The new type of image sensor is called a SPAD (single photon avalanche diode), according to a report from Nikkei Asia. It differs from a tradition sensor in that instead of examining the number of photons each pixel in the sensor can detect over a period of time, it’s able to multiply the effects of sensing just a single photon, which is why it’s referred to as an “avalanche.” As Gizmodo notes, this cascading effect allows the sensor to amplify the effects of the detected photos, giving it over 10 times the sensitivity to light than a traditional sensor, making it possible to truly “see in the dark.”
This illustration shows the SPAD sensor’s cascading ability to increase accuracy of light information received by the sensor. (Image: Canon)
Though SPAD technology has been around for quite some time, the big deal with Canon’s announcement is its new sensor is 3.2 Megapixels, which makes it the world’s densest of its kind, and more than three times denser than the company’s previous version, which it announced in June of this year. Also, not only does it have enhanced light sensitivity, it can also recognize objects by measuring how long light particles take to reach the sensor, and it can do that in full-color too. According to Canon, “The SPAD sensor measures the distance to an object based on the time it takes for the reflected light from the object to return, and can capture space in three dimensions.” This means it can “see” an environment in three dimensions, even in super low light and in full-color, which is why it’s such a good fit for security cameras, and self-driving cars.
This marks a significant advancement over previous methods for low-light photography and video, which required just taking long exposures to allow enough light to reach the sensor or using Infrared, neither of which are suitable for rapid analysis of images or detailed video. Taking long exposures is just too time-consuming for anything other than still photography, and Infrared only provides high-contrast images without detailed color information, making it great for Navy Seals trying to see bodies or buildings in darkness, but less so for security cameras trying to capture a video feed of what’s happening in a certain location in real-time.
As if all that wasn’t enough, the SPAD sensor has one more trick up its sleeve: a super-high frame rate. Unlike CMOS sensors that have a rolling shutter that activates the pixels in the sensor one at a time, the SPAD has a “global shutter” that can turn them all on at the same time, allowing it to reduce exposure time to 3.8 nanoseconds, which are billionths of a second. This allows it to have an astonishing frame rate of 24,000 frames per second, granting it the ability to capture events that occur almost instantaneously. In its PR for the previous version of this sensor, Canon demonstrated this capability by capturing beams of light in slow motion traveling through smoke.
As we stated in the intro, this groundbreaking technology won’t be making its way to your iPhone or DSLR anytime soon, but as Nikkei Asia notes both Panasonic and Sony are working on similar technologies, primarily for self-driving cars and robots.
Now Read:
- Canon R5 / R6 Cameras Overheat More Quickly Than Shutterbugs May Like
- New Android-Powered Mirrorless Camera Uses Canon Lenses
- Canon Targets Nikon, Sony With $2,300 Mirrorless Full-Frame EOS R
No comments:
Post a Comment