Camera manufacturers have been racing to offer ever larger sensor resolutions, but Ben Davis wonders if we have already reached “peak megapixel”
An image from Ben’s travels, taken on a high-resolution DSLR
In 2012, Nikon announced the D800 – a 36.3MP full-frame DSLR – and I simply had to have one. At the time I was shooting with a D700, a camera I loved and had no pressing need to replace, but the idea of tripling the resolution of my images was too enticing to resist.
Leaping from 12MP to 36MP was like switching from standard definition TV to HDTV. Seeing all the extra detail was extremely satisfying, and my D700 began collecting dust.
Fast-forward five years to the present, and Nikon’s largest-resolution offering has remained static at 36.3MP. Five years a long time in the technology world, and this perhaps suggests that sensors are close to reaching their maximum – or, rather, optimum – pixel count. And if not, then the “pixel wars” are certainly slowing down.
Prior to this stalling, the resolution of camera sensors had been approximately doubling every five years. Way back in 1999, Nikon launched their first professional DSLR: the D1. It came with a 2.7MP sensor. In 2004, there was the D70 which had a 6.1MP chip, and in 2010 the D7000 boasted a 16.2MP resolution. If this pace of increasing pixel count continues, can we expect a 70MP sensor in the imminent future?
While Nikon have yet to bump their maximum sensor size, two years ago Canon took the top spot in the DSLR resolution league table with the 5DS and 5DS R, both offering a whopping 50.6MP sensor. I’m sure other manufacturers are seeking to overtake Canon as the current leaders in the DSLR market; after all, the megapixel count is one of the key selling points of a new model.
Photographers who suffer from pixel-envy may look to the sensor spec first when eyeing up their next camera, but are there benefits beyond bragging rights? There were two key things for me when switching to a much higher resolution sensor: the actual pleasure I took from capturing enhanced detail, and in turn the much tighter crops I was able to make as a result. It’s not hyperbole when I say it revolutionised my shooting experience.
Of course there’ll be pedants and purists who say you shouldn’t need to crop because you should get it right in camera, or that by cropping you’re wasting pixels on your sensor. But in the real world you don’t always have a lens that is long enough for your subject, or enough time to ensure the framing isn’t a bit loose. Capturing higher-resolution files allowed me to sidestep these problems, and helped me create the images I was envisaging.
An extreme crop of the image at the top of this post, made possible by the generous resolution of the original file
So would I like even more megapixels on my sensor? You betcha! But I don’t need them. Almost no one does. And increasing the pixel count is not without its downsides – it’s true, for instance, that cramming more pixels onto a sensor will actually produce more noise in images. This is because to increase the megapixel count, the photosites themselves have to be made smaller in order for them to all fit on the chip. This amplifies the signal-to-noise ratio in the camera, making for grainier pictures, and the image quality suffers accordingly. This is the same reason why larger sensors – like those in full-frame cameras – offer better image quality than their smaller counterparts.
Higher resolution images also means bigger file sizes, which slow down your camera, fill up your memory card, and ultimately mean you end up buying more and more external hard drives to store them all on. There’s currently 12TB of additional storage hanging out of the back of my computer. Massive camera files also make Lightroom and Photoshop perform more sluggishly.
There are other drawbacks of having more megapixels, too. Capturing more detail will result in more pronounced camera-shake if your shutter speed is too slow or your hands are unsteady, especially if you then chop out a sizeable crop from the image.
Not all lenses are worthy of a sensor with a high resolution, too. Cheaper optics don’t have such a good “resolving power”, meaning they’re not great at rendering finer details, so to make the most of more pixel power you really need pro-level lenses.
This all adds up to make photography more expensive, not just from needing better lenses, but also requiring all the extra storage capacity as well.
The thing is, in most circumstances, we already have way more megapixels than we actually need. The biggest I ever really print my images is approximately A3-size, and to do this at 300dpi you only need a file that measures 3507x4959px, which you’ll get from a 17.3MP sensor. That means I’m casting aside detail whenever I make a print from a shot I’ve taken on my Nikon D800 or D810. Even if I wanted the image to fill a billboard it would still be adequate: using a lower resolution print is no problem if the image is viewed from some distance. But quite often, images we take only ever live online, and never even make it to print. If you only upload an image to Facebook then it’s resized to 1200x900 pixels, which is roughly the size of a 1-megapixel sensor!
It’s worth knowing that if for whatever reason you need a much higher resolution file than your sensor can capture, you could always shoot overlapping images and stitch them together in Photoshop. Usually this technique is used for creating panoramas to produce a much wider field of view than an ordinary lens would allow. However some photographers use what’s known as the Brenizer Method, which requires you shoot overlapping frames both horizontally and vertically so that each individual shot effectively acts as a jigsaw piece of a much larger image. This approach is often used for portraits, though is only suitable for subjects that remain entirely static. While this will create a much higher resolution file, a bigger pixel count isn’t its raison d'être. The key benefit is creating an enhanced depth of field to mimic the look of medium format cameras. In truth, whenever I employ this technique I always then downsize the final image, as having a 200MP file is both impractical and totally unnecessary.
But just because we already have more megapixels than we truly need, doesn’t mean I would automatically reject having even more at my disposal. I enjoy capturing high resolution images. It feels almost scientific when you zoom into an area, maybe a subject’s eye or a distant part of a cityscape, and you can view detail you didn’t know was there.
I know it’s unnecessary for 99% of applications, is ultimately detrimental to the image quality, and slows down my workflow, but I imagine I’ll be tempted to upgrade whenever a higher resolution sensor is released.
About the Author
Ben Davis is an award-winning professional photographer with more than 10 years' experience in the industry. His internet home is www.cambridgeshireweddingphotography.com