Smartphone Camera’s Innovation: Hardware vs Software
Back in February 2019, HMD Nokia had unveiled an array of their latest product during Mobile World Congress (MWC) 2019 in Barcelona. Out of all the products present, the one where everyone have been continuously talking about was the long-awaited 9 PureView (9PV).
The latest phone to carry the iconic PureView badge, it sported five identical 12MP ZEISS camera consisting of two RGB lens and three monochrome lens alongside a time-of-flight (ToF) sensor. Furthermore, the camera was developed in conjunction with Light, the company that brought out the mad L16 camera. Little wonders then on why everyone were having high hopes for HMD Nokia to regain back their crown in the world of mobile photography.
But right now, camera performance aren’t my concern. What interest me with the 9PV is the phone design. According to Juho, all the camera was housed on an 8mm thick body. This meant a super slim design without the need for any camera bump whatsoever. Nokia’s decision on creating a powerful multi-setup camera under a slim design did sound like a perfectly common sense, isn’t it? Or does it?
Let’s be honest on one thing: Even for a mobile photography aficionado like me, smartphone cameras still have a long way to reach the equal photo quality level of a proper DSLR camera despite various innovation. All the recent innovation such as Night Sight and SuperZoom were made for software processing. Software processing spelled for computational photography. The idea is that a photo is taken through rigorous algorithms of the camera’s imaging software to create an output equal to a proper DSLR or Mirrorless camera.
The problem, however, is that smartphone camera tend to have a rather small sensor. In the case of the 9PV, all the identical 12MP camera have a sensor size of 1/2.9". That is way, way less than you will find a 1-inch Mirrorless, let alone a proper Full Frame DSLR camera.
At this point, you might be wondering to yourself, “Why should I be bothered with all of this sensor nonsense? As long as my phone can take a nice photo, that’s all that mattered right?”
Well here’s the thing: When it comes to mobile photography, I always believe that there is one area where smartphone camera always struggle the most: low light photography.
Photo taken by yours truly via a Microsoft Lumia 950 XL without any edit whatsoever.
For me, low light condition has always been the ultimate true test in mobile photography. This is where you can see whether physics or AI delivers the best image processing. Sure that Google gave us Night Sight as mentioned earlier, but that doesn’t mean they are the real winner. A similarly low light photo taken from a phone with larger sensor and manual setting (i.e shutter speed) without any AI interference may take equally good or even better photo than the one with software and AI trickery.
> But how relevant camera sensor is when it comes to dark places?
Put it simply, it’s all about light and details. The bigger the camera sensor, the more area the light can gather inside the sensor. Also in return, you got more details captured in the camera. This prevents the camera from generating a very noisy image.
> So what’s the toast? What do I think?
Smartphone manufacturers of today focuses deeply on giving consumers versatility. It’s all about choosing the best multi camera setup. One would go monochrome, another one would go depth sensor, while there will be someone who would go all the way with ultrawide and telephoto zoom.
Whichever route the manufacturers go, they will always be focusing on the software bit. It feels like as if they have realized that the consumers didn’t really want to see big bump on the back of their phone or that most people are not really that interested with the engineering behind the camera as long as it looks good on social media, particularly Instagram. I honestly believed that in terms of photo quality, we are actually going nowhere. Yes, there are vast of innovations came up by manufacturers as mentioned earlier in the post, but what’s the point of that if the photo we got looks much like a Photoshop painting of a seven year old?
So back to Juho’s point about the 9PV’s super slim design without camera bump. Undeniably, credit to them for achieving such feat despite the five camera and ToF arrangement. But personally? I rather have a bump for larger sensor rather than no bump with more camera but tiny sensors. Looks like we really are still far away from DSLR photo quality in the pocket.
Originally published at https://otakupetrolheadslife.blogspot.com on May 1, 2019.
