Paul Briden 23/03/2018 - 4:37pm

Google's third-gen Pixel phones are coming in late 2018, but how could the Pixel 2's stand-out feature be enhanced?

Google's Pixel 3 series is going to arrive later this year, likely during October as it has done the last two years running. There's currently a drip feed of information leaking out about the forthcoming handsets, we know, for example, that it'll run a Qualcomm Snapdragon 845 processor, but for the most part we remain in the dark, or have vague hints to go on.

Amongst the leading rumours is the suggestion there will be three individual models; a Google Pixel 3, a Google Pixel 3 XL, and a third model- as was intended to be the plan in 2017, but which was scrapped in favour of just a pair of phones once again. On the third model, not much is known, but speculation is rife that it could be a cheaper model aimed at developing markets and the classic Nexus fanbase.

One of the standout features for the Pixel 2 series was the camera. The first-gen Pixel cameras were nothing particularly special, but with the second-gen Google - with a little help from HTC and LG, of course - smashed it out the park. The Pixel 2 series cameras received rave reviews and are considered by many to still be some of the best phone cameras in the business right now.

So where can Google go from here for the Pixel 3? You can be sure that many consumers will be watching eagle-eyed to see if Google improves the formula for its imaging tech.

There are indeed rumours that Google will update the sensors inside the Pixel 3 series, but so far nothing is really known. We can only speculate.

One thing worth noting is that the Pixel 2 series uses a single 12MP sensor with a nice wide f/1.8 aperture, optical image stabilisation (OIS), and dual-pixel-phase-detection autofocus, and combines this with some highly sophisticated software to get the stunning results it achieves. Here's an excerpt from our Pixel 2 review to illustrate how this works.

"Basically, the sensor, once you’re in Portrait Mode, isolates the subject of the image via machine learning, using the dedicated chip inside the phone. Once it has done this, the background is blurred, creating that DSLR style finish.

The INS and OUT of how this technology actually functions are rather complicated, but the resulting effect is designed to create a dual-lens-style shot with a single lens setup, thereby negating the need to have a dual-lens camera on the rear.

“Starting from an HDR+ picture,” said Google, “we next decide which pixels belong to the foreground (typically a person) and which belong to the background. This is a tricky problem because unlike chroma keying (a.k.a. green-screening) in the movie industry, we can't assume that the background is green (or blue, or any other colour) Instead, we apply machine learning.”

This technology is also available on the front-facing camera too, so if taking selfies is something you do a lot, you can get some pretty good results using Portrait Mode.

The idea behind this technology is to mirror, as closely as possible, the way a DSLR captures images, with that tell-tale background blurring.

Granted, you’re going to get better results with a $2000 DLSR, but for a phone, Portrait Mode is a great addition to an already very impressive piece of kit."

You can be pretty sure such software-based tech is going to make a return in the Pixel 3 series. Google could feasibly improve the results even further by actually tweaking the hardware as well.

A few possiblities include:

  • Adding a dual-sensor setup; this could possibly allow for optical zoom, but probably wouldn't improve the image quality much past what we see already.
  • Adding a wider f/1.7 or f/1.5 aperture, or even adding a variable aperture which adjusts, like the human eye, to light, as we see on the Samsung Galaxy S9 series; the Galaxy S9 dynamically adjusts between an f/1.5 and f/2.4 by using a disc-like mechanicsm.

In fact there's no reason why all of the above could not be implemented in the same handset.

Apart from this it's difficult to imagine precisely how Google could improve its current camera offerings, and whether it should do the above options is another question entirely. They're sure to cost a fair bit to develop and produce; a cost which will inevitably be passed on to the consumer in the final RRP of the device.