"The inherent tradeoff between smaller pixels and improved sensitivity or image quality continues to drive innovation. The OV10820 delivers the best of both worlds: 10.5-megapixel high resolution imaging with unparalleled low-light sensitivity from RGBC in a compact footprint," said Devang Patel, senior product marketing manager at OmniVision. "The sensor offers 4K2K and fast frame rate FHD video in a native 16:9 aspect ratio, addressing the need to capture video for increasingly popular widescreen mobile displays and next-generation TVs with extremely high resolutions. Furthermore, the OV660 companion chip converts the RGBC data into RGB Bayer format, allowing manufacturers to integrate the camera solution with standard ISPs."
The 1/2.6-inch OV10820 utilizes a 1.4um OmniBSI-2 pixel with a RGBC color filter pattern. It records full-resolution 10.5MP video at 30fps at 296mW power, and supports 4K2K (3840 x 2160 pixels) at 30fps and 1080p60 video while maintaining full FOV with binning for 10b RAW output. The OV10820 fits in a camera module form factor of 9.5 x 9.5 x 6.4 mm, including AF.
The OV660 converts OmniVision's RGBC RAW data into standard RGB Bayer RAW data. The OV660 supports a primary RGBC camera sensor of up to 20MP in resolution and a secondary RGBC front-facing camera sensor. The OV660 eliminates the need for manufacturers and backend processors to modify their existing standard Bayer imaging pipeline and algorithms. The OV660 is available in a 4 x 4 mm WLCSP package.
The OV10820 features a high-speed 4-lane MIPI output interface to support the required high data transfer rate. The OV660 features two MIPI input ports and one MIPI output port. Both devices are currently shipping in volume production.
Update: Phonearena, Engadget, and other sites speculate that the new Omnivision sensor is inside Motorola X smartphone.
Phonearena has published many comparison pictures of the same subject shot by Motorola X, iPhone 5, Samsung Galaxy S4, and Nokia Lumia 1020.
What is the clear pixel SNR10 value
ReplyDeleteI thought Aptina will come to market first with clear pixel product but it looks like they loose to OV again. Its nice to see OV leading the pack though, heard that even Sony could not get clear pixel processing working.
ReplyDeleteYou are missing the July 18 post at this site. Or, look to the left under "5 Most Popular Posts".
DeleteLeading the pack? You mean after Toshiba and Kodak/Truesense Sony and Samsung? I think the problem was finding a customer willing to use RGBW in a consumer product. The image sensor technology has been there for quite a while. In fact, I think the earliest Kodak/Bayer patents included no-filter as a kernel element.
DeleteThanks but Aptina does not have its sensor in a phone that I know of. Do you know any?
DeleteInterestingly, from the initial Moto X/OV images in public domain it does look like the camera does not show a clear benefit over iphone. Aptina white paper suggests better improvement but who knows what issues that comes with.
Getting to market first here is irrelevant - OV's RGBC suffers the same problems as SONY's RGBC... Apple was just smart enough NOT to ship SONY's RGBC with the iPhone5. Aptina's ClarityPlus will ship later this year and it provides far superior results on every metric that matters...
ReplyDeleteOops, actually the R-G-B-W sampling was included in Banning's US Patent No. 2,755,334 and referenced by Bryce Bayer in his subsequent patent 3,971,065. Banning's was more of a color-wheel approach compared to the Bayer spatial sampling approach. RGBW CFA may have been first described by Honda et al. of Toshiba: "High Sensitivity Color CMOS Image Sensor with WRGB Color Filter Array and Color Separation Process Using Edge Detection"
ReplyDeletein the 2007 IISW.
Does anyone have more details on OV660 (especially power consumption numbers if used together with 10MP sensor at 30fps)?
ReplyDeleteAmazing! Googling OV660, I get Louis Vuitton belts, Ladies boots, door knobs , Harley Davidson muffler and Turkish number plates on the first page. The details are very well hidden!
ReplyDeleteHmm, not a single muffler in my search. Apparently Google take into account your recent search history. Thanks for sharing your real interests :-)
DeleteThat's a real image-sensor nerd!
DeleteI would have searched for an exhaust (not a muffler), and have never driven a motorbike - or looked for clothes on the internet. Google does have a mind of it's own sometimes :)
DeleteJust checked Phonearena picture comparison.
ReplyDeleteIs it just me or OV RGBC looks far inferior to everything else.
The picture is more noisy and there seems to be some color artifiacts in the dark area of the bottom left corner.
Is this suppose to be state of the art?
This is not a fair comparison of the sensors' performance. As one Phonearena comment pointed out, the Moto X only gets 1/3 the integration time compared to iPhone in the first picture. That should explain the higher noise.
ReplyDelete