4 ms·

Well, looking at the keynote slide, the samples are heavily zoomed in so 'old' pixels are depicted as tens of screen pixels high. For an honest portrayal, it's

Well, looking at the keynote slide, the samples are heavily zoomed in so 'old' pixels are depicted as tens of screen pixels high. For an honest portrayal, it's easy: 'new' pixels just need to be depicted as half-as-high. Sorry, but the '300dpi' blather is complete reality-distortion nonsense. :)

jfoutz 16y ago
Do you have the slides? I just saw the image from the guy's site. I'm not sure how much the angle of the photo will compress the distant (high quality) letter. those displays are huge, so maybe the letters are 30' apart? it's not clear how far away the camera is. They are kind of weird samples. For example the big pixel side is highlighting 25 pixels, but you could easily draw a 16 pixel square and assert the light row should go in the next sample. Also, the big side is taken from a point with a -1 slope, but the little side is much steeper, maybe 5? It seems like you'd want to compare the same section of both letters. Also, it's not really clear to me why each red box in the lower left sample has 2 or more colors. That seems like a good indicator that the red box != one pixel.
- confuzatron 16y ago
  Sorry, no, when I say slide I'm looking at the photo of Steve Jobs with a big screen behind him showing two letter 'a's. One is heavily pixellated, and the other one is smooth in comparison - at the resolution of the slide display. Given the size of the old pixels, the new pixels should be easily visible but aren't. Of course it could be a misleading screen-grab and the slide was not purporting to show old iPhone resolution next to new iPhone resolution here. Maybe it's 'old' compared to 'perfect'. (I should point out, my participation in this crucial discussion is dependent entirely on the slowness of my compiler).
- DougBTX 16y ago
  The second, black and white, example is a screen shot from the Apple website: http://www.apple.com/iphone/features/retina-display.html http://www.apple.com/iphone/features/retina-display.html The interactive examples further down appear to be realistic though.

jfoutz 16y ago

Do you have the slides? I just saw the image from the guy's site. I'm not sure how much the angle of the photo will compress the distant (high quality) letter. those displays are huge, so maybe the letters are 30' apart? it's not clear how far away the camera is. They are kind of weird samples. For example the big pixel side is highlighting 25 pixels, but you could easily draw a 16 pixel square and assert the light row should go in the next sample. Also, the big side is taken from a point with a -1 slope, but the little side is much steeper, maybe 5? It seems like you'd want to compare the same section of both letters. Also, it's not really clear to me why each red box in the lower left sample has 2 or more colors. That seems like a good indicator that the red box != one pixel.

confuzatron 16y ago
Sorry, no, when I say slide I'm looking at the photo of Steve Jobs with a big screen behind him showing two letter 'a's. One is heavily pixellated, and the other one is smooth in comparison - at the resolution of the slide display. Given the size of the old pixels, the new pixels should be easily visible but aren't. Of course it could be a misleading screen-grab and the slide was not purporting to show old iPhone resolution next to new iPhone resolution here. Maybe it's 'old' compared to 'perfect'. (I should point out, my participation in this crucial discussion is dependent entirely on the slowness of my compiler).
DougBTX 16y ago
The second, black and white, example is a screen shot from the Apple website: http://www.apple.com/iphone/features/retina-display.html http://www.apple.com/iphone/features/retina-display.html The interactive examples further down appear to be realistic though.