How Google Real Tone Camera Technology Works

Google developed Real Tone camera technology with a mission to make the camera and its image products more evenly distributed when shooting with more colorful results, especially in distinguishing skin tones. Historically, the lack of diverse testing meant that today’s cameras were not built to accurately depict the various skin tones of objects.

Unfortunately, this also means that smartphones continue to carry that bias, ultimately resulting in photos that are unattractive to people of color. But it’s not just smartphone cameras that struggle here. Previously there were various technological issues with cameras when it came to non-light skinned people.

So with the launch of its Pixel 6, Google is trying to solve this problem using Real Tone technology. With this feature, the camera’s hardware and software work in tandem to match the actual tones of the people in the frame. And it works really well, especially for first-generation products.

The way it works is also not as simple as one might think, and it seems more like there is a miracle here. But really, it’s no different from the other aspects of computational photography found in a number of other Android smartphones. Google puts a lot of effort into maximizing the resources and costs of delivering this technology.

The moment you tap the shutter button in the camera app, the job of producing great photos has already begun. The camera’s lenses and sensors aren’t particularly smart, but they’re great for their intended purpose, which is to collect and focus light. Using multiple electronic captures and filters, sensor data is collected at very specific points of light in red, blue, and green.

The data is not a photo, it is just data. It takes some processing to turn millions of points of light into an image, and that’s where Google always gets really great at this kind of game.

Using Machine Learning and AI, camera streams can look for shape and contrast to construct faces, distinguish between subjects and backgrounds, and more. One of the most important things that needs to happen to camera sensor data is to make the colors in the finished photo look the same as what the human eye can see.

That’s where Real Tone shows its strength. The camera software is refined to evaluate the skin tones it sees and expand the number of different color gradations. Think about the people you know and how different their skin tone is from each other or from your own. Humans aren’t just beige or tan, and camera software shouldn’t act that way. If you’re white, you’ve probably never had a problem with your skin tone in photos.

This is not an easy endeavor, but it is necessary. Exaggerating darker skin or applying less contrast won’t work, and there are billions of photos that prove it. Something like Google’s Real Tone project is a long overdue technology that many people have been waiting for.

Drop Your Comment

Proudly powered byWordPress. Theme byWeblizar.