How good is the new iPhone's camera? We tested it against some older phones and a few compacts
It has been years since a camera phone has been run through the photographic gauntlet that is the Popular Photography test lab, but with the iPhone 4S getting so much attention for its new 8MP camera, we figured it was time. We put the smartphone through our three primary tests--color accuracy, noise, and resolution. Then we took a look at the lens distortion. It's worth noting that the native camera app doesn't offer any level of control over the exposure settings, so we had to leave it up to the camera to decide ISO, aperture, and shutter speed. Since we use bright light on our targets for consistency, the only results we could offer were at ISO 64.
The final results are mixed. Compared to 8MP camera phones from a few years ago (the Sony Ericsson K850, Nokia N82, and Samsung G800), the iPhone 4S yielded more resolving power. The iPhone 4S turned in 1730 lines, compared to the Sony Ericsson K850's 1370 (at ISO 100), the Nokia N82's 1410 (at ISO 100), and the Samsung G800's 1250 (at ISO 50). In our color accuracy test, the iPhone 4S scored an Extremely High rating with an average Delta E of 10.0. While that sounds nice, the truth is that almost all the cameras we test end up with an Excellent rating in this test. The Sony Ericsson K850 scored basically the same as the iPhone 4S with an average Delta E of 9.9. But, the Nokia N82 achieved an excellent rating with an average Delta E of 6.5 and the Samsung G800 also scored Excellent with an average Delta E of 7.1 (lower numbers are better).
Noise is where the iPhone 4S did the worst. Even at such a low sensitivity setting of ISO 64, the iPhone 4S only mustered our second lowest rating of Moderate with a noise score of 2.5. In this case, lower scores are better and our cutoff for an Unacceptable noise rating is 3.0. The Sony Ericsson K850 (2.4; Moderate, at ISO 100) and Samsung G800 (2.6; Moderate, at ISO 50) both turned in similar results to the iPhone 4S. The four year-old Nokia N82 was the winner here with a 1.8 for a Low rating.
But, how would the new iPhone's camera compare to a compact? We knew a dedicated camera would fare better than the iPhone, but ran Canon's PowerShot S95 through the test lab to find out exactly how big that difference is. Comparing the JPEG-based test results of the S95 to the iPhone underscores why a real compact camera still makes sense if you care about image quality. Under the same ideal circumstances as the iPhone's test results-- i.e. at the S95's lowest sensitivity setting of ISO 80-- the Canon resolved 1930 lines per picture height. That's 200 lines more than the iPhone, which is a noticeable difference. Of course, the 10MP S95 has 2 million pixels more than the 8MP iPhone 4S, but the S95's main advantage is noise control. The S95 scored a 1.4 for a Very Low rating at ISO 80. That's three rating bands better than the iPhone 4S rated at its lowest setting of ISO 64. Furthermore, the Canon had a noise score of 2.1 at ISO 800 where it still (barely) outresolved the iPhone 4S with 1750 lines. We shudder to think what results from the iPhone 4S would look like at ISO 800.
At higher ISOs, you can see the loss of detail that occurs due to noise. Photo: John Mahoney
Though both rated Extremely High in color accuracy, the S95 did so at the lower end of our scale for that rating with an average Delta E of 8.6, while the iPhone 4S barely eked into that rating band with a 10.
The one place that the iPhone 4S had any sort of advantage over the Canon S95 is in lens distortion. But, that's only when the Canon S95's 3.8X zoom lens was at its widest or at its midpoint. At those settings, the S95 had 0.3% and 0.33% barrel distortion respectively, which we rate as Visible, while the iPhone 4S showed 0.15% barrel for a Slight rating. Zoom the S95 all the way out to 3.8x, aka a 105mm equivalent in 35mm-photography terms, and distortion fell to a mere 0.08%, which we deem Imperceptible. All told, the Canon S95 offers a vastly better shooting experience, with a robust control system, and plenty of options to help you control what kind of image you'll get.
But, for what it is, the iPhone 4S captures decent images.
Photo: John Mahoney
Given that superzooms use the smallest sensors of any dedicated cameras, it's fitting to compare one of those to the iPhone 4S as well. The Sony HX100V, a 16.2MP camera with a 30x zoom lens and a 1/2.3-inch (7.77mm diagonal) sensor only managed to resolve 1770 lines per picture height in our test. Plus, its color accuracy also rated Extremely High with an average Delta E of 9.2. However, the Sony HX100V's noise result at ISO 100 was 0.9, for an Extremely Low rating-- the best rating you can get in this test. The HX100V maintained lower noise than the iPhone 4S ISO 64 result all the way up to ISO 1600, where the Sony scored Moderately Low with a 2.3.
Furthermore, the phone made some strange choices for aperture and shutter speed. All of the test images ended up with an aperture of f/2.4 and shutter speeds between 1/180 sec. and 1/450 sec. While we understand that it's probably programmed to assume that the phone is handheld while shooting, we wish the iPhone 4S would make use of its motion sensors and choose a slower shutter speed and smaller aperture to maximize depth of field if the camera is steady. There are so many accessories available to steady the iPhone, including more tabletop tripod accessories than we'd care to count, that it makes sense to us that the phone might be stabilized while shooting.
Photo: John Mahoney
Ultimately, we were impressed with the amount of resolving power Apple has gotten out of its latest iPhone, but it seems that that comes at the expense of noise. That's not a surprising finding considering the relatively small sensor. Furthermore, the ability to adjust our exposure settings, even just a little bit, is something we'd like to see Apple include down the road.
Of course, the battle of cell phone cameras vs. compact certainly doesn't stop at these numbers. When you consider sheer convenience and connectivity, the iPhone dominates. If you're more concerned with image quality and control -- as we assume most of you reading this are -- a dedicated camera still offers some noticeable advantages. But, in the end, it certainly wouldn't hurt to have both handy.