/Explained: The megapixel myth

Explained: The megapixel myth

Ah, the magnificent megapixel. The term so commonly used in marketing that it has become a staple on any spec-sheet which relates to a camera, whether that’s the one bolted onto a smartphone or an actual full-frame DSLR. Unfortunately, it is also the term that is so commonly misjudged by people, and the one term that gets thrown around far too often in internet arguments, and also a grossly misinformed one in that case.

There was a time and place where each megapixel mattered. However, in 2017, we have seen a shift in philosophy, shying away from pixel count in favor of other metrics. While many didn’t mind, there were pixel fanatics who shouted heresy. Personally, I think it’s high time we debunk the megapixel myth and why a higher megapixel count is not always better.

WHAT IS A MEGAPIXEL?

This slideshow requires JavaScript.

A megapixel is approximately 1 million total pixels. If an image resolution is 6000×4000, that would be a 24 megapixel image. It’s not uncommon to see double-digit megapixel counts in even budget-tier smartphones these days.

Basically, it’s a unit to measure resolution. The more pixels there are, the higher the resolution. In theory, this would mean fine details can be more easily resolved and overall creates a more detailed picture. A higher resolution also helps in cropping, as you still have a bunch of pixels to play with so you don’t lose a lot of detail if you decide to crop into a specific part of an image.

SO MORE PIXELS ARE BETTER, RIGHT? WHAT GIVES?!

Well, notice that I said “in theory”. However, this is not such the case in the real world. Because truth be told, there are several other factors that go into determining overall image quality. Megapixel count is just one of them. There are others that are just as important, while some of them may be considered even more important.

SO, WHAT ARE THOSE FACTORS?

Well, since you so kindly asked, here are a few that count, although there are way more.

Sensor size – The actual size of the camera sensor. They range from very small sensors typically used on front-facing cameras on smartphones to big medium-format sensors used on medium-format cameras like the Fujifilm GFX 50S. A larger sensor can house more megapixels with larger individial pixels and as a side effect, also create much softer “bokeh” effects than a smaller one.

Pixel size – The size of each individual pixel. A larger pixel size is typically better, since they can be more sensitive to light, this means that they can have an expanded ISO range, typically lower-noise when paired with a larger sensor, and also helps improve dynamic range.

Sensor type – There are 2 types of digital sensors available; Charged-coupled device (CCD) sensors and Complementary Metal-Oxide Semiconductor (CMOS) sensors. CCD sensors typically have better image quality but CMOS has been the most-common type of camera sensor due to its low-cost and low power consumption. Recent CMOS sensors have caught up with the image quality of CCD sensors, although the latter may still be used in niche products.

Lens quality – The lens itself also plays a role in photo output. A good lens will have optics carefully engineered to reduce ghosting, flaring, chromatic aberrations, and fringing while increasing natural sharpness and overall clarity.

Aperture – The part of the lens that lets light into the sensor. A wider aperture will let in more light, leading to better performance in non-ideal lighting conditions, but as a consequence, may cause some loss in corner-sharpness and some vignetting.

Stabilization – Either optical or digital, this helps stabilize an image to make it more likely to capture a sharp shot. Optical stabilization is especially useful in low-light environments while digital stabilization will have a bigger impact on video output.

Image Signal Processor – Arguably one of the most understated parts of a digital camera, the image signal processor or ISP takes the digital data captured by the sensor and processes that information to create a view-able photo to our eyes. The processor also controls other factors like white balance and autofocus and is key to features like HDR photography. Many of the sophisticated automatic functions on your camera is in this little guy’s hands.

Automatic post-processing – When shooting in compressed formats like JPEG, your camera firmware will apply some form of post-processing to the RAW file before compressing it. Usually on smartphones, this post-processing will follow a set algorithm defined by the manufacturer regarding factors like saturation and sharpness. However, some smartphones and most interchangeable-lens cameras allow you to adjust the post-processing to a certain degree.

OKAY, SO WHY AREN’T MORE MEGAPIXELS ALWAYS BETTER?

Deciding on the resolution of a camera sensor is always a balancing act, because changing the megapixel count will affect other variables, especially pixel size.

When you add more megapixels but keep the sensor size the same, the pixels will get smaller. Smaller pixels are less sensitive to light compared to larger ones, although a slight reduction (like say 1.55 microns to 1.4 microns) typically does not impair performance by a huge degree. A significantly smaller pixel size can cause narrower dynamic range due to the pixel’s limited ability to sense a wider variety of changing light alongside weaker low-light performance due to the need to jack up the ISO, especially when paired with a smaller image sensor. Usually, a higher resolution will come alongside a larger sensor to compensate, alongside features like OIS and improved post-processing like HDR+ on the Google Pixel (which requires a good signal processor).

However, the other extreme is not ideal, as HTC found out with the first-generation UltraPixel sensor on the One M7 and M8. While the camera sensor featured an enormous pixel size of 2 microns, it came at the expense of resolution. While 4MP was ample at the time, it showed some downsides in terms of detail. While many considered it fine for casual social media use, when put aside many of its competitors, its daylight performance stacked up rather roughly. Thanks to recent advancements, however, the current HTC flagship, the U11, has a much better all-rounder camera, and has a sensor that manages to finely balance low-light performance with reasonably large pixels plus stabilization and a wide aperture while having a very reasonable 12MP resolution for detail in daylight.

BUT WHY ARE SOME CAMERAS ABLE TO SHOOT GREAT PHOTOS EVEN WITH SMALLER PIXELS?

Great question.

The digital camera tech landscape has seen a rapid improvement in recent years, fueled by the popularity of the modern smartphone and the increasing reliance on its cameras by users. While it is very difficult to bend the laws of physics, advancements like backside-illuminated sensors and pixel supersampling have helped to alleviate some of the downsides of smaller pixels.

In recent memory, we’ve seen smartphones like the LG G4 and Samsung Galaxy S6 being able to take very solid low-light photos (for a smartphone) despite their smaller pixels due to the combination of a large sensor, wide aperture, OIS and solid post-processing (it is worth noting however that Samsung has recently moved to a lower-resolution 4:3 image sensor on the S7 and S8 while retaining a similar sensor size and having a wider f/1.7 aperture. As a result, the pixels are larger, meaning better lowlight performance although daylight performance suffers a little, due to the loss of some horizontal pixels which made up the 16:9 aspect ratio of the S6). Huawei’s recent flagships may feature relatively small pixels compared to the competition, but its use of a secondary sensor which omits a Bayer RGB filter, allowing it to be more sensitive to light, has compensated quite a bit for its smaller pixels. Coupled with the wide aperture and OIS on the P10+, the phone is very capable of producing very solid low-light photos with some grain due to Leica’s method of post-processing, which favors some noise over heavy noise-reduction.

This slideshow requires JavaScript.

Another method is using sophisticated post-processing. HDR+ on the Google Nexus and Pixel phones is a good example. While the current Pixel has a large sensor coupled with large pixels, HDR+ can actually benefit those smaller sensors through the use of computational photography and uses a mixture of image stacking, pixel re-aligning and averaging to deliver an image with a significantly wider dynamic range and less noise without resorting to conventional noise-reduction techniques.

This slideshow requires JavaScript.

Manufacturers seem to be focusing on efforts to help bring out more out of these smaller-pixel sensors, which should be a boon to many smartphone photographers, and should help traditional camerabugs when many of these technologies make their way to ILCs, like how the Sony Alpha A7R Mark II features a full-frame backside-illuminated sensor, the first time where BSI technology has made its way into a full-frame sensor.

CONCLUSION

 

The claim that a camera with the higher megapixel count will always win might probably be regarded by many in the photography community as one of the most ill-advised statements throughout the camera industry, and marketing material has not really done much to disprove that notion.

A higher megapixel count will definitely have some benefits, but its benefits should be weighed against its potential costs in other areas. As mentioned earlier, it has always been a balancing act and manufacturers typically have to decide on a good mix of both unless that camera is designed for a specific purpose.

 

It’s also worth noting that a higher megapixel count does not always result in a sharper picture. There has been an old saying that a sharp image from a lower-resolution sensor is better than a blurry image from a higher-resolution sensor, and that holds true. A high megapixel count can’t compensate for blurriness due to camera shake, and in extreme examples, the smaller pixels lead to more noise in low-light environments, which results in a blurrier photo due to the extra noise-reduction applied.

Keep in mind that this article is not trying to say that megapixels aren’t important. In some circumstances, they indeed are important. However, you should consider more factors before evaluating camera performance, as these other factors are often just as important, or even more important than the megapixel count alone.

Take note of that and you should be golden.