At some point over the last two years, millions of tech employees everywhere have come to the same conclusion: “Wow, my webcam sucks.” After days, weeks, months of conducting Zoom calls through the blurry haze of a MacBook Air camera, techies everywhere started to get fed up. Some rigged up elaborate replacements that turned a DSLR into a webcam; others mounted their smartphones; dedicated webcams, even the ones that barely suck less than the Air, became so popular that they were out of stock for months at a time.
Veeraj Chugh and Stefan Sohlstrom decided to fix their webcam problem in a different way: by building a better one. Chugh and Sohlstrom are the co-founders of Opal, which is the name of both the company they created and the webcam it’s launching on Tuesday. At $300, it’s among the most expensive devices on the market. But Chugh and Sohlstrom think people will pay the steep price as they come to recognize how much better they look on calls and how much that can matter. More than that, they think Opal is at the forefront of a home-office revolution, a burst of new startups and products designed to make remote work better.
Chugh and Sohlstrom began talking about webcams in the summer of 2020, while they were working together at Uber, after having their own wow-webcams-suck moments. “Just like everyone else,” Sohlstrom said, “we went straight to Zoom. And for me, I was like, ‘Wow, this is way worse than in person.’” The two men began discussing a simple question: Why are webcams so bad? Sohlstrom, a lifelong photography enthusiast, began digging into camera software and Photoshop filters, looking for ways to make himself look better. Chugh went to Alibaba and bought dozens of different webcams to test and take apart.
They discovered that there’s a pretty big reason why most webcams suck: There hasn’t been much incentive to make them better. Even the webcam giants like Logitech admit that, to some extent, they’d fallen asleep at the webcam wheel pre-pandemic because nobody seemed to care. “We're like, all right, we're at 90% market share, our products are pretty good, no one's really doing anything new there,” Scott Wharton, who runs Logitech’s video collaboration business, told me earlier this year. The pandemic, and the rush of demand for home-office gear, caught all these companies off guard.
Veeraj Chugh and Stefan Sohlstrom are the co-founders of Opal.Photo: Opal
But the better diagnosis of the webcam business, Chugh said, is because even the ones with high-end hardware don’t have the software chops to match. A high-end sensor and fast chip is wasted without the right software and neural processing engines, and the best software can’t solve for badly captured light. “It’s because hardware companies make webcams,” Sohlstrom said. “The companies that make that kind of stuff, they’re not necessarily known for computational photography and computer vision. Whereas if you take your iPhone, and you look at the difference between the last four of them … it’s not really hardware, right?” In recent years Apple, Samsung, Google and others have shifted from competing over megapixels to adding automated filters, software that can turn a blurry picture into a crisp picture, and much more.
At first, they had thought building a webcam might be a fun side hustle, or something they did just for themselves. But the more they looked at the space, the more they realized others might want the same thing they did. They also realized getting it right wouldn’t be easy. By the end of 2020, they’d both quit their jobs to start a webcam company. They started to put together a team, beginning with Kenny Sweet, who has worked on everything from Beats headphones to Dell laptops.
From the very beginning, the Opal team made a clear distinction internally. They weren’t just building a better webcam. There’s a whole supply chain and feature set made just for webcams, from the chassis to the chip to the images sensor, and Opal wanted no part of it. “As soon as you start talking to webcam vendors and webcam suppliers,” Chugh said, “you're literally picking between D+ and D- options.” Instead, they decided to build the best camera they possibly could, and then just mount it to a user’s computer. They bought processors from Intel and sensors from Sony, and cobbled together something approximating a smartphone without a screen and all the radios. They built a mesh of microphones to improve the audio experience and are working on the ever-complicated problem of canceling echos and background noise.
Some suppliers were actually surprised by how high Opal was aiming, Chugh said. “That was when we knew we were onto something good, when people were like, ‘Are you sure? For a webcam, this feels a little high-end.’ And we’re like, yeah, that’s the point.” Sweet and the design team took inspiration from Dieter Rams’ minimalism and Teenage Engineering’s emphasis on the technology, and built a camera that looks less like an eye perched atop your monitor and more like, well, a camera.
All along, Sohlstrom said, the hardware has developed ahead of the software. When we got on a Zoom call to talk about Opal — Chugh and I using Opal cameras, Sohlstrom on his MacBook camera because the Opal software he was currently running was too buggy — Sohlstrom acknowledged that even as Opal launches, there are big bugs to fix. The camera Opal sent me to try has a tendency to constantly search for focus, which he noticed immediately and said would be fixed in the next day or so. But in a way, that’s the point: Opal believes it’s shipping a webcam that’s already better than the one you’re currently using, and it will get even better over time. The hardware is just a platform for the ever-iterating software.
In the run-up to launch, Sohlstrom and his team are working on optimizing the camera for every video-chat app on the market, which itself is a huge amount of work. They’re mostly focused on the Mac, he said, where “it’s all this technology that is, frankly, old and untouched and unloved by Apple.” They spent months building manual controls so users can tweak brightness and bokeh and focus to their liking. Opal hopes most users will never need any of these settings, and can just plug in their camera and look great. But people who drop $300 on a webcam are also more likely than most to want to tweak their settings.
Opal's camera looks nice, but ultimately the hardware is only half the battle. Maybe not even half.Photo: Opal
The team is already starting to dream about what it can do with all the extra hardware power it has, too. Nearly all of the camera’s processing happens on the device itself, which both gives Opal extra control over the system and means it can do more without overloading the user’s computer. “We actually built a gesture-control engine,” Sohlstrom said, “that can basically recognize your hands … and allow you to use gestures to control different parts of your computer.” You could mute yourself with a closed-mouth gesture, or drop a hand-raise in Zoom by literally raising your hand. Or — and this is Sohlstrom’s favorite — what if you could end a Zoom call just by waving at the screen, instead of doing the awkward dance of finding the Leave button? Opal could eventually support AR features, metaverse-style avatars and much more.
The camera that’s shipping now is still something of a beta test. Chugh and Sohlstrom said they’re confident people will like the Opal better than their existing webcams — it helps that that’s an incredibly low bar — and then be along for the ride as the device gets even better. Given the price, though, the team doesn’t have much room for error, so they’ve been expanding the pool of Opal users slowly even though they say 16,000 people are currently on a waitlist to buy one. “We’re shipping out first in the hundreds, then the thousands, then the tens of thousands,” Sohlstrom said.
Exactly how big that market will get is the other question facing Opal. The company’s not necessarily attempting to eat the whole webcam industry. It just wants to make the high end much higher. “I think our target market is people who make a living on Zoom,” Chugh said. “Those people aren’t looking for value … what those people want is best in class.” Opal is betting that just as Instagram and FaceTime turned cameras into the most important feature of smartphones, Zoom and Teams might do the same for PCs. A few years from now, we won’t all be rectangles on a black screen talking to each other. We’ll be something more immersive, something more interesting. And that only works if the camera does.
