Category: Computational Photography

Photography’s future is software & computation

This post is by Om Malik from On my Om

In an interview with DPReview, Sigma CEO Kazoo Tamaki made some observations about the impact of computational photography and the rise of camera phones. 

The technology that has impressed me the most is computational photography. The image quality from smartphones has improved drastically over the past several years, mainly due to computational photography. I’m amazed. This might change imaging technology. Camera and lens manufacturers need to learn something from it. Of course, we shouldn’t just copy the technology because we have much better hardware: bigger and better sensors, and better optics. But that kind of software is very powerful.

I believe they [the camera manufacturers] recognize the importance of computational photography. Still, as a camera and lens manufacturer, I feel we need to satisfy the very serious photographers and the history of photography culture. We don’t need to satisfy customers who just want to play with images or want a cartoon look. We have to follow the tradition of photography. People have been pursuing better picture quality in photography for over 150 years, right? So we have to respect the photo culture. But if there’s technology we can use to enhance picture quality that can also contribute to photography culture, why not? We should use it.

My takeaway from his comments was pretty simple: the camera industry is caught in the classic innovator’s dilemma. If they were smart, and if they could move faster, as an industry, they could try and embrace the change. They have bigger devices, historical knowledge, and the (Read more...)

The Smartphone Megapixel Race!

This post is by Om Malik from On my Om

Smartphone photography keeps marching on — and why not. After all, cameras, screens, and battery life are the key distinguishing features of most phones, especially in the Android ecosystem. And that is why we continue to see Android hardware makers — Samsung, Xiaomi, Huawei, and others try to one-up each other with camera technology and megapixels. 

Samsung will soon launch a new Galaxy (23) model featuring a new 200-megapixel camera sensor. The new sensor, the ISOCELL HP2, will pack 200 million 0.6-micrometer pixels in a 1/1.3″ optical format. This isn’t the first 200-megapixel sensor made by Samsung. The higher pixels allow for “pixel binning,” which allows the sensor to perform better. So, for instance, four pixels can be binned together to create 1.2μm size pixels to output 50-megapixel images. Bin 16, and you get to a 12.5-megapixel image, which can lead to a better quality of images. Apple’s iPhone also uses Pixel Binning in the latest iPhone 14 models. Apple uses Sony sensors.

Samsung says it has a new technology –Super QPD that leads to faster and more accurate auto-focusing, especially in low-light environments. In addition, Samsung says the sensor uses a “Dual Vertical Transfer Gate” that leads to better colors, less overexposure, and fewer washed-out colors. 

Since Samsung supplies these sensors to others, such as Xiaomi, it shouldn’t come as a surprise to see new phone models show up with these sensors. What will distinguish one phone from another is how the software harnesses the capabilities of these new (Read more...)

Why Camera isn’t just a camera

This post is by Om Malik from On my Om

I woke up this morning thinking about the new Apple Studio Display’s webcam hiccup. It has reaffirmed my belief that the camera, and by extension, the visual sensor, is becoming a key interface to the information and how we interpret it. What keyboard and mouse were to what was textual computing, visual (and other sensors) will be a key to computing in the future.  

As I said, it is the camera stupid

An article in the New Yorker laments that smartphone photography is too algorithmic. Similar laments were made when William Eggelston started experimenting with color film. Since then, our everyday memories have been captured on color film, each generation getting better than the others. It is the same for computational photography — we started with the grainy photos off Nokia, Blackberry, and the first iPhone. I remember the first iPhone and the photos that came off its puny sensor.

We have already come so far in this journey, and what writers overcome with nostalgia think is just a camera, isn’t just a camera.

But back to the Studio Display camera problems. 

Looking beyond, the speed with which Apple can fix the problem by issuing a software upgrade will reaffirm the advantage of what I wrote earlier about putting “smarts” into previously dumb devices. Apple’s ability to take all the gains offered by its iPhone business & its scale gives the company a significant leg-up in its ability to reinvent products. It will help it become the (Read more...)

Apple ProRaw + Adobe Super Resolution = Amazing!

This post is by Om Malik from On my Om

In December 2020, with the release of the iOS 14.3, the owners of iPhone 12 Pro (and ProMax), got to experience Apple’s new photo format, ProRaw. In simplest terms, the iPhone camera captures multiple image frames, picks out the best bits from these frames, and stitches them together in a photo with higher amount of data that can be manipulated for editing later. These are big files — about 10-12 times the sized on normal files captured by the iPhone.

In more recent days, Adobe announced a new version of Photoshop (and Camera Raw) image editing software especially for the M1 Mac. As part of these upgrades, the company unveiled a new feature called Super Resolution.

“The term ‘Super Resolution’ refers to the process of improving the quality of a photo by boosting its apparent resolution,” Adobe engineers write on the company blog. “Enlarging a photo often produces blurry details, but Super Resolution has an ace up its sleeve — an advanced machine learning model trained on millions of photos.”

Think of this as turning a 10-megapixel photo into a 40-megapixel photo. While I don’t need Super-Resolution with my Leica digital files, it is an interesting proposition when applied to cameras with smaller sensors, especially smartphones. I thought it made perfect sense to marry the ProRaw files from Apple’s iPhone 12 ProMax with Super Resolution. So, I did.

Last evening, I took a handful of photos with the iPhone’s normal and short-tele lenses in ProRaw format. I applied the (Read more...)