Megapixels… more isn’t always better


The Redmi Note 10 Pro’s biggest selling point is its “108mp” camera

Companies always go for the top notch and people can be really be intrigued. One of the biggest trends in this “going-for-the-top” is the higher megapixel count. A few years ago, Xiaomi announced the Redmi Note 7. It took the world by storm because it has a 48MP camera alongside a 5mp depth sensor, and soon other companies followed suit and many CMOS manufacturers started topping each other by increasing the numbers

The question is… do you we really need that high of a megapixel count?

What are Megapixels?

By its simplest definition, a “megapixel” is one million pixels. This means a 12MP image has about 12 million individual pixels. The more pixels, the more details an image gets, and also the higher the resolution and bigger the image size. This is why 8MP images has lower storage size than a 12MP, albeit the 12MP has more details.

How does a phone capture, let’s say, a 48MP image?

The technique is simple, it’s called “Pixel binning”. But before we can get right into it, we need to know how pixels are measured first.

The Pixels in terms of Pixel binning, are the physical elements in which a phone’s camera sensor capture light in photography. Pixels are measured in microns (symbol: ┬Á). Microns are a millionth of a meter. Most flagship phones like the iPhone 12 and Google Pixel 3 have large micron pixel size so that more light can be captured, thus resulting a better image even during low light situations.

So now that you know, let’s get right into Pixel Binning.

Pixel Binning

The most simple way to explain Pixel Binning is that, it takes data from four pixels and then combines them into one. This means that a 48MP shot, when used with Pixel Binning, is actually just 12MP multiplied by four. Pixel-binning also divides the image resolution by four.

Now let’s take this to the extreme. The Redmi Note 10 Pro touts itself as a midrange phone with a high resolution of 108MP. While the phone indeed does have 108MP, it does not take native 108MP. Instead, it will use pixel binning to achieve that 108MP. Effectively, the image resolution is only 27MP natively.

Quad Bayer filter

Pixel binning is made possible through a Quad Bayer filter. A bayer filter is a colour filter found in any digital camera sensor. It uses the data taken from pixels and captures the image in an RGB (Red-Green-Blue) format.

Process of a Quad Bayer Filter. Courtesy of: Android Authority

A Standard Bayer filter has 25% red, 25% blue and 50% green. So, for example, a 12MP image has 6 million pixels that see green, and 3 million pixels for red and blue. With this in mind, a 48MP image has 24 million pixels that see green and 12 million pixels that see red and blue colours respectively.

So, is a Quad Bayer filter any different? Not really. It’s actually a slight misnomer but a Quad Bayer filter use the same ratio of 25-25-50. The biggest change is not the filter itself, but rather, the sensor. The sensor will combine four pixels into just one square, as shown above.

In reality, a 48MP image cannot provide more detail than a 12MP can, the same way how a 108MP image cannot provide more detail than what a 27MP camera can. Even then, its minute and subtle.

It’s not all about Megapixels

If you have seen our smartphone guide, then we told you not to base on MP alone, and instead base it on other factors such as software, sensor size, micron pixel size and aperture.

Companies know of this fact and, instead of providing more detail, they just raise the numbers. People will think that having a higher megapixel count offers higher quality. Megapixels are NOT a measure of quality.

Also, manufacturers will need to constantly find a way to squeeze those high megapixel count into an already cramped smartphone body, meaning pixels get smaller, and we all know that smaller pixel size capture less light.

So, in order to remedy this, smartphone companies rely on AI and camera software to churn out high-quality images. This is why Pixels and iPhones only have 12MP, as their software, and Google’s “computational photography” allow for great quality photos without squeezing too many pixels in a smartphone, and more often than not, they have better low-light performance without even the use of night mode thanks to a higher micron pixel count. These phones can defeat midrange phones that have 48MP, or even 64MP available in the market today.

Software and hardware tuning are the main reasons why, despite the similar sensor (say, Sony Exmor IMX586, the most common 48MP sensor), photo qualities turn out different as companies have to determine the quality of the image based on the chipset used as well as processing power.

Higher resolutions require more processing power, so if the processor is not efficient, then you might get lower quality photos. It also means that processing is slower as the chipset is determining the data captured from the sensor and then applies pixel-binning in order to get a high-quality image. This high processing power can temporarily slow down your phone.

Are higher MP counts bad?

Like the title suggests, higher MP is not always bad, but you should not use them to judge a picture’s quality, but rather, use them to judge a photo’s resolution. Remember that higher resolution does not mean higher quality.

There is a benefit to having higher MP. According to Anwesha Ghosh, marketing manager of DXO Image Labs, while 12mp camera is often “good enough” it doesn’t mean that phones should get stuck to just using 12MP. She also said that 48MP can attain for better optical zoom, as a higher megapixel count can zoom in 2x just through cropping and without losing quality and without digital interpolation, This is also the reason why the 108MP camera is present on the Galaxy S20 Ultra, so that, when you zoom in, quality is far less degraded, especially on moon shots.

With 108MP cameras, users can now have 3x optical zoom without losing so much detail. But again, only the resolution benefits while the quality degrades, especially when posting on Social Medias that compress images. Remember, that the sensor and pixel count are more important than megapixel numbers alone.

In Summary:

The Galaxy S20 Ultra showing multiple aspect ratios. Courtesy of: AndroidCentral

Megapixels should not be used to determine image quality, but rather pixel resolution. Having a higher resolution does not mean higher quality, and in fact, can degrade quality as manufacturers have to squeeze that many pixels into an already cramped smartphone. Having lower micron pixel count means noisier images and weaker low-light performance.

High Megapixel counts, especially on midrangers are actually just pixel-binned images. Pixel-binning is done through a Quad Bayer colour filter which combines four pixels into one. So that means, a 108MP pixel-binned image produces the same quality image as a 27MP one.

Higher MP can provide lossless optical zoom without digital interpolation. This is why smartphone companies often use them on periscopic or telephoto cameras on their flagships, as now you can take a crystal-clear image of the moon and zoom in without losing too much detail.

Higher MP counts need more processing power, thus slowing down the phone temporarily and also causes the image to be processed slower. Depending on the processor, these images may turn out to be high-quality, or very low quality if the processor is inefficient for the task, not counting apps running in the background or other processes.

If we wanted to determine the quality of the image, then software algorithms and the sensor are what we should depend on and it’s where the innovation should be moving forward. While smartphones cannot replace a DSLR, the images they produce is very close to what you can get in them, but smartphones are far more limited as their sensors cannot be replaced so easily, and with other factors, such as adjusting aperture, is only done through software and nothing much changes through its hardware.

Sources: