Pixel XL vs Canon 50d
a useless comparison
If you’re shopping for a new phone and you’re hoping this article will help you decide, you’re sure to be disappointed. Just to get it out of the way, I’ll sum up the conclusions of my phone camera research in a single sentence: The Apple iPhone 7 camera is the best in low light (such as inside your house in the evening) and for telephoto (zooming in), the Samsung S7 camera has the fastest auto focus (making it good for pics of kids and pets in bright daylight), and the Pixel XL has the best looking photos straight off the camera from a wide variety of scenarios, no processing required. For my purposes, any of these would have been great cameras, and the selling factor for me was that the Pixel XL offers free unlimited full resolution storage and backup of every photo and video taken with the Pixel through Google Photos (which also integrates nicely with your DSLR photos via the desktop uploader). Really Google, you’re giving me free storage of all my photos and 4k videos forever? Forever-ever? Challenge accepted.
Yesterday was my first day with the Pixel XL, Google’s first smartphone camera (yeah it does other things too, but so do all of the other phones, so all I really care about is the camera). Planning a day of walking around the city doing casual street photography, I decided to test out the Pixel XL’s 12 MP rear camera with the other 12 MP camera I had laying around — the 8 year old, heavily used and abused Canon 50d, lovingly named the Tank, as seen above.
The Pixel XL looks exactly like the iPhone that you’re probably reading this on, so I won’t bother to post a photo of it. Just look at your phone for a second. Pretty boring, right? The 50d looks way cooler, and is better as a blunt force weapon.
The lens on the 50d is a Canon 40 mm f2.8 prime lens (meaning I can’t zoom in or out), which on the camera’s crop sensor behaves more like a 50 mm. The Pixel XL offers an unknown lens (as reported by Lightroom), but it registers as 4.67 mm focal length with a locked aperture of 2.0. That is ridiculously wide, such that any DSLR lens that could approach that would have to add a ton of fish-eye distortion. So in order to get some images that I could compare between the cameras, I had to crop the Pixel XL photos down a LOT to get a similar field of view to that of the 50d.
But wait, doesn’t that completely negate this camera quality comparison because I’m comparing a full 12 MP image on one camera to a cropped 12 MP image on the other? Absolutely. Didn’t you read this article’s subtitle? Useless.
Functionally, I don’t care. A 12 MP source photo cropped down to something like 5 MP should still produce a good quality image, one good enough to print or publish online. And if I’m using the super-wide angle lens that the Pixel XL offers for serious shoots, I need to trust in my ability to heavily crop the images later, so this is a test of my workflow as much as it is a test of the cameras’ respective quality.
I also chose to leave both cameras on full auto settings. Normally, for a street shooting scenario I would use the DSLR in AP or manual mode, locked onto settings such as ISO 100, 1/250 or 1/500 shutter speed, and f8.0 aperture. However, I figured I should let the computer make the decisions on both cameras since the Pixel XL doesn’t give you that amount of fine-tuned control (at least not in the default camera app). It is worthy to note that the Pixel XL by default applies an HDR effect to most photos, which means that it boosts saturation and contrast and does some automatic shadow recovery, among other things.
OK, now on to the photos. For each of these, the DSLR photos are on the left and the cropped Pixel XL photo is on the right. All of these are straight off the camera and completely unprocessed, save for the cropping or where noted otherwise.
And here’s the original uncropped Pixel XL image:
The first thing I notice is that the 50d chose a very different white balance for the shot, giving it a heavy blue tint. The Pixel XL is more true to life in color reproduction, though it does slightly over saturate, which is fine with me. The other thing I notice is the heavy lens and solar flare from the Pixel XL. Perhaps I need a case with a lens hood for it! While lens/solar flare can look great and trendy, it generally leads to desaturated photos and a loss of dynamic range. Since the Pixel XL is compensating with its HDR effect, it ultimately ends up looking fine for my tastes.
The obvious differences here are saturation. The 50d is less saturated (the sky is pale blue), and has less contrast and is generally a brighter image. The Pixel XL has deeply saturated colors, more contrast, and darker shadows. The full uncropped shot shows that pervasive lens flare again.
I chose this somewhat sad looking cactus as a particular challenge for each camera’s auto focus. “God dammit,” the cameras seem to say, “do you want me to focus on the tip of the cactus spine? or the base? or the fleshy green part?” I ignored the camera’s whinging and snapped away. Both cameras returned somewhat muddy and unclear images, as both chose wide apertures, resulting in a narrow depth of field that highlights parts of the cactus while rendering other parts blurry. Here’s the interesting thing: the 50d chose f5.0, which I think shows off a nice bit of the cactus and makes the image more interesting. The Pixel XL was stuck with its only option of f2.0, which should have highlighted an even narrower slice of the cactus and blurred the rest, but instead the depth of field on the Pixel XL’s photo seems to be deeper. This is the opposite of what I would expect to happen. I’m becoming increasingly less trusting of the Pixel XL’s reported EXIF data, as I think there’s a lot of computing going on in the background that drastically effect the end results, at least as far as aperture is concerned.
The color tone and saturation of these two photos are also drastically different, and for this straight-off-the-camera comparison the Pixel XL clearly produces more pleasing colors. However, since I’m shooting raw on the 50d, this comes at no surprise. Raw produces a more flat image that is meant to be post-processed before it is published (I also have the camera set to neutral picture style, further dulling the image and necessitating post-processing). If I really liked this photos (I don’t) and I wanted to publish it in my portfolio (I don’t), I would give it a few minutes of Lightroom processing and end up with something like this:
This version offers much more true to life color reproduction and an overall better image, since you are encouraged to focus on the bit of cactus in the foreground and subtly ignore the background, instead of trying to focus on everything as the Pixel XL did.
This is probably the point where you start muttering to yourself about the gall I must have to compare JPGs to RAW. Again, I’ll refer you to the article’s subtitle. The Pixel XL’s default camera app doesn’t offer a TIFF/RAW option, and to shoot the 50d is JPG mode would be handicapping it too much. I don’t need to explain myself to you, but I will anyway because it makes me feel better.
Moving on…
This is the first image where I genuinely prefer the Pixel XL’s results to the 50d. The Pixel XL’s auto HDR effect definitely kicked in pretty heavily here, and I’m sure many people prefer that it be dialed back a bit. I like it, and use ‘bold’ or surrealistic HDR in my photography often, though I’m typically combining multiple exposures, not just adding an HDR filter. However, I would believe that the Pixel XL is actually taking multiple exposures during the split-second shot (maybe it is real HDR, not just an HDR filter? anyone know?), and if so then I say it is fair play. By the way, the Pixel XL’s EXIF data for this shot is somewhat comical:
Is it just making these numbers up? Probably not. My guess is that they’ve started treating ISO as a full sliding scale, ignoring the traditional notches at 100, 200, 400, etc. This makes sense to me, as there’s no reason for electronic sensors to conform to film standards. 4.67 mm? Good lord, that’s ridiculously wide. f2.0, which again doesn’t really make sense based on the photo’s apparent depth of field. Finally a shutter speed of 1/2100 is admirably fast, but still within the range of the 50d’s max shutter speed of 1/8000.
Since the HDR effect is so prominent in this photo, I decided to give the 50d’s image a chance to redeem itself with a Lightroom HDR filter:
Meh. Its closer to the Pixel XL’s final image, but the HDR effect is a bit too obvious to me and the image is too heavily processed for my taste (the dark halos around each bar in the fence is distracting to me, and generally ruins the image).
Here’s that lovely golden hour light that we photographers love so much. The 50d chose f5.6, which offers a nice image of the pup while blurring the building in the background. The Pixel XL creates an image where everything is somewhat in focus (even the inferior dog that ran into the shot), but nothing is particularly sharp. This leads your eye to wander the photo looking for the thing you are supposed to care about, only to realize that there is no clear subject and you should just give up and reconsider your life choices. From the full image we can see that the Pixel XL was probably trying to focus on the grass in the foreground, which still doesn’t make for a terribly interesting shot.
With less than a minute of processing I can bring those colors in nice and purdy on the 50d’s image:
Day and evening light shots should be well within the wheelhouse for both cameras, but as the sun set I wanted to push both cameras to perform where I expected them to fall flat: night photography. Specifically night photography without a tripod or artificial lighting. So, street photography at night? Sure.
While I don’t have a comparison image for this one due to the lack of a wide enough angle lens for the 50d, here’s an example image from the Pixel XL that I find truly impressive:
The Pixel XL’s colors continue to please, but getting a good sunset photo from a phone should be one of the most basic of functions these days. What I find impressive is the quality of the dog and bike in the foreground. Sure its a bit dark since the camera is exposing for the light on the horizon, so let’s brighten it up a bit in Lightroom:
This is a challenging environment for a camera, and I think the Pixel XL performed well above my expectations here. The shadow recovery from the foreground is exceptional. OK, so perhaps the Pixel XL performs better during the blue hour than the golden hour? Interesting.
My initial attempts at night photography on full auto mode aren’t worth sharing, so at this point I stopped trying to compare the cameras head to head and started trying my best to use the camera to get photos I liked. I wanted the reeds to be in focus with the city lights blurred in the background. You can see with the 50d, the f2.8 aperture gave me the effect I wanted, but it was too challenging to get the reed in focus. If I really wanted to do this photo right, I would need a tripod and probably a flash. But the bokeh! Yeah, pretty pretty bokeh.
and here’s where we get to briefly touch on the difference between bokeh and blur. bokeh is a property of your camera’s lens hardware, and it cannot be recreated with software or effects in quite the same way. blur is simply out of focus light, which can be recreated with software and effects in a generally acceptable way. ultimately, all that really matters is that you like the images you are getting, so use whatever camera and software makes you happy.
The Pixel XL’s photo is so boring, I’m sorry that you had to see it. Again the autofocus isn’t sure what to work with, so it ends up focusing on the green stalk of the reed instead of the feathery bit (which makes sense, since the autofocus is looking for hard edges with high contrast). The cityscape in the background is recognizable but not really in focus.
I wouldn’t publish either of these photos in my portfolio, but the 50d’s photo inspires me to head out again and keep experimenting with the site and lighting, whereas the Pixel XL’s photo inspires me to stop trying to shoot night photos.
In this case the Pixel XL performed a bit better, but still not as well as the 50d. When viewed on a small screen both of these photos look generally fine, but when you zoom in or view them on a larger screen you start to see the quality degradation:
The 50d gets grainy as we expect at high ISO, whereas the Pixel XL gets splotchy with a lot of color bleed. Grainy I can work with, and I often do for effect, but splotchy isn’t a tool I particularly want in my toolbox.
The 50d’s image is flat and out of focus, but quite a bit brighter at ISO 3200. The Pixel XL’s image (in which I took a few steps closer to the stop sign) offers more depth and contrast and is slightly more in focus, at least around the letters on the stop sign, but is surprisingly dark with its reported ISO of 4519 (!?).
The full image from the Pixel XL shows of course a much wider field of view, but also demonstrates that the quality and sharpness fall off pretty quickly as you get away from the center of the image.
In conclusion
Hooboy, you still with me? You must be thoroughly bored by now. That’s the end of the head-to-head comparison. So which camera won? For me, the 50d is still going to be my preference for a street shoot like this, as it gives me the flexibility and options to create the shot I want, not just the shot that a computer thinks I should take. That said, I’m glad to know that the camera in my pocket can sub in for certain shots that the 50d will struggle with, and that the shots I take with the Pixel XL will require much less (if any) post-processing. If you were actually shopping for these as cameras, the 50d offers a much better quality to dollar ratio. But it won’t let you make calls or send texts or read stupid articles like this one.
I did want to share a few things from each camera that you can’t really compare. As with all tools, each one has its strengths and weaknesses, and knowing how and when to use each one is the best way to get the results you want.
End of article.
This is my first time creating an article like this, so I would appreciate your feedback! Too many images? Not enough images? Too much jargon? Too much snark? Did you read this hoping it would benefit your life in some way and find that it was a total waste of time and spun you into an existential crisis about how you spend your precious time on the subway scrolling through never-ending mediocre content on your phone until your thumb is tired and your brain is mush and you have no choice but to walk away from your life, shun electricity, and disappear into the woods for 7 years while living off the debris of campers and holiday cabins and fashioning shoes from pine cones and then realizing that your pine cone shoes are actually pretty cool and you should probably return to society to launch a startup around your eco-friendly pine cone shoes and you have to find an incubator and do some crowd funding and market research and all of that sounds terrible so you just go back to scrolling mediocre content on your phone? Great! I want to hear all about you and your emotions (in 500 characters or less, I’m a busy man).
