Grainy night photos are the flaw no phone launch ever seems to fix, so a potential solution that starts with a 3D printer got my attention.
Researchers at Singapore’s University of Technology and Design, along with partners in China and Japan, have 3D-printed tiny structures called color routers. Think of them as microscopic traffic directors for light. Instead of filtering light and throwing much of it away, they guide red, green and blue light to the right spots on a sensor (via Nature Communications).
How does a color router beat a camera filter?
Almost every camera sensor sits behind a Bayer filter. It comprises of a grid of red, green and blue filters that lets one color through at each pixel while blocking the others. In the process, only 30% of the incoming light makes it through, while the rest goes to waste. This is one of the reasons why photos taken in dim conditions can look grainy, blur easily or require longer exposures.
A color router takes a different approach altogether. Rather than blocking unwanted colors, it redirects light in a way so that more of it reaches the sensor. Each tiny unit contains four precisely shaped pillars made from resin.
Researchers print them using called two-photon polymerization, a technique which uses a focused laser to harden liquid resin into incredibly tiny structures. The four pillars work together to steer red, green and blue light toward their respective areas in the sensor.

Earlier color routers relied on materials such as titanium dioxide. While they could bend light strongly, it was difficult to manufacture them with high precision. However, to tackle this, the team took a rather clever approach. They used ordinary resin (which bends light less effectively) and made up for it with carefully designed pillar shapes and dimensions.
How much better is it, and when could my phone get one?
On a patch just one millimetre wide, the printed routers directed about 70% of incoming light to the intended sensing areas in real-world tests (40% more than a typical color filter).
Real-world camera tests are also promising. The router-based camera setup achieved similar brightness with an exposure of about four milliseconds instead of 10. This could actually help reduce motion blur in dark or poorly-lit scenes. The color reproduction was also closer to reality.

However, at the end, this is still a lab experiment. It was conducted with a monochrome sensor rather than a complete smartphone camera. Further, there’s some color leakage as well, wherein about 25% of the blue light reaches the red channel.
Lead researcher Joel Yang says the design needs to be adapted to materials and manufacturing processes that are already used by chipmakers. The next step is to integrate a color router directly onto an actual image sensor and test how well it performs.
In my experience, getting from a promising lab experiment to a feature in your phone can take years. So enjoy the idea, but don’t put off that upgrade just yet.
RELATED COVERAGE:






