3 ms·
This reminds me of the BadAstronomy writeup about Jobs and the resolution of the iPhone 4: http://blogs.discovermagazine.com/badastronomy/2010/06/10/resolving-
by waivej 14y ago
This reminds me of the BadAstronomy writeup about Jobs and the resolution of the iPhone 4: http://blogs.discovermagazine.com/badastronomy/2010/06/10/resolving-the-iphone-resolution/ http://blogs.discovermagazine.com/badastronomy/2010/06/10/re...
- CoolGuySteve 14y agoYa, I was thinking the same thing. At 440ppi, this seems like overkill, considering someone with 20/20 vision only needs around 300ppi for things to be acceptable. I was hoping with the retina marketing blitz, Apple would put the density comparisons to bed and we could start thinking about displays with wider colour gamuts, wider viewing angles, better outdoors performance, 3D, mini projectors, etc. I understand hard numbers like ppi are easier to compare between and display in huge letters in a BestBuy flyer, we called them 'checkbox features' at a place I used to work at, but at some point they distort the overall quality of the product. For example, the Pentium 4's clockrate and megapixel ratings on cameras are cases of a checkbox feature actually decreasing the quality of most products in the market.
- bane 14y agoInterestingly, in high-end printing, 300 dpi is considered unacceptably poor, most print shops will simply refuse to print artwork at 300 dpi. 600 dpi is usually what's considered minimally acceptable for artwork and around 1200 dpi for lots of text.
- CoolGuySteve 14y agoDPI is very different though, a screen sub pixel can vary in luminescence (typically index in most displays with 6bits of accuracy, minus crappy colour gamut on the display) whereas a printed dot is limited to the ink colours used. Some printers overlap dots, print thicker dots, and have other tricks to work around this aspect. I don't know as much about this area, I've worked mostly in graphics, but my understanding is that because the colour gamut for each printed 'pixel' is less expressive, you need more dots in the same area to get a similar result. Furthermore, dpi expresses printed 'dots' whereas a pixel on most displays can be considered three fixed 'dots', one for each display channel. Although lately, some LCD substrates are cheating a little and displaying only 2 subchannels per pixel. Anyways, that's everything I know about dpi, as I said, I'm more versed in video rather than printing. So take this with a grain of salt: I think you need to multiply PPI by 3 to get DPI, ignoring cleverness on the part of either display or printer manufacturers.
- bane 14y agoThat's basically right in some cases. For example in full-color (4 color) printing you essentially need 4 dots (halftones) to make any color. However, they aren't all necessarily lined up in a grid like pixels (it's an analogue medium). But simply dividing 1200 dpi by 4 doesn't really provide a correct calculation for the resolution at the end. http://blog.savvydog.com/?p=106 http://blog.savvydog.com/?p=106 And then you actually get into measuring with LPI (lines per inch) which can easily go above 300 for things like good magazines (though around 130-150 is more typical). 130-150 is actually around the retina display's 300dpi (a good conversion estimate is that 1LPI=2DPI but some will go as high as 3DPI). Black and white (non-process) printing can be very different and uses solid colors. Higher quality artwork is necessary for this. That's where the 1200dpi comes into play. Fine books, black and white artwork, etc. need artwork at 1200dpi and the LPI can easily go above 500.