WHAT MAKES THE PIXEL AND IPHONE 11 IMAGE QUALITY CLASS LEADING?

WARNING! Before you shalaye, take into consideration the fact that this talks about image quality not video recording or 100x zoom or any kind of versatility.


It’s no secret that the Google Pixel phones and the iPhone have always topped the charts when it comes to sheer image quality, year in year out. Sure, Samsung and Huawei may have always had the better hardware, but that alone is hardly ever only what makes a good photo in a smartphone camera.

The secret is computational photography. It’s basically software processing extraordinaire to get more out of your camera hardware. That’s really important as you can only fit so much camera hardware into a phone with emphasis on the lenses and image sensors. You can’t cheat physics.

HDR was one of the first notable implementations of computational photography. The small image sensors found in smartphones could and still can only take in so much light meaning low light photography suffers immeasurably. So, by taking two or three photos at different exposure levels, software algorithms could produce significantly higher dynamic range, meaning brighter shadows and well exposed bright areas / light sources with more detail.
But unfortunately, due to inferior software and less processing power at that time (relative to now), HDR shots had a tendency to look really artificial. Thankfully, we’ve come a long way in technological advancement since Apple pioneered the HDR with its iPhone 4 in 2010.

THE PIXEL
The pixel is a product of Google’s excellent software engineering prowess. In fact, it took Google five years to perfect its native camera app – Gcam before launching it in its first pixel phone. Just by looking at the name – Pixel, you already know what to expect.

HDR+
Google made a significant leap in HDR technology with its advent of the HDR+. It basically uses the same method as the conventional HDR technology in that it takes quick bursts of shots and compares them to give an overall cleaner image…albeit on steroids. Just like the iPhone’s Deep Fusion, Google’s HDR+ starts ‘secretly taking photos’ once you open the camera before even taking the shot. Once you tap the shutter button, it takes the last raw image data and gets to work. HDR+ doesn’t just ensure you get a well exposed image, but one with reduced noise and accurate color reproduction. It works really well in day time and even better at night. The conventional HDR from rival companies bar apple aren’t as good. Sorry Samsung. In recent times, Huawei’s hardware has been closing the gap though.

SUPER HIGH RESOLUTION (ZOOM)
With the pixel 3 and 4, Google also introduced super high resolution. It was and still is a fundamental improvement to a technique known as demosaicing. When your smartphone camera takes a photo, it captures only red, green or blue data for each pixel (unlike Huawei’s RYYB filter). Demosaicing fills in the missing color data so each pixel has values for all three color components.
Super resolution takes advantage of camera wobble due to the unsteady nature of our hands (which results in camera shake) to capture quick bursts of photos. This lets the camera figure out the true RGB data for each pixel element of the scene without using the conventional demosaicing methods. This translates to much better digital zoom quality than several phones out there, even matching optical zoom of some smartphones at close zoom distances, say 2x or 3x for instance.

PIXEL VISUAL CORE/NEURAL CORE
But what good is all this if you don’t have the processing power to support it? Introducing Pixel Visual core and Pixel Neural core. The former was used in the Pixel 2 and 3 while the latter, in the latest pixel 4 and pixel 4 xl.
Aside from the conventional, general-purpose CPU, you may have heard of processors dedicated to more specific tasks like the NPU (Neural Processing Unit) in Huawei’s Kirin chips, TPU (Tensor…) in Google’s cloud servers, GPU (Graphics…) for gaming, IPU (Image…) etc. These are processors optimized to perform few specific complex mathematical tasks. The Visual core is an IPU while the Neural core is an NPU.
According to Google, the Visual core can compile HDR+ images 5 times faster than a normal CPU would with 1/10th the power consumption. The core handles the complex machine learning/neural networking tasks relating to the camera. It uses this core to augment the ISP and DSP already present in Qualcomm chips meaning quicker computational photography and a chance to include even more complex machine learning. In short, Google’s in-house silicon crunches far more numbers than your typical CPU can. The Neural core is a bigger upgrade to the Visual core, so it’s alright to expect better results.

UNFETTERED ACCESS TO DATA/EXPERIENCE
Google arguably has the largest access to data of any company in the world. Their search engine became so popular that it was coined as a dictionary term – google, meaning to search for information online. Not to mention its plethora of apps that have become indispensable in the lives of many. An app like Google’s photos grants the company access to millions of pictures of mind-boggling variety to play with. This it takes advantage of, combining it with neural networking to constantly train and retrain its AI to perfect the art of computational photography – sceneries, astrophotography, portraits, buildings, landscapes, animals etc.

CREATING A NICHE
With the introduction of the Pixel phones in 2016, Google chose an aspect of the market to focus on – CAMERA (PHOTOS to be more precise). That is arguably the biggest selling point of its phones as is made obvious by its name. It’s why a person will choose it over a Oneplus. By committing most of its resources towards this area, it was able to do much better than its aforementioned counterpart who focused on performance and others like Samsung who offered a one-size-fits-all approach. It made a name for itself, establishing itself as the go-to brand for image quality in the android space.


THE IPHONE

Apple have always had some of the best engineers the industry has seen. This is to no small extent relevant to their dominance in the smartphone industry (there’s also marketing and fan base). I believe in recent times, the iPhone 11 has seen the biggest generational leap in image quality.

DEEP FUSION
The iPhone 11’s Deep Fusion - an equally sophisticated variation of Google’s HDR+ takes four pairs of images – four long and four short exposure ones (before you hit the shutter button), then a longer exposure one upon taking the shot. Complex algorithms then find the best combinations, analyzing the shots to figure out what kind of subject matter it should optimize for, PIXEL BY PIXEL. This is where the A13 Bionic shines. The fact that the iPhone SE 2020 could shoot iPhone 11 grade quality shots whilst having the same camera specs of the iPhone 8 goes to show you how much processing power is needed for every shot taken by the iPhone thus highlighting the importance of the A13 bionic chip and the complexity of the iPhone’s Digital processing.

PHOTOSEGMENTATION
Apple has another one called Photo Segmentation. This uses machine learning to split a photo into tiny pieces and then process each one individually in an instant. So, if you take a portrait shot for instance or a selfie, the software can decide to push its resources towards your face, clothes, hair, beard, makeup etc while doing less to the background. Thus, sometimes detail may vary in some areas- usually parts of the image you hardly notice or care about. This also happens in low-light.

PORTRAITS
How about Portraits? The iPhone and Pixel have been dubbed severally as the kings of portrait, with your victor being a matter of personal preference. Even with a single camera as seen on the iPhone SE 2020 and Google pixel 2 and 3, both companies make a depth sensor/ telephoto lens seem nearly unnecessary. In contrast (if I remember correctly), the Samsung s9 failed woefully in portrait shots, with edge detection being its weakest point.
Google’s access to data is phenomenal. Having made its services nearly indispensable in a large part of the world, the software behemoth has millions of images to play with, to constantly train and retrain its AI to perfect the art of bokeh. The pixel visual/neural cores work alongside the available ISP (image signal processor) to calculate a depth map of the image using data from the dual pixel autofocus. The AI does some kind of abracadabra and then, the computation begins. How this computation is carried out makes the difference.
Apple’s single camera smartphones employ a similar process, but we have little information on the A13 Bionic chip and thus the peculiarities of how the process is carried out.

BRAND REPUTATION
Apple has over the years established itself as the celebrity smartphone, literally and figuratively. Not to mention its wallet-ripping pricing. It’s been the default device of celebrities for taking photos for as long as I can personally remember. Because of these, any dip in its ability to act as the perfect smartphone replacement for a dslr could prove costly. For this reason, Apple makes no compromises in its handling of its camera section. Splurging of the cash and recruiting of the best do well to ensure that the first trillion dollar company’s trademark product never falls short of the competition’s imaging capabilities.

CONC.
It’s not that every other brand out there doesn’t use computational photography. Basically every smartphone out there currently does. It’s just that Google and Apple have taken it a notch higher, no pun intended…for Apple.

Modified from the Web.

Just simply say optimization done properly.......

Unlike some Tecno crap.

software

Huawei is good, currently p40 pro plus is the best when it comes to photography.
No other oem, takes better night shots than Huawei
Well every oem has thier flaws, it all boils than to software optimization.

Detailed.
Unintentionally explains why battery takes a nosedive too when using camera app.

Karex81:
Huawei is good, currently p40 pro plus is the best when it comes to photography.
No other oem, takes better night shots than Huawei
Well every oem has thier flaws, it all boils than to software optimization.

Thank u, people just believe in general assumption, i can list 5phones that beats those two phones hands down in raw photography.
anytime ignorant prople like this talk, i just refer them to dxomark rating.

Comon Oppo reno 3 pro and xiaomi mi cc9 pro beats those two phones hands down.

seunfly:

Thank u, people just believe in general assumption, i can list 5phones that beats those two phones hands down in raw photography.
anytime ignorant prople like this talk, i just refer them to dxomark rating.

Comon Oppo reno 3 pro and xiaomi mi cc9 pro beats those two phones hands down.

.

I weep for you.

Qsand:
.

I weep for you.

Abeg weep for ur self.

Pixel and iPhone are not class leading anymore. Huawei has been sitting on the top spot for at least 2 years now.

Because they don't do things like tecno and infinix



.na dere my brain carry me reach oo, nobody should qoute me abeg

chatinent:
Unlike some Tecno crap.

Am with you,

You have my support

Comrade

As at now best camera phone is Huawei p40pro+ iPhone 11promax has the best battery, s20ultra is versatile good in all aspects also. All have good camera and it's their processor that do that shit