Through those improvements, Canon doubled the saturated amount of electric charge and widened dynamic range. Specifically, it increased the saturated amount of electric charge from 8,100 electrons with a frame rate of 120fps to 16,200 electrons with a frame rate of 60fps.
The dark temporal noise of the new sensor is 1.8 electronsrms. As a result, its dynamic range is 79dB with a frame rate of 60fps and reaches 111dB when HDR is applied.
The method of reading out electric charge accumulated several times within time for one exposure is similar to the method that Canon employed for its GS-based 4k CMOS image sensor and announced at IEDM 2016 (See related article). The technology announced at IEDM 2016 accumulated electric charge four times, but the company reduced it to two times this time.
When the number increases, the size of memory needs to be increased. So, when the pixel size is not changed, it becomes necessary to decrease the size of PD, lowering sensitivity.
The pixel size of the sensor announced at IEDM 2016 is 6.4μm, which is relatively large. However, the pixel size of the latest sensor is as small as 3.4μm. So, the new sensor accumulates electric charge only twice in the aim of preventing the memory size from increasing and the PD size from becoming too small.
Still, the PD size became smaller, lowering sensitivity. So, Canon improved sensitivity by forming a "light guide," which collects light, in front of PDs. The sensitivity (parasitic light sensitivity) of the new sensor reaches 28,000 electrons/lx·second. This is the second improvement.