The Truth Behind the iPhone 4 Retina Display

Thanks to his ability to restructure the time-space continuum in accordance with his will, Steve Jobs normally gets away with pushing the credibility barrier a bit during product launches. But when he recently announced that the iPhone 4 sports a “Retina Display,” questions were raised. Were those tiny little pixels really so small that, when the iPhone 4 is vigilantly held a particular number of inches in front of your nose, the human eye cannot detect them? Calculators were whipped out, numbers were crunched, and experts batted about their sometimes-opposing views. We sorted through all the noise, interviewing key players (via email) to bring you the definitive answer. And that answer is: close, Steve, but no cigar.

But the really important question isn’t whether Steve’s math is perfectly correct. No, the more germane question is whether an ever-growing plethora of teeny-tiny pixels really matters in the greater scheme of things. Do more pixels equate to a better gadget experience for users? Are these little red-blue-green dots the holy grail of mobile displays--and if they are, can we trust companies to provide us with accurate specifications?

The math discussed in media stories and blog posts about the iPhone 4 was enough to give many of us brain cramps. Nonetheless, others were inspired to jump right in and wrestle with the numbers to extrapolate real-world usage data. Among them was Phil Plaits, a scientist who spent a few years calibrating a camera on board the Hubble Telescope and who now blogs for Discover Magazine. 

“My first reaction to the announcement and the ensuing coverage was interest and curiosity. I figured [Jobs] wouldn't lie outright, so what really is the limit of human vision as far as pixel size? The math is simple, if you know it, so I did the calculations, and found that while his claim wasn't perfect, I thought in context it was fine,” said Plaits, who helpfully provides charts and graphs illustrating the central concepts in his blog post about the iPhone 4’s display.

But as usual, the devil is in the details.

“It's easy to make assumptions about pixel size and distance, but then you have to take into account human vision--which is complicated--how the pixels are laid out, whether there is space between them, how they're refreshed, and so on,” said Plaits, who adds that he opted not to worry about slicing the data that finely. 

“I just wanted to see if, given some simple assumptions, you could make a smooth, continuous-looking display. The answer is yes, and the new iPhone display will be pretty close to that ideal.“

Some experts, however, say more accuracy is warranted. Dr. Raymond Soneira, president of DisplayMate Technologies Corporation, which produces video calibration, evaluation, and diagnostic products, said that exaggeration over display specifications has been “building for many years and has now become outlandishly unreliable, with many of the consumer specs being exaggerated by 1000 percent or more.”

Blame it on the snowball effect: Once one manufacturer exaggerates its exacting specs a tiny bit, all the competing companies do the same. The stakes are raised again and again, and pretty soon that little innocent snowball has morphed into an out-of-control abominable snowman.

“The manufacturers have actually painted themselves into a corner because the numbers they are quoting now are so ridiculously large that there is nowhere to go,” said Soneira. “The only realistic solution that I see to stop spec abuse is an organization (completely independent of the manufacturers) that develops a set of straightforward objective standards for measuring and advertising display specs for consumer monitors, laptops, HDTVs, smart phones, and other mobile displays.”

Soneira, whose statement about the iPhone 4 display sparked the furor over whether Apple was fibbing about the specs or not, envisions that manufacturers that meet those independent standards would be allowed to advertise their specs with a special controlled trademark, like the EnergyStar program. And consumers would learn to only trust specs with that trademark. 

“I proposed this back in 2003, but it went nowhere because too many manufacturers resisted the idea,” Soneira said. “It's time for this solution to be implemented--or just imposed. It's in everyone's interest except for the manufacturers that can only compete using fraud.”

Soneira added that his initial analysis and comments on the Retina Display “were widely distorted and transformed into an attack on Apple and Steve Jobs--they were not. I simply did a quantitative analysis of what was said in the context of my campaign to eliminate (or more realistically, reduce) exaggeration in display specs. Apple's claim falls under glorious wording rather than numerical-spec abuse—and even with quantitative analysis, it's minor compared to what other manufacturers are saying. I sent Steve Jobs an email explaining that and got a reply from him.”

Over at the NASA Ames Research Center, Andrew B. Watson, Senior Scientist for Vision Research said in an email interview that he was pleased that Apple had dubbed the new screen the "retina" display and emphasized that the company was trying to match the display to the requirements of human vision. “That is our mantra, and as a vision scientist, I felt flattered. However, I was slightly skeptical about the claims (Jobs) made.”

Watson, whose work involves developing display metrics that are meaningful to the human eye, looks forward to the day when visual resolutions are quoted as a display spec. He explains that in order to know the true effective resolution of a display, you need to know not only the resolution in pixels/cm, but also the viewing distance. Together these give you the true "visual resolution" of the device in pixels/degree.

Resolution is one of the key contributors to the users' quality experience when using a device display, according to Watson, the benefits provided by high resolution depends somewhat on the content you’re viewing--for example, you’d notice high resolution more in images with sharp edges such as text.

“And with resolution, there is a point of diminishing returns. From a theoretical point of view, the absolute limit is around 120 pixels/degree. But it becomes progressively harder to tell the difference above 60 pixels/degree, a visual resolution is reached at a viewing distance of about 27 cm (around 10.6 inches), which is a reasonable natural distance for an iPhone,” said Watson.

Besides pixel density, the most critical factors in creating a great display for a small gadget like the iPhone include color gamut (the total space of colors that the display can produce), field of view, and contrast ratio (the ratio of brightest to darkest pixels). That latter, Watson said, is an area of raging controversy. 

“Nowadays, contrast ratio mainly has to do with how black you can make a pixel, which is ironic. Another is brightness, up to a point. And another is how well the screen suppresses reflections, although this may be less important for handhelds, where you can rotate the screen to avoid reflections.”

Watson added that not all quality limits are imposed by small-sized display form factors. We have, he points out, a digital infrastructure that is hard to change. For example, it’s conventional to code each pixel signal with 8 bits each for red, green, and blue, but that is simply too small a space to achieve high dynamic range displays. 

“To some extent, current displays are already reaching the limits imposed by the eye and the infrastructure. Future developments will probably be in making displays cheaper, lighter, lower power, or extending them into new domains such as creating flexible, 3D, immersive, wearable, integral, or holographic displays,” said Watson.