular resolution. If we speak strictly, highest tested angular resolution that test subjects were able to benefit from is 156 pixels per degree [of person's field of view]. Data suggests actual limit would be about 200 pixels per degree.

If we take 200 ppd (0.3 arcminutes per pixel) as a limit, 8K would match that limit when 8K display itself occupies 37 or less degrees of person's field of view horizontally, meaning we would need even higher resolution if we want our display to occupy more that 37 degrees.

If we're talking about 0.5 arcminute per pixel (120 pixels per degree):
4320p would mean optimal sitting distance of 1.6x (1.59) picture height or farther, making display occupy 58.4 degrees of field of view horizontally or less. With 2160p, optimal sitting distance would be 3.2x (3.18) picture height or farther, making display occupy 31.24 degrees of field of view horizontally or less.

Direct link | Posted on Aug 1, 2012 at 20:41 UTC as 12th comment
Total: 12, showing: 1 – 12