New ams Color Sensor Designed to Eliminate Flicker in Rolling Shutter SensorsNovember 05, 2019 by Gary Elinoff
The highly sensitive flicker sensor is designed to deliver distortion-free images and videos captured on a smartphone camera.
Because of artificial lighting, photographs and videos are subject to a phenomenon known as flicker. Flicker occurs because artificial lights, like fluorescent lights, turn off and on at certain frequencies (though these on-off events are imperceptible to the human eye), resulting in a flickering effect in photos and videos.
ams recently announced a new sensor aimed to eliminate banding and other flicker byproducts from photos and videos taken on a smartphone's outward-facing camera. Specifically, the TCS3408 color sensor is designed for smartphone cameras employing rolling shutter image sensors.
Examples of blurred images from temporal light artifacts. Image used courtesy of Ethan Biery (Lutron Electronics) and Jim Gaines (Philips Lighting)
As described by David Moon, Senior Product Marketing Manager in the Integrated Optical Sensors business line at ams, “The increasing pixel count of today’s high-resolution smartphone cameras enables image capture with extraordinary detail and clarity; however, pictures and videos can often be ruined with unsightly banding distortion effects from flickering light sources. Smartphones can now capture vibrant, detailed and artifact-free images in all lighting conditions, thanks to the TCS3408 high-sensitivity flicker sensor.”
What is Global Shutter and Rolling Shutter?
A device that employs rolling shutter first samples one row of pixels then moves down to sample the next row of pixels. This finite time delay can result in distorted or blurred renderings, even in the best of conditions. A system that sports global shutter samples all the pixels of the sensing device at the same moment—regardless of location—which eliminates this problem.
Rolling shutter vs. global shutter. Image from Oxford Instruments
Because most smartphone cameras have rolling shutters, they've been prone to image distortion in the past. The TCS3408, with on-chip flicker detection, provides an accurate measurement of color and brightness thanks to five concurrent-reading, ambient light-sensing channels—a wideband and a clear reference channel in addition to the three RGB channels.
The unit can also store a sequence of flicker measurements in its internal memory. This allows the camera’s video processor to detect flicker frequencies ranging up to 2 kHz, which are typical to today’s Pulse-Width Modulated (PWM) LED lighting systems.
Rolling shutter is typically associated with a distorted banding effect or blurred renderings from flickering lights. The TCS3408 device incorporates a flicker detection engine that will detect, for example, the 60Hz flicker generated by incandescent or fluorescent lights.
Block diagram of the TCS3408. Image from ams
Upon detection, the camera’s video processor is designed to synchronize the shutter with the “on” portion of the relative ambient light output in a scene. This effectively eliminates distorted banding artifacts and produces the image that the user sees.
The TCS3408 is available in a 0.5 mm high, 2.5 by 2.0 mm package with eight active pins and four pins intended for grounding.
ams offers a complete evaluation kit for TCS3408. In addition to all relevant hardware, the kit includes drivers, documentation, and the necessary GUI software to help OEMs get their TCS3408-based products to market as quickly as possible.
Around the Industry
To combat distorted images from flicker, designers might also turn to a few other color sensors on the market.
The BU27006MUC-Z from Rohm senses the flicker noise typical of room lighting. While this device is standard grade, the company also produces specialized devices for automotive purposes.
Similarly, STMicroelectronics’ VD6281 includes a flicker engine that can work with light frequencies ranging from 50Hz to 2kHz and can output flicker information in both analog and digital formats.
What do you look for in a color sensor? Share your thoughts in the comments below.