Follow Slashdot blog updates by subscribing to our blog RSS feed


Forgot your password?
Iphone Technology

A Professional Perspective On Apple's Retina Display 346

Reader BWJones, who is a retinal scientist, sends in this detailed analysis of the iPhone 4's "retinal display," which includes photomicrographs of the display pixels of earlier generations of iPhone as well as the iPad. Well worth a read. "... as you can see from these images of the displays I captured under a microscope, the pixels are not square. Rather they are rectangular, and while the short axis is 78 microns, the long axis on the iPhone 4 pixel is somewhere in the neighborhood of 102 microns. ... While [an earlier analysis by] Dr. Soneira was partially correct with respect to the retina, Apple's Retina Display adequately represents the resolution at which images fall upon our retina. ... [I] find Apple's claims stand up to what the human eye can perceive."
This discussion has been archived. No new comments can be posted.

A Professional Perspective On Apple's Retina Display

Comments Filter:
  • by Anonymous Coward on Friday June 25, 2010 @10:19AM (#32690366)

    People love to whine about all the Apple stories. I would defy any of them to submit their own stories about all the other computer companies that are breaking new ground with this type of research. Do you think Dell for example has a team of physics PHDs figuring out these technologies and pushing their vendors to tool up for them? No, THOSE are the guys just packaging off-the-shelf reference designs. Or waiting for the exclusivity on Apple's deal with [insert obscure pacific rim manufacturer here] to expire so they can make a similar looking phone a few years later.

    Do you think Apple has a team of physics Ph Ds figuring this out? (Hint: no.)

    Apple deserves credit for identifying this technology and bringing it to market. That's a worthwhile and necessary pursuit, no matter what the Slashdot detractors say. But Apple is not doing groundbreaking research into materials science or manufacturing here; it's merely bringing them to market in an attractive way.

  • Why would dell? they don't make electronics. They build computers byu assembling other peoples electronics.

    I have submitted stories about real ground breaking technologies from:
    and many, many others. I stop submitting 2 years ago because I had not had a submission accepted since 99.

    Now, I don't mind the apple stories. It's not like this is a limited space newspaper.

    BTW that tech isn't as ground breaking as you seem to think. It's like there isn't much there in regard to new tech, so people are glomming onto and straws they can grasp to justify waiting hour to buy a product that they could walk in and buy in 2 weeks. Hell it might even be fixed by then.

  • So I played with one briefly yesterday. I thought, "oh, this is nice, it's about the same speed as my 3gs...this screen doesn't LOOK a whole lot better." Then I realized I really needed to clean my glasses. With my glasses off, and the screen 6 inches from my nose, it looks AWESOME. From any distance away, through my myopic eyes, dirty glasses, and the pollutants in the air, it's much better than it needs to be.

  • Re:slashdotted (Score:1, Informative)

    by Anonymous Coward on Friday June 25, 2010 @10:41AM (#32690646)

    The images aren't here, but here's a Google cache of the site, so you can at least read the article. I'm working with Bryan to get the box statically caching and back up, be patient. :)

  • by atrain728 ( 1835698 ) on Friday June 25, 2010 @10:41AM (#32690656)

    It said "2 meters".

    What is this "meter" you speak of .. I only know of the metre. Unless you were using some new fangled musical notation that I don't know of. []

    The metre (or meter), symbol m, is the base unit of length in the International System of Units (SI). []

    In British usage, some words of French, Latin, or Greek origin end with a consonant followed by -re, with the -re unstressed and pronounced /r/. Most of these words have the ending -er in the United States. The difference is most common for words ending -bre or -tre: British spellings centre, goitre, kilometre, litre, lustre, mitre, nitre, reconnoitre, saltpetre, spectre, theatre, and titre all have -er in American spelling, as do calibre, fibre, sabre, and sombre.

    Happy to clear that up for you.

  • by Anonymous Coward on Friday June 25, 2010 @10:49AM (#32690770)

    .. which is why "laser quality" is 300 dpi. We knew fax looks like shit because it's 200 dpi. It's why chemical photo prints are almost 300 dpi. Why print artwork is done at 300 dpi. The "300 dpi pleases the retina" thing is like 25 or more years old. 300 is the most important point on the resolution scale.

    As someone that used to work for an agency:

    300 DPI represents an reasonable economics tradeoff between accuracy and cost. You can have 1270 DPI (professional photo typesetters run that high), but how much do you want to pay for it? Those machines are $500k+. And yes, your eye can tell the difference between 300 DPI and 1270 DPI.

    Artwork developed by print agencies is done at 300 DPI and no higher because of limitations on file transfer size and professional printer RIP speed. Trust me, if an 8" x 10" photo could be squashed into 3-5MB and rasterized in a short amount of time at 1270 DPI, there would be printing equipment and printers offering that as a service overnight.

    Finally, standard 35mm film is around 10,000 DPI, dude.

  • by clone53421 ( 1310749 ) on Friday June 25, 2010 @10:53AM (#32690826) Journal

    78 um and 102 um. They used the Greek letter Mu, which Slashdot helpfully strips out.

  • by Anonymous Coward on Friday June 25, 2010 @11:04AM (#32690986)

    Just smear vaseline on the lens like cameramen did for women in 1930s/1940s movies, to make them look better.


  • by Sockatume ( 732728 ) on Friday June 25, 2010 @11:05AM (#32691002)

    The face camera isn't high-res. I think it's something like 640x480, standard sort of webcam view. So, uh, I guess I'm saying you shouldn't blame the phone here, you'd look like that on regular old video.

    (I tried to come up with a more polite way of putting that. Sorry!)

  • by Graff ( 532189 ) on Friday June 25, 2010 @11:28AM (#32691282)

    There's a huge difference between 300dpi in printing where your C,Y,M or K is either on or off, and 300dpi in systems where the C,Y and M or your R, G and B come in 256+ levels.

    That's not actually how halftoning [] works in print. You have various levels of ink coverage at each "pixel" location, what they do is vary the dot size from large enough to cover all of the paper at that location to no ink at all. You easily achieve 256 levels of intensity at each location when halftoning.

  • There are few contenders between Droid's 25 and iPhone's 325: []
    There are entries about and over 300 dpi.

  • by omnichad ( 1198475 ) on Friday June 25, 2010 @11:34AM (#32691356) Homepage

    The pixels ARE square - at least in the display sense. At 78um x 102um it comes close to a 4:3 display turned on its side. Now, 4:3 is what computer graphics have been using since CGA at least.

    However, that measurement (FTA) doesn't take into account the pixel pitch. There's a gap of approximately 33um (eyeballing this) between each pixel horizontally, where there's maybe a 5um gap vertically. That makes each pixel take up approximately 111um x 107um which you might as well consider perfectly square, since I am not measuring this.

  • by Pandrake ( 1513617 ) on Friday June 25, 2010 @11:47AM (#32691492) Homepage

    Hmmm, I also used to and *currently* work in the print industry. I don't have time to explain the relationship between DPI and LPI when it comes to reproducing halftones in print, plus the effects of the only three different kinds of printing technology (ink based, toner based, and ink jet based) *and* the effect that has on different kinds of paper stock.

    Essentially, the "sweet spot" for color halftones is 340 dpi when you have the option of a very high linescreen with very low dot gain. There's sort of a formula for dpi, which is 2.5 X the linescreen (LPI); but it cannot be followed exactly if you're worried about tracking and other quality issues. 300 dpi is a nice round number that makes for "easy" to email files as well as a good standard for target resampling in PDFs and the like. Anything higher than 300/340 dpi and you'll have to have an LPI so high in order to prevent banding you can't keep the dots on the page (again, depending on which printing technolgoy you use). If you're LPI is too low it looks like the old newspaper photos, and also runs the risk of rosetta morie patterns being visible if the angle of the 4 inks (if using only 4 inks) isn't adjusted properly.

    For toner and ink jet based printing, it's easier to get away with as little as 150 dpi for color photos since the LPI is moot at a certain point where dot gain and near continuous tone transfer of pigment is possible.

    For greyscale halftones, you usually want 600 dpi for SWOP printing; again, it's not forumulaic, since that high of resolution may or may not result in banding depending on the LPI and the distance-to-amount of gradiation.

    For black and white, such as text, you want the dpi to be as high as your output dpi - so you don't see any of the jagged edges between the points of black and the white of the paper underneath.

    Finally, standard 35mm film is around 3,200 dpi depending on the emulsion chemistry.

  • by Bigjeff5 ( 1143585 ) on Friday June 25, 2010 @12:13PM (#32691840)

    It's impressive, but it isn't super impressive. The Toshiba Portege G900 already had a screen at 313ppi, and Sony Erickson X1 is 312ppi.

    What makes it newsworthy, is Jobs said (I paraphrase) "It's as good as the human eye can perceive." That's why he named it the "retina display".

    A scientist with a Ph.D. came along and called bullshit, saying that the human retina can perceive pixels much higher.

    The Ph.D. in this article respectfully disagreed, and said the previous scientist:

    A.) Used the wrong figure for retinal resolution when he made his calculations (0.5 arcminutes instead of the 0.78 arcminutes established by a recent, authoritative study) and

    B.) Failed to factor in losses in the optics of the human eye regarding how much light will actually hit the retina.

    With A fixed and B factored in, the scientist concludes that the practical limit of the human retina (what it can distinguished given the amount of light that hits it) is 286ppi when held at 1 foot away from the eye (the ideal distance for viewing detail). The iPhone is well above this, at 326ppi, which means Jobs was right, and the name is apt.

    It's worth noting that there are quite a few phones that beat the 286ppi limitation, but the iPhone has the highest.

    Basically it looks like we don't need any higher resolution than what the iPhone and others have achieved, anything more would be pointless.

    That, to me, is very impressive.

  • Re:Too literal (Score:2, Informative)

    by Anonymous Coward on Friday June 25, 2010 @12:36PM (#32692202)

    "Now, I don't know if this is just me, but I can still pick up sound higher than 32KHz"

    Nah...I'm pretty sure it is just you. Because that is twice what the average person over 18 can hear. And 12Khz above what 'perfect' hearing is around birth.

    Pretty much, it means you are full of shit, a nerd on the internet that likes to exaggerate, or you just don't know what you are talking about. Works out to be about the same.

    There might be things that are biting away at the 32Khz range, but you aren't hearing may be hearing distortion much lower levels. An accurate sound meter would be able to isolate the frequencies to tell you what you are hearing.

    I've worked in audio for a while and on occasion, I'd have artists or nerds claim that our specs weren't high enough. They'd pull out numbers pretty much like you did...and we had equipment that was calibrated to much higher than that (mainly to avoid the brickwall much lower...not because we needed anything much past 20khz)...considering my four graduate years were spent human research, I'd occasionally set up an experiment to prove a point.

    Funny thing is (and with my background, actually, it isn't)...even when their claims were dispelled, we'd still be told we were wrong and hear the artists repeat these claims in magazines or make boasts about competitors products that didn't tell them they were an idiot. We sold more because we were truthful than we would have with a celebrity endorsement...

    tl;dr: You're full of shit.

  • Re:Too literal (Score:5, Informative)

    by khellendros1984 ( 792761 ) on Friday June 25, 2010 @02:52PM (#32694476) Journal
    No one hears frequencies that high. People that claim to do so are generally hearing much lower frequency harmonics of the sound that they claim to hear. The commonly quoted maximum human-audible frequency is around 24kHz. Secondly, 3840 X 2400 is the largest display resolution I could find any mention of (aside from multi-display arrangements). You could find a display with over 3x the pixel count of a 1080p display, but not with 3x the resoution. That's a 22-inch display by Toshiba, which means it's about 2/3 the dpi of the iPhone screen. Essentially, you shouldn't be able to see the pixels from over 2 feet away, 3 feet if you have excellent eyesight.

    So....can I have your autograph, Kal-El?
  • Re:Too literal (Score:3, Informative)

    by Khyber ( 864651 ) <> on Friday June 25, 2010 @05:28PM (#32696696) Homepage Journal

    "You're working with screens with resolutions on the order of nanometers?"

    Why, yes. Learn about our technology sometime, and quit focusing upon regular computer stuff. We're using QD phosphors on our screens. Yay 90+% gamut.

  • Re:Really? (Score:3, Informative)

    by tknd ( 979052 ) on Friday June 25, 2010 @07:47PM (#32698354)

    First of all, marketing is not just advertising. There is a lot to marketing that is not advertising like focus groups, surveys, strategies, and measurement.

    And although you claim that Microsoft spent more as a percentage of revenue on advertising, they also spent way more than Apple on R&D:

    all numbers in millions

    AAPL (52 weeks ending 2009-09-26)

    Total Revenue 42,905.00
    Research & Development 1,333.00
    R&D percentage of revenue: 3.1%

    MSFT (12 months ending 2009-06-30)

    Total Revenue 58,437.00
    Research & Development 9,010.00
    R&D percentage of revenue: 15.4%

    DELL (52 weeks ending 2010-01-29)

    Total Revenue 52,902.00
    Research & Development 617.00
    R&D percentage of revenue: 1.16%

    Financial information taken from

"I prefer the blunted cudgels of the followers of the Serpent God." -- Sean Doran the Younger