Video: Have we reached the limits of smartphone camera quality?
While much of the spotlight was on Apple’s latest camera updates during the recent iPhone launch, we almost missed this excellent explainer video from Rachel Yang at TED-Ed.
It breaks down why there's a ‘ceiling’ smartphone camera development is fast approaching, while also simplifying the reasons smartphones capture and process images differently to standalone cameras.
Yang’s video is a great, easy-to-follow explainer on why today’s mirrorless cameras and DSLRs might use the same buzzwords as the latest iPhones, Google Pixels, or Samsung Galaxy phones, but produce their images in completely different ways.
She breaks it down simply: image quality mostly comes down to sensor size and the optics in front of it, which affect three main things — resolution, dynamic range, and noise.
Over time, while standalone cameras have increased the size of their sensors, with a Type 1, or “one-inch”, sensor typically being the smallest used today and medium format the largest, smartphones haven't. They don’t have the space for a sensor nearly that big, and most don’t even have space for a Type 1.
To overcome this limit, smartphones have rapidly advanced their use of computational photography in recent years.
“When you snap a picture on your phone, this pocket-computer starts running complex algorithms, which often begin by secretly taking a string of photos in rapid succession,” Yang says. These are then layered together, helping manage high dynamic range, noise and light issues.
The problem is that sensor technology isn’t moving as fast. New sensors can take years to hit the market, and they often come with trade-offs, like resolution versus noise or readout speed, which is key for making computational photography work well.
It's also one of the reasons why smartphone manufacturers, like Apple, are increasingly leaning on machine learning and AI to do more heavy lifting, as it's hoped this will give smartphones better performance, despite the limitations of physics.