By eliminating fixed-pattern noise using on-chip noise reduction technology and suppressed random noise using complete pixel charge transfer technology, Canon succeeded in effectively eliminating noise. Random noise normally increases with fast signal processing. The Canon CMOS sensor therefore has unique technology that amplifies signals according to the sensitivity level generated just as they’re read. Signals with high S/N ratios are then sent to a high-speed amplifier. With this technology, low-noise sensors can function well in shooting environments of high light sensitivity and long exposure. Fixed-pattern noise comes from uneven signal boost among different pixel amplifiers. This is why the noise may appear on the same pixels even when taking pictures of different subjects at different times of the day. To remove this type of noise, Canon’s 2nd generation on-chip noise reduction circuit reads the amount of fixed-pattern noise and then removes it to produce a pure optical signal. Random noise is suppressed when the sensor resets the photodiodes that store electrical charges. Canon therefore designed the photodiode and the signal reader independently. By first transferring the residual discharge — light and noise signals — left in a photodiode to the corresponding signal reader, the Canon sensor resets the diode while reading/holding the initial noise data. After the optical signal and noise data have been read together, the initial noise data is used to remove the remaining noise from the photodiode and suppress random noise, thus enhancing signal-to-noise ratio. Page top