Camera Makers Are Starting to Get The AI Message
Canon recently announced they’re building deep learning image processing technology to improve the images their cameras record. It seems camera makers are starting to get the AI message. This online article, Deep Learning Image Processing Technology, is presented in English via Google Translate. The report is fascinating, since some of their photography examples reminded me of what my iPhone can do. I’ve been saying for quite some time that if the camera manufacturers don’t get image processing built into their cameras, as we have in our phones, they will eventually go by way of the dinosaur.

iPhone Pushing the Boundaries
The quality of images an iPhone can produce is not surprising, since Apple has over 800+ engineers working exclusively on the iPhone’s camera. I recall reading an article about this little-known fact back in 2015. It came from an interview 60 minutes did with Apple, where they announced this 800+ iPhone camera engineers factoid. Over 800 engineers! Is anyone surprised that the iPhone is getting more capable and outperforming real cameras in many situations?

And I discovered firsthand how the iPhone uses AI technology to punch way above its weight in small camera sensors. Let me share how I gave up on smaller high-end, point-and-shoot cameras in exchange for an iPhone 13+.
Traditional Point and Shoot No Match for the iPhone
It all began with a reasonably expensive, $1300.00US Sony RX100 Mark Vll. I purchased the RX100 specifically for its much larger 1-inch sensor that, I expected, would give me better image quality than a typical point-and-shoot camera with a much smaller sensor.

I was pleased with the Sony until late one evening on our Japan Winter Photo Tour. Tanya and I were high above the city of Tokyo. As I gazed out our hotel window, I became enthralled with the gorgeous lights of this big city’s urban landscape. I grabbed my Sony, turned it on, placed it in Program Mode with Auto ISO and began producing pictures.
Night Scenes of Tokyo

After a few images, I started reviewing them on the back LCD and was less than pleased with what looked like a lot of noise. At the time, I had an iPhone 6, if you can believe it. I’m not one to upgrade my phone with each new model release, and at the time, I wasn’t even thinking about using my phone for serious photography.

That’s why I bought the RX100; it was supposed to be a serious point-and-shoot that could give me professional results. Not happy with what I saw, I asked Tanya to hand me her iPhone 11, and I began to shoot.

Immediately I could see a significant difference on the iPhone screen. But once I got them into the computer, I was dumbfounded. The quality of the iPhone 11 images was beyond comparison to the Sony. The differences were astonishing!

I’ve been thinking about sharing this story for several years. The Canon AI story gave me the excuse to dig these images back out. Unfortunately, the Canon article does not say, as far as I could find—it’s a long article—if they are building this AI technology into a camera or if the image quality benefits are coming via Canon software on a computer. Either way, it shows what AI software technology can do. I’m using AI in several different desktop processing programs to improve the quality of all my digital and film images. But that’s a future blog post.
Let me know your thoughts on where this is all going and what you think of the iPhone 11 night shots compared to the same with the Sony RX100 Mark Vll.