May 06, 2011
Conversation with Ambarella's Chris Day about the state of still/video imaging in mobile computing devices
In a recent blog entry I wrote about the generally low quality of cameras built into rugged mobile computers compared to the very rapidly advancing state-of-the-art in miniaturized imaging technology. It doesn't seem to make sense that high quality, costly tools for important jobs should be saddled with imaging hardware that ranges from only marginally acceptable to quite useless. Still and video cameras are now in tens of millions of smartphones and many of them now can take very passable high res still pictures as well as excellent video. I would expect no less from vertical market mobile computing hardware.
Why is that important?
Because the ability to visually document work, problems, situations or details is increasingly becoming part of the job, a part that can dramatically enhance productivity, timeliness and precision, as well as enable quick problem solving by consulting with home offices, etc., and it also helps create documentation trails. Add technologies such as geo-tagging and mapping, and the presence of high quality hybrid imaging functionality has an obvious and direct impact on return on investment as well as total cost of ownership. However, that only applies if the computer's still and video capturing capabilities are at the same high quality and performance level as the computer itself.
Over the past several months I have asked several of my contacts in the mobile computing world why the cameras aren't any better, especially since many of them highlight those cameras as productivity-enhancing new features. Which they can be, but generally are not, or not yet. The cameras are slow, produce unacceptable pictures (low resolution, artifacts, color, compression, sharpness, large speed delays, interface), and video is generally almost useless (very low resolution, very low frame rate, etc.). I did not receive any compelling answers, just tacit agreement and somewhat vague references to space and cost considerations.
So I decided to seek opinions from people at the forefront of today's miniaturized image processing solutions and get their side of the story. Molly McCarthy of Valley PR was kind enough to arrange a call with Chris Day, who is Vice President, Marketing and Business Development at Ambarella and has one of those very cool British accents.
Why did I seek out Ambarella? Because when we took apart a video scuba diving mask I had been testing, I found Ambarella chips inside. The product was the Liquid Image Scuba Series HD mask that has a high definition still/video camera built right into the mask. It can shoot 5-megapixel still pictures and also 720p HD video (see our review). The mask including the camera costs less than US$250 and it records on a microSD card. We also reviewed another tiny sports camera that includes Ambarella technology (the ContourHD), and that one can do full 1080p HD video.
What is Ambarella? It is a Silicon Valley company that was formed in 2004 to be a technology leader in low power, high definition video compression and image processing semiconductors. Chris explained that their main thrust is H.264 video compression, a technology that generates very good video at file sizes much smaller than conventional formats. Their largest market is what's called consumer hybrid cameras, the rapidly expanding segment of small cameras that can do both high quality, high resolution still images as well as superb high definition video. Ambarella is probably the leader in that area, and also the first to truly merge high-res video and still imaging.
Ambarella's hottest market right now is sports cameras, the kind that generate incredible HD video of skiing, skydiving, car racing, and all sorts of extreme sports (including, of course, scuba diving). They also do cameras for security and surveillance where the days of the grainy b&w low-res video often shown in "the world's dumbest criminals" type of TV shows is rapidly coming to an end. Ambarella also supplies other markets that rely on high compression but also high quality in their sophisticated imaging and forecasting systems.
About 400 people work for Ambarella these days, 100 of them at the Silicon Valley headquarters. For the most part, Ambarella makes chips, but they are also getting closer to providing full products, and already offer hardware/software development platforms.
I told Chris of my puzzlement over the primitive state of cameras built into most current mobile computers, especially considering that the professionals using those expensive high-quality computers could definitely use reliable, high-res cameras built into their equipment. Chris said that Ambarella did have discussions with several notebook manufacturers three to four years ago, but nothing ever came of it, primarily for cost reasons.
Now it must be understood that a good part of Ambarella's value-added consists of the chips that do very fast, very good video compression, and general purposes processors can do some of that, so perhaps consumer notebook makers simply didn't see the need for the extra speed and quality when most notebook users don't ask for more than basic webcam functionality.
Notebooks are one thing, of course, and tablets and smartphones another. Also to be considered is the fact that there are really two types of cameras used: vidcams for video conferences (increasingly referred to as "front-facing" cameras), and the much higher resolution documentation cameras (generally called "rear-facing") used like regular digital cameras. Most better smartphones and tablets now have two cameras, one for each purpose.
To that extent, Ambarella created their iOne smart camera solution that brings full HD camera and multimedia capabilities to Android-based devices. The iOne's SoC (System on Chip) supports live video streaming, WiFi upload of video clips, and full HD telepresence applications. It also has multi-format video decoding for playback of Internet-based video content up to 1080p60 resolution (i.e. better than HD TV). Chris felt that sooner or later one of the media tablet makers would truly differentiate itself with a superior built-in camera.
Ambarella also offers full development platforms for digital video/still imaging that contain the necessary tools, software, hardware and documentation to develop a hybrid DV/DSC camera functionality (see Ambarella consumer hybrid camera solutions here).
The bottom line, Chris Day said, is that "it is possible to have a mobile computing device that is also a world-class camera." We're just not seeing them yet. I am convinced that the first professional mobile computing product that offers the still/video recording capability of an inexpensive consumer camera will have a definite strategic and marketing advantage.
But what about the size and cost? As is, there are any number of imaging modules for those handy smartphones that are getting better all the time. They are tiny and inexpensive and light years ahead of what we now see in actual vertical market mobile handhelds and tablets.
A step up are the imaging modules that go into standard digital cameras. Those are larger and more complex, but judging by the tiny size of today's consumer point & shoot cameras that often offer 14 megapixel and 1080p video, those electronics should also easily fit into many mobile computing devices. They cost more, of course, but given the fact that many consumer cameras are now under US$100, it should be possible. Consider one product that uses Ambarella technology, the Sony Bloggie Touch. It can do 12.8mp stills, 1080p video, has 8GB of memory and a 3-inch touch LCD, yet it's hardly thicker than half an inch and costs under US$150. The guts of this in a rugged tablet or handheld would make an extremely attractive combination.
So the experts have spoken. It's doable. And it wouldn't even cost that much.
Video/imaging integrated into cellphones has changed the world. A lot of reporting now originates from smartphones before CNN ever gets there. And there's already talk that smartphones may essentially replace the conventional low-end camera market. The technology is there.
State-of-the-art DV/DSC Video/imaging could bring great value-added to rugged mobile computing hardware. Being able to document work, situations, conditions can be invaluable and truly open many new possibilities to get jobs done better and faster. But the pictures must be good, and users must be able to rely on the camera. Current camera modules cannot do that. HD video, likewise, could change everything. And it is truly lightyears ahead of the slow, grainy QVGA and VGA videos that most current computer cameras are limited to.
Posted by conradb212 at May 6, 2011 02:16 AM