How can smartphones have so many megapixels? Pixel binning explained

Pixel Binning is one of the key techniques that allow smartphone manufacturers to pack more and more pixels into its cameras every year. But how does it work? Why is it necessary? Find out in this video which goes over the basics.

00:00 Introduction
0:35 What is a pixel?
1:58 What is pixel binning?
3:08 What are the advantages of pixel binning?
3:35 What phones use pixel binning?
4:01 What phones DON'T use pixel binning?
4:11 What is the disadvantage of pixel binning?
4:56 Conclusion

----------------------------------------------------
Follow us on social:

Affiliate Disclosure:

Follow the Team:

[sidewaysfcs]

In reality, the light-gathering surface of a 8x6mm size 12MP sensor with 2 micron pixels is equal-to-or-less than a 8x6mm 48MP sensor with 1 micron pixels, because there is always some small space between the pixels.

Also. In daylight, the 48MP sensor in unbinned mode is still useless, because the diffraction limit is given by the lens, in smartphone the lenses are so small that the effective resolutions for tipical sensor sizes is limited to about 16-20MP at best.

It is physically impossible to resolve details beyond 16-20MP resolution on such small sensors (smaller than 1'') with lenses so small.

It is also impossible to resolve details fully on pixel-sizes less than 0.8 microns, since red light as a wavelength of 0.8 microns at max.

[branimirteodorovic]

Cameras that use pixel binning almost always don't produce more details when using full-resolution mode. The reason is the quad-Bayer filter. Even though the sensor resolution is 48 MPx, it sits behind a 12 MPx color filter.
Therefore, the only benefit of high-resolution cameras with pixel binning is bragging rights. Smartphone manufacturers can market their cameras as high-resolution without producing terrible images in lower light.

[LeBeautiful]

Love how pixel binning is more recognizable now than what it was almost 10 years ago

[tanaka]

I like how you dismantle these complex concepts keep it up man💪💪

[lucasvanderleidearaujo]

the solution is simple, lower the resolution ( 8 or 12MP is more than enough) and increase the sensor size.

[TeoSluga]

You have got this all wrong from the beginning... More pixels is more light? Absolutely not true. More pixels pro ide more details. What is really crucial for providing light is the size of the sensor. You can put 108 millions of pixels to a small sensor: it will not help a bit with a low light photography. But hey, now you are asking how is it possible to produce a decent night photo with a small sensor? Here is when artificial intelligence comes in hand. The low light photos are artificially enlightened. So the best sensors out there are big in size (1:1) and DO NOT have huge number of pixels. The number of pixels in best cameras is somewhere between 12 and 36 million pixels.

[lucafogliati]

One thing that is not clear to me is if pixel binning impacts on the size of the image.without pixel binning the same camera (same sensor same lens) is supposed to produce different image size if the picture is shot at 4k resolution or 1920x1080 for example.lower res image is supposed to have smaller horizontal and vertical angle of view compared to 4k res.i Guess this is not happening with pixel binning that actually uses the full size of the sensor at every shooting resolution or am i wrong ?

[jyothyelizabethmartin]

What-
Does Brandon Harvard from MKBHD work here at Android Autority? 👀

[georgesabikhalil]

Thanks for this video! Now I know when to use pixel binning and when not.

[nomore-constipation]

I think the explanation of Pixel binning could be better if you try and show an example.

Having a hard time trying to explain this myself to another person but they still are kind of glazed over when I explain it or shared this video (watching at the same time)

Maybe a picture example and then a close-up. I tried using this verbally but I think they cannot see the math so to speak about how this is done

[KuranIsAFantasyBook]

The real question is "how smartphones got so many megapixels and still do shitty photos?" Even the top iPhones and Sammys still struggle with focus and noise.

[Aadith__menon]

1:52 1:53 these people are mkbhd's editors

[tobiasnitzsche]

Nice video! FYI - Bayer is pronounced "Buyer"

[anshulsingh]

Can you tell from which smartphone pixel binning or oversampling method introduced

[siddharthchakravarty]

I still do not understand why the influencers are still not calling out bullsh*t on quad bayer and pixel bining...!
I am a professional photographer and I can explain why pixel bining and quad bayers are crap and do not provide ANY advantage over just having a 12MP sensor directly. You want high resolution in day time? then over sampling is the work around which still is crap but some how you have been trained to expect your tiny 1 inch smartphone camera to capture too much detail for posting it on social media. I do not have a YouTube channel so if you wanna know just email me.

[B.D.F.]

Oh, so like the iPhone 5 in low light. It would be neat if the iPhone 14 had it on the 10th anniversary of the feature’s introduction to the iPhone. (Phil Schiller only talked about it for a few seconds on stage so most people missed it)

[AndreasLudwigPhD]

this makes no sense. the question to answer is why splitting the available sensor surface into 50 or 200 pixels is advantageous over just using larger pixels if you pixel bin and just provide 12MP pics at the end. My canon r5 has 45MP, my Google 6 claims to have 50. loading the raws into Lightroom you see the real quality. the ultra high MP of phones are just marketing bs to win the spec sheet war.

[Rushtik]

i still dont understand how it works, combining pixel, into 1 pixel, the 4 pixel has its own filter but combining them how do they identify rgb light since it act as one

[kenmastersmaster]

So... More light in night view and more detail in daylight? Wouldn't bigger pixels still have more detail?

[judenjilah]

So the best would be to take a high res pic and a pixel binned image and fuse into one image.