The December 2022 issue of IEEE Spectrum is here!

Close bar

Piecing Together a Picture of Photography Startup Light

Just funded for $9.7 million, Light is keeping its plans behind closed doors, but a few beams of information have slipped out

3 min read
Piecing Together a Picture of Photography Startup Light
Photo: Tekla Perry

Palo Alto-based startup Light hit the news this month when it closed a US $9.7-million round of venture funding. It revealed little about its plans, other than it’s “a team of creative technologists on a mission to reimagine the art and science of photography.”

The founders are Dave Grannan, who was CEO of Vlingo, a company that does speech recognition for mobile phones (including the first Siri app and Samsung’s S-Voice), and Rajiv Laroia, a founder of Flarion Technologies, a company acquired by Qualcomm that developed OFDMA, the technology behind LTE. Other executives identified include a manufacturing guy who worked on camera modules for Flextronics and a marketing guy who worked on mobile media for Disney. So it’s not too hard to figure out what this company would like to do—develop disruptive technology for mobile phone photography.

It’s been two years since Lytro made an attempt to disrupt camera technology with its focus-everywhere camera. The Lytro development showed the power of image processing algorithms, however, Lytro cameras didn’t fly into consumer hands. That’s likely because most people have simply given up carrying dedicated cameras—they carry phones that take good enough pictures, and have gotten used to accepting good enough instead of great.

So most engineers working on photography aren’t trying to design new stand alone cameras, they’re looking at ways to make the pictures we take with our phones better. Given that phone pictures are starting with a small lens, making a big leap in quality is going to take a major innovation in hardware or software—or both. Probably every camera and camera module manufacturer is trying to figure out what that innovation will be.

As is Google. It recently revealed software called Lens Blur, designed to simulate the lens of an SLR camera to allow users to select a depth of field. Its algorithms do this by creating 3-D models of a scene from multiple frames. Selective depth of field was one of the important features of the Lytro technology. Lens Blur solves one frustration—but isn’t a quantum leap.

Light clearly needs to do more than just deal with the depth of field problem. But what? I’d ask the company, if they were talking. They’re not. Here’s what I do know, based on a few digital bread crumbs left as the company has been evolving.

Photo: Tekla Perry

For one, it looks like Light is not just developing algorithms, they’re planning some kind of innovation in camera module hardware. Laroia, an IEEE Fellow, was in 2013 identifying himself as a founder of Tinz Optics, “a company developing technology that miniaturizes lenses/cameras while preserving the quality.” Tinz Optics has the same address as the current Light, and its URL now redirects there, so while there has been no official name change announcement (part of the whole secrecy thing, I guess), Tinz clearly morphed into Light. (The office, by way, is next to a Jamaican restaurant, so I have to wonder if Tinz is a a play on that restaurant’s habit of identifying items in its menu as "tings". Or a derivative of the Jamaican soft drink Ting. Something else I’ll ask Light if I ever get to talk to them.)

Of course, it’s not that Light will ignore depth of field algorithms. And they are also planning to use software to combine multiple frames into a single image. This information comes from a recent job posting on Linked In, seeking an  

...imaging scientist with a deep understanding of computational optics and image processing and familiarity with imaging sensors and geometric optics... [responsible for] developing innovative computational algorithms to combine multiple recorded images into a single image, image processing to reduce noise and increase image dynamic range, [and] extracting depth information in a scene.

That listing has been closed and more browsing around Linked In tells me that Light may have filled that post with a research scientist out of camera maker Ricoh who has written a bit about multi-aperature imaging, that is, using several small cameras instead of multiple shots from a single camera to get higher resolution images. Could this be Light’s approach—make each camera smaller and putting lots of them in the phone?

There’s a big market to capture for a company that revolutionizes mobile-phone-based photography. Will Light be the one? I’ll be watching—from a comfortable patio chair at the Jamaican restaurant, with a bottle of Ting.

The Conversation (0)

Deep Learning Could Bring the Concert Experience Home

The century-old quest for truly realistic sound production is finally paying off

12 min read
Vertical
Image containing multiple aspects such as instruments and left and right open hands.
Stuart Bradford
Blue

Now that recorded sound has become ubiquitous, we hardly think about it. From our smartphones, smart speakers, TVs, radios, disc players, and car sound systems, it’s an enduring and enjoyable presence in our lives. In 2017, a survey by the polling firm Nielsen suggested that some 90 percent of the U.S. population listens to music regularly and that, on average, they do so 32 hours per week.

Behind this free-flowing pleasure are enormous industries applying technology to the long-standing goal of reproducing sound with the greatest possible realism. From Edison’s phonograph and the horn speakers of the 1880s, successive generations of engineers in pursuit of this ideal invented and exploited countless technologies: triode vacuum tubes, dynamic loudspeakers, magnetic phonograph cartridges, solid-state amplifier circuits in scores of different topologies, electrostatic speakers, optical discs, stereo, and surround sound. And over the past five decades, digital technologies, like audio compression and streaming, have transformed the music industry.

Keep Reading ↓Show less
{"imageShortcodeIds":[]}