Shootout: How Does A High-End Smartphone Camera Compare To A $3,400 Dslr?

Our recent review of the Thrustmaster HOTAS Warthog joystick-and-throttle combo was notable not only for the really cool, really expensive piece of gaming equipment it featured, but also for the much-more-expensive full-frame DSLR used to take the article’s pictures: a $3,400 Canon EOS 5D Mark III.

This is a fair amount of scratch to lay down for a camera, especially when the Internet is full of examples of pro photographers going the opposite direction, ditching bags of expensive gear in favor of smartphone cameras for most applications. The idea here is that the person, not the gear, takes the picture. And there is a (likely apocryphal) story that tells the tale of an encounter between famous novelist Ernest Hemingway and famous photographer Ansel Adams. In the story, Hemingway is purported to have praised Adams’ photographs, saying, “You take the most amazing pictures. What kind of camera do you use?”

Adams frowned and then replied, “You write the most amazing stories. What kind of typewriter do you use?”

Past a certain minimum point, skill trumps equipment when it comes to photography (and the same is true for almost any pursuit or profession). However, equipment does play a role—sticking two wildly different cameras next to each other and taking two pictures of the same scene will usually yield wildly different results. Those pro photogs whipping out iPhones instead of big DSLRs know a hell of a lot about how to make the most of the light they can get into a lens—and capturing light is what photography is all about—and they know how to maximize the strengths of their cameras under whatever shooting conditions they’re given.

But for the rest of us who aren’t pros, what do you get when you look at the output of a multi-thousand dollar camera with interchangeable lenses next to the picture from a good smartphone camera? How do the two stack up without being boosted to their maximum potential by a trained eye? What kind of images are produced when you try to shoot scenes with roughly equivalent camera settings on two wildly different devices?

It turns out that this is a really complex question to attempt to answer, but we’ve taken a crack at it, pitting our expensive 5D Mark III and fancy lenses next to the big new camera in Apple’s big new iPhone 6 Plus. The iPhone 6 Plus was picked over other smartphone cameras primarily because it’s new and shoots great pictures—it’s a great example of the best that smartphone cameras can be. We’ve done our best to snap some representative photography under several different lighting regimes, but there are so many variables to account for that it’s impossible to do a truly controlled test and comparison.

Rather than striving for tremendous accuracy and meticulously tweaking camera parameters, we sort of winged it, pointing and shooting with minimal settings adjustments and trusting the usually quite skilled automatic aids built in to both our smartphone camera and our DSLR. Because in real life, this is what the vast majority of casual photographers are going to do anyway.

The steep curve of learning

Learning how to take pictures is, like most skills, something that’s easy to start doing, but difficult to do well. It takes time and practice—practice to understand the different parameters of a camera and how changing each affects the resulting image. As with most skills, knowledge is acquired often in fits and starts, a stair-step pattern of puzzlement and revelation. And as with most skills, a person’s subjective assessment of their skill doesn’t tend to track very well with their actual skill.

The point labeled “Dunning-Kruger peak” is that surge of self-confidence one gets when one acquires some small amount of skill in a thing and, emboldened by that skill, decides that they are awesome at that thing. However, repeated use of the skill typically disabuses one of that illusion (unless it’s a skill like rocket car racing, in which case you’ll probably just die before you get good at it). Once you’re across the peak, reality settles in, and you realize that the things you don’t know massively outweigh the things you’ve learned, entering what I’ve labeled as the “Jon Snow Trough.”

Only with a lot of time and effort can you realistically self-assess your skill. If I had to place myself on that chart, it’d be right at the second trough—I’ve learned, again, that there’s a lot I don’t know. But I am not anywhere near the same league as Ansel Adams, and I still have a fascination with gear and a reliance on technology to push me over gaps in ability. The genesis of this entire piece came from wondering what I’d do if I had to rely entirely on smartphones for product photography—and, frankly, I’m not sure I could do it. I need the extra techno-oomph to overcome lack of skill.

So enough navel-gazing and dithering about: let’s meet our two test cameras.

The contenders: iPhone vs. EOS

In one corner, we have Apple’s latest and biggest smartphone, the iPhone 6 Plus. The new iPhone’s camera has a scratch-resistant sapphire outer lens, a fixed f/2.2 aperture, fancy optical image stabilization to help make pictures clearer, and a 1/3″ 8-megapixel sensor, which produces stills with a max resolution of 3264×2448 pixels. There are a number of other sites that have good breakdowns of some of the extra fancy technology Apple has crammed into the phone in order to make it take excellent pictures, but there’s a lot going on inside that little chunk of electronics. All told, a new iPhone 6 Plus will set you back at least $299 with a two-year commitment to a cellular carrier; off contract, the iPhone 6 Plus starts at $749 and rapidly goes north.

In the other corner, we have—well, we have a whole mess of Canon gear. To start with, there’s the EOS 5D Mark III, a “prosumer” model of DSLR—that’s “digital single-lens reflex,” meaning that the view in the viewfinder is coming in through the lens, instead of through a separate viewfinder hole. The 5D Mark III comes with a 22-megapixel “full frame” sensor, meaning the sensor that actually takes the picture is about the same physical size as a piece of 35mm film (the 5D Mark III’s sensor is 36mm by 24mm), which yields a picture size of 5760×3840 pixels.

Of course, a DSLR body doesn’t do anything on its own, aside from make clicky noises when you press the shutter button. We’ve also brought a few Canon lenses for covering a range of shooting situations—our everyday 24-105mm f/4 L lens, a 70-200 f/2.8 for the far away stuff, a 100mm f/2.8 macro for close-ups, and a 50mm f/1.8 “nifty fifty” for sharp portraits. And to keep everything well-lit, we’ve got a couple of flashes (or “speedlights,” as you call them if you’re a photographer, to differentiate them from big studio flashes).

Before I started actually shooting photographs on a DSLR for Ars, I was totally bewildered as to why anyone would need more than one lens. After all, point-n-shoot cameras and smartphone cameras don’t have interchangeable lenses. What the hell is all of this heavy, expensive glass for? Fear not—we’ll explain.

For now, it’s enough to know that the gear here was all bought new, some by Ars and some by me. The total price for everything in this picture is a bit over $8,000.

Photographers out there are nodding their heads; most other folks are probably scraping their jaws up off the floor. Rest assured, though, this is a good-but-not-awesome collection of DSLR gear that pretty well matches my current photography abilities.

Match settings and fire

Taking pictures with the same “settings” on an iPhone camera and a DSLR isn’t as straightforward as it might sound. For one, the iPhone’s optical image stabilization lets it use shutter speeds that are a lot slower than you can safely use with a DSLR in your hands, even one with in-lens mechanical image stabilization. (That is, we suppose, unless you happen to be a world-class neurosurgeon with the steady hands of a marble statue.)

For most of the comparisons below, I took three pictures—one with the iPhone, using an iOS app called Manual to see what the camera’s settings were without having to pull the image off and inspect it (the iOS default camera app hides its juicy settings away from the user, so using the third-party app was the easiest way to see what the camera was doing).

The next two pictures were done with the 5D Mark III; the first taken with settings as similar as possible, and the next with settings that resulted in more correct exposure. To keep things simple—and to minimize haulage—I kept a single lens on the DSLR: Canon’s jack-of-all-trades-master-of-none 24-105mm f/4 L with image stabilization. This lens has a maximum aperture of f/4, compared to the iPhone 6 Plus’ fixed aperture of f/2.2, which means that a given shutter speed on the iPhone 6 Plus will produce images about one full stop brighter (and, oh, how I wish I had the column space to go into the exposure triangle, but that’s a whole other article!). I also did some small amount of post-processing on both the iPhone images and the DSLR images—mostly cropping and straightening.

There are four different regimes under which I took sample photos: indoors in poor lighting, indoors in normal lighting, outdoors in direct sun, and finally, in the make-shift white box I use for product photography. For the last, I took out all the stops with the DSLR and used off-camera speedlights and a lot of post-processing—exactly as I would with any other product image on Ars. I also gave the iPhone image similar treatment, though of course off-camera flashes for the iPhone weren’t an option with the gear I had available.

Indoors, bad lighting

One thing we’ve seen over and over again in smartphone reviews is that smartphone cameras struggle with low light. However, all cameras struggle with low light—that’s why generally the first piece of gear you buy for a DSLR is a speedlight. More light means better images, period.

We’re starting off with some really, really bad lighting conditions—blinds drawn, indoor type of lighting conditions. Relatively dim to the human eye and nightmarishly impossible for cameras.

These are pictures of a glass checkerboard in my dining room. The iPhone picture is unrecognizably dim, the DSLR pictures, you’ll notice, are brighter—but much, much noisier. That’s because I pumped the ISO setting progressively higher on the DSLR for the second and third pictures, amplifying the sensor’s output. At a max ISO of 102400, you can tell what’s in the photo a lot better than you can with the iPhone, but the image is far too noisy to be useful.

There’s a bit more light in these sets of images—not a lot, but some. The iPhone is dark, but the DSLR’s first stab at ISO 25600 isn’t terrible. Pumping the ISO up makes things brighter and worse.

This column was in an indirect pool of light reflecting off the walls and floor of a hallway, and now we’re starting to see a noticeable difference in the detail being picked up. The iPhone’s optical image stabilization lets it hold its shutter open for a crazy-long 1/4 of a second—far longer than I can keep the DSLR still, even though the 24-105mm lens has its own built-in mechanical image stabilization (Canon’s IS system uses a motorized element in the lens to counteract movements, whereas the iPhone’s stabilization moves the sensor itself; the sensor is much tinier than a big chunk of lens glass, and so it can compensate for even tiny vibrations).

Note also the areas behind the column. The iPhone doesn’t have the lensing necessary to produce a nice blurred background, while the DSLR manages to get a nice bokeh effect.

Indoor, moderate light

Next, one of the Apollo posters I’ve got on my office wall. The thing to look at here is the sharpness of the poster’s black lines against the white background (most apparent when looking at the full-size images). The iPhone struggles with the details and produces a fuzzier image, while the DSLR snags a much sharper image—especially once I amp the ISO and drop the shutter speed a bit.

This was shot in full indirect sunlight—the room was comfortably bright, and the blinds were open to a sunny day. The iPhone does fine, but the indirect sun doesn’t push enough light onto its sensor for a noise-free image, even with a shutter speed 1/15th of a second. The DLSR, on the other hand, can use its huge sensor and much bigger aperture to slurp in the light.

Outdoors in the sun

Get outdoors in full sun where there is an abundance of light, and suddenly the iPhone transforms into an awesome picture-taking machine. The details on the flowers in the basket are crisp and beautiful. I had to do quite a bit of fiddling with the DSLR to dial in a good shot, and I ended up with one underexposed image and one overexposed image.

Ordinarily, I’d take these both back into Aperture or Lightroom and correct them—but therein lies a huge strength of a good smartphone camera. You can generally get it good enough in-camera without having to rely on post-processing, especially when you’re just faffing around or taking selfies or vacation photos. With a DSLR, getting it perfect in camera without having to pull out the post-processing tools isn’t always an option, especially if you need to approach professional quality with your images.

Notice that the bricks on the DSLR image look… a little bendy. This is called “barrel distortion,” and most lenses exhibit it to one degree or another, especially zoom lenses like the 24-105mm I’m using. It’s easily correctable in post-processing (applications like Adobe Lightroom can fix it automatically when pictures are imported), but I’ve left it uncorrected as an example of the kinds of things that you just don’t have to care about with a smartphone camera.

In addition to the barrel distortion, the corners of the DSLR image are a bit dark—this is called “vignetting.” Again, this is a trait that most lenses exhibit, some more than others, and it’s easily correctable. But it’s another thing you simply don’t have to bother with on smartphones. (However, you might sometimes want to actually add vignetting for a more stylized look, if you’re the Instagram type.)

Here I’ve shot another flower, but this time I’ve used the fact that I’ve got a zoom lens on the DSLR to foreshorten the image, letting me blow the background out into a cool artsy blur. The iPhone produces a perfectly pretty image, but you’d have to go to a lot more trouble to match the effect. On the other hand, the details on the blossoms themselves are every bit as sharp—in fact, on the DSLR image, I realized that I’d screwed up and used an aperture value that left my plane of focus too narrow. Look at how the middle of the cluster of blossoms is in focus, rather than the entire cluster. And then compare that to the iPhone pic, which has the whole thing in frame.

Lesson learned: if you’re going to be artsy, don’t let the tools get the better of you.

Pulling out all the stops

There are tradeoffs indoors and outdoors with smartphones and DSLRs—the fact that the small (smallish, anyway) iPhone can produce images that are almost as good as the bulky DLSR and lens combo means that there aren’t too many situations where a DSLR is required. Someone skilled at making the most of their available light can, in almost any circumstance, produce wonderful images with a smartphone (and I’m not saying I produced wonderful images here—merely that it’s easy to see how someone with more skill could).

However, now we come to the area where the smartphone, as hard as it tries, cannot compete: controlled studio photography. In this case, I wanted to take a glossy, high-resolution, gloriously bright product shot of something. So I pulled out all the stops: I built my usual whitebox out of poster board to bounce light around, and I set up a pair of off-camera speedlights to throw prodigious amounts of light around (I usually operate with three, but I’m temporarily down to two right now). The setup looked like this:

The tiny lens and sensor on the iPhone 6 Plus just can’t hang with the DSLR and the light provided by the flashes. Even with the iPhone’s own on-camera flash—which this close-in produces a terrible overblown effect thanks to the tyranny of the law of inverse squares—the images it produces just aren’t anywhere in the same league as the DSLR.

Comparing the iPhone’s best picture to the DSLR’s shows that the iPhone is hopelessly outclassed. I’ve got the DSLR’s aperture stopped way up to maximize my focal plane; in retrospect, I also should have swapped out the 24-105mm zoom lens for the “nifty fifty” 50mm prime in my kit for a far sharper image. Even without that, the DSLR steals the phone’s lunch.

So what have we learned here?

You’ll likely notice that even though in some pictures the iPhone and DSLR have similar settings, the pictures look wildly different. This is due to a whole confluence of factors—mainly because we’re dealing with hugely different lens and sensor sizes, and the actual physical value of camera settings like the f-stop are dependent on the dimensions of the lenses being used.

A true apples-to-apples comparison would be a lot more difficult to pull off than simply shooting a few sets of images under a few different lighting regimes, and it would likely yield a lot more objective data. However, what matters in a lot of these cases isn’t megapixels and histograms in a vacuum, but how the images look to your eyes. We’ve definitely got enough here to show that under many conditions, a smartphone that costs a few hundred dollars is mostly as good as a DSLR that costs eight-to-ten times as much.

We’ve also demonstrated—pretty amply, I believe—that the skill of a photographer matters more than the equipment. This is pretty obvious because even with that expensive DSLR, the images Imake aren’t anywhere near as good as the images that can be banged out by a more skilled photographer. In fact, one of the most extreme examples of just how far skill trumps equipment was put together by DigitalRev TV and features BAFTA-winning cinematographer Philip Bloom attempting to scrabble together a movie using a 240p-resolution Barbie Video Girl camera. The results are surprisingly good—far better than most folks could probably manage even with a much better camera.

So raise your iPhones up high and snap away—unless you need to spend a lot on a camera, you’re almost certainly fine. Step away from automatic shooting and learn the camera’s settings a bit, maybe, but don’t feel like you need to spend thousands to take better pictures. Instead, remember this photography aphorism: “Amateurs worry about gear; professionals worry about money; masters worry about light.”

Source: http://arstechnica.com/gadgets/2014/10/smartphone-camera-vs-dslr/3/