The two top dogs in the smartphone world right now are the Apple iPhone 5 and Samsung Galaxy S3. My doughty CNET colleagues have thoroughly reviewed and compared each one, but so far we hadn't actually sat down to measure the performance of their screens. Until now.
I subjected the screens of multiple GS3 andsamples to a series of measurements using the same state-of-the-art gear and software I use to evaluate and calibrate TV screens. The phone screens are quite a bit smaller, but the process is very similar.
While I found major differences between the two, and advantages and disadvantages to each, my tests indicated that in general the iPhone 5's screen is better than that of the Galaxy S3. I'll be more specific.
The iPhone can get much brighter. This leads to punchier whites and better handling of bright ambient-light situations, particularly outdoors or under bright overhead lights. The Galaxy S3 has much deeper black levels. The AMOLED screen of the Galaxy S3 can produce a much deeper, more realistic shade of black than the IPS (in-plane switching) LCD of the iPhone 5. This advantage shows up best under dim ambient-light conditions, such as outdoors at night or in a darker room. Color on the iPhone is much more accurate. The iPhone hews significantly closer to the standard gamut used to produce most of the photos, videos, and other consumer content you'll view onscreen. That means it produces highly accurate colors. The Galaxy S3, on the other hand, has a gamut far beyond the standard, which oversaturates colors.
The table below sums up the average measurements I took from three samples of each phone, with the brightness sliders set to half and then again to full. Luminance (light output) levels are in nits and color temperature in degrees Kelvin. Check out the Methodology section below for my reasoning, as well as the nuts and bolts behind the results.
|Test||Apple iPhone 5||Samsung Galaxy S3|
|Color temperature (average)||7,219||7,706|
|Color temperature (average)||7,444||7,694|
Brightness, black level, and contrast
The Apple iPhone 5 is capable of nearly four times as much light output as the Samsung Galaxy S3. Viewed side by side, it's obvious the iPhone's screen can get much brighter, and that difference is especially important in situations where a lot of ambient light is striking the screen, such as out of doors and in brighter indoor rooms with strong overhead lighting. The iPhone's brighter screen also makes it easier to read on than the screen of the Galaxy in bright lighting.
On the other hand, with both of their screens turned up all the way, the Galaxy can get nearly 50 times darker than the iPhone 5. In a completely dark room, like the one in which I measured the phones, a black (0 percent stimulus) test pattern filling the screen of the Galaxy S3 was barely visible; the same pattern on the iPhone was much brighter. The difference was less obvious, although still visible, under normal lighting.
In its simplest form, contrast ratio is the difference between the brightest image a display can create and the darkest. According to my measurements in a completely dark room, the Galaxy S3 has a contrast ratio of more than 11,000:1, compared with the iPhone 5's ratio of about 1,100:1. That's roughly 10 times as much contrast.
To say that these screens are as different as night and day isn't too much of an exaggeration. In a dark environment, the better contrast of the Samsung's screen will cause its image to look punchier, especially with high-contrast material like dark scenes in a movie or a nighttime photo. Under normal room lighting or brighter, particularly a daytime sky, that advantage will dwindle and the brighter screen of the iPhone will shine through, overcoming ambient light to provide a more legible, clearer image for all types of material.
Which phone looks better overall in terms of contrast depends on ambient lighting, what material you're looking at, and the brightness you have the screen set to (at half brightness, for example, the contrast ratio of the Galaxy S3 drops roughly in half, while the iPhone's remains about the same). Given the fact that phone screens are often viewed in variable lighting or outdoors, the light output advantage of the iPhone is more important in my opinion than the contrast/black-level advantage of the Galaxy S3.
The advantage of the iPhone in this area is much more clear-cut. Its color comes extremely close to the sRGB/Rec. 709 standard gamut, meaning that the colors it puts on the screen closely match the colors of most signals. The gamut of the Samsung, since it's a good deal larger than the standard gamut, causes colors to look more saturated than they should. Green, in particular, is much, well, greener than it should be. The iPhone produces much more realistic, natural color.
The color temperatures of the two phones are more similar, within an average of 250 degrees Kelvin of one another, at full screen brightness. The iPhone is still better, however, especially in terms of RGB balance -- the Samsung is minus-red and plus-blue, and though the iPhone is too, it's not as bad.
Photos viewed on each screen bore out the test results. Colors from skin tones to tree leaves to flowers to neon signs looked punchier and more saturated on the Samsung's screen than on the iPhone's, where shot after shot looked more natural.
Subjectively, some people might find the Galaxy S3's improper gamut preferable to the iPhone's proper one. Put another way, if you prefer oversaturated colors (think Dynamic or Vivid mode on a TV) you'll like the Galaxy S3 better, and if you want colors that more accurately portray the source, the iPhone 5 is the clear choice.
Into the weeds: Methodology
I tested three different samples of both phones. I measured each phone once at 50 percent brightness and once at 100 percent brightness, keeping the ambient light sensor turned off.
Update 10/16/2012: To match the experience of a typical user as closely as possible, I did not change any of the other display-affecting settings from their defaults. Some of these, such as disabling "Auto adjust screen tone" on the Galaxy S3, may improve test results, and I may cover them in a future update, but they're largely beyond the scope of this test.
The phone screens were measured by a state-of-the-art spectroradiometer, the Konica Minolta CS-2000, in a completely dark room. The phones were held using a small tripod at a uniform distance (16 inches) perpendicular to the lens, and I focused as close as possible to the exact center of the screen. I measured full-screen 1,280x720-pixel JPEG test patterns within the default photo viewing app (Photos on the iPhone and Gallery on the Galaxy S3). The patterns originated from SpectraCal, manufacturers of the CalMan 4 software I used to power this evaluation. Black, gray, and white luminance was measured from 0 percent black to 100 percent white in 11 increments of 10 percent each, while color gamut was measured at 75 percent luminance.
In the interests of keeping these tests as objective as possible, I measured only black, gray, and white levels and color. I didn't address resolution mainly because I don't consider it a major differentiating factor -- the iPhone's screen can get sharper, but the Samsung's is plenty sharp for most people. I also ignored viewing angle since it's less relevant to phones; a phone is usually held in your hand, so you can freely move it to the ideal perpendicular viewing angle.
I didn't test reflectance, which is the measure of how exactly the screen handles reflections and ambient light. It's an important factor, but I currently lack the expertise and facilities to evaluate it properly. If you're curious, Raymond Soneira at DisplayMate tests reflectance, among many other points of interest, in his excellent "Flagship smartphone display technology shoot-out."
I also didn't go into the same depth of color testing as another evaluator whose work I greatly respect, Chris Heinonen. His article at AnandTech, "The iPhone 5 display: Thoroughly analyzed," came to similar conclusions about the iPhone 5's color accuracy and contrast. Heinonen's and Soneira's work was helpful in developing my test, as was that of Geoff Morrison.
Update 10/17/2012: Better late than never, here's the raw data used for the chart above, in the form of charts and a Google spreadsheet averaging the measurements I took from three samples of each phone.
How much does image quality matter on a phone?
I typically review TVs, a product for which the image quality produced by the screen at a given size and price is (or should be, IMHO) the most important factor. Cell phones are a whole 'nuther ball of wax.
In my personal opinion, the quality of the images produced by the screens of today's phones is much less important than handset design, brand, carrier, software, and ecosystem. I actually plan to buy a Samsung Galaxy Note 2 myself, mainly because I like its size and capabilities, and I'm a longtime Android user who loves the low rates of T-Mobile. I expect its screen to perform similarly to a Galaxy S3's, and worse than that of the iPhone 5 for example, but that's not an important-enough factor to prevent me from buying one.
It's still worth talking about the quality of cell phone and tablet screens, however, especially because they can be so different. As we consume more data, videos, and photos on these small screens, how capable they are of producing accurate, realistic images becomes more and more important. And as with any product you'll be committing to using frequently for two years or more, it's nice to know what you're looking at.
Correction: Originally this article included measurements of and references to gamma that were incorrect for the Galaxy S3 due to its tendency to dynamically adjust the screen. They have been removed. Thanks to Raymond Soneira at DisplayMate for pointing out this issue.