When they announced the Google Pixel Buds, the company’s devices chief Rick Osterloh made a pertinent comment about smartphones.
“I don’t envy those of you who have to write reviews for a bunch of smartphones with very similar specs. Megapixels in the camera, processor speed, modem throughput, battery life, display quality… These core features are table stakes now… To be honest, it’s going to be tougher and tougher for people to develop new, exciting products each year, because that’s no longer the timetable for big leaps in hardware alone,” Osterloh said.
What Osterloh essentially said was that even as we buy a phone every six months, the industry itself is facing massive technology saturation. Chipset makers have developed more efficient smartphone processors, but the once predicted “doubling of performance” doesn’t happen anymore. Battery technology has never progressed as fast as others, while display quality though advanced, has also reached a tipping point. OLED displays, which were once considered to be too expensive, have made it even to the more affordable segments.
And with this saturation, companies are going back to the basics — the megapixel (MP) wars.
It started with 48MP sensors announced by Sony and Samsung, which companies like Xiaomi, Realme, OnePlus and many others have used this year. Xiaomi even announced a smartphone with a 64MP camera sensor recently, while its competitor Realme has announced that a device with a similar camera is coming soon. Since both these use a Samsung ISOCELL sensor, we could expect Samsung to come up with such devices too.
The real difference this time though is that the focus is on software, and interestingly we can credit Google for that. The first 48 megapixel sensors were really never about the resolution. They used those extra pixels for a process called pixel binning, which combined four pixels to make one big pixel, thereby enhancing low-light imaging. The resulting photograph was a 12MP photo, which one could argue was where these cameras did their best work.
Samsung mentioned similar capabilities in its upcoming Bright HMX 64MP and 108MP sensors as well, naming it “Tetracell technology”. But here comes the catch.
When using high resolution sensors like this, many manufacturers default to the pixel binned resolution. That is why phones with 48MP cameras default to 12MP resolution. The explanation for this is usually that 48MP photos are also huge in size, meaning they’ll take a toll on your phone’s storage.
However, since smartphone sensors are considerably smaller than those on DSLRs, pixels here are really jostling for space. That means the pixels are packed really closely together, thereby increasing noise in the photographs you take. And companies again use algorithms to tackle this aspect. The better a company’s understanding of this problem better the results, but it’s never without trade-offs.
A look at full resolution shots from these sensors will often show colour smearing, where photos often look like they’ve been painted with water colours. You could try this yourself by taking a full resolution photo and doing a 100% crop on them.
With pixel binning, smartphone sensors are able to take in more light on paper, but there are other factors involved — like the lens aperture and how light is focused on the sensor. Smartphone sensors are usually limited when it comes to wider aperture lenses as compared to DSLRs, which is another reason why high resolution sensors may not give you exactly what you expect today.
It’s because of limitations like these that mobile cameras have to focus on software. Interestingly, the best proof of this are Google’s Pixel devices, which have sported some of the best cameras in the last three years without using any of these high resolution sensors. Google does this by using the vast amounts of data it has on virtually everything, showing us the beginnings of computational photography creeping into smartphone cameras.
So, at the end of the day, these “super high resolution” sensors may not be the answer to the issues smartphone cameras face today, bringing the MP war to a naught.
While these sensors, of course, sound great on spec sheets, they don’t necessarily make smartphone cameras better.
When the megapixel wars first began, the higher resolutions were making up for lacklustre software. But now, with the vast amounts of data, higher resolutions may remain just specs, while artificial intelligence (AI) and other software take center stage.