A method for noise reduction in an imaging device comprises a 4T pixel in operation, whereby the 4T pixel comprises a pinned photodiode and a floating diffusion node. The method includes the steps of: detecting a signal impinging on the 4T pixel of the imaging device and integrating the charge of the detected signal simultaneously in the photodiode potential well and the potential well of the floating diffusion node; deriving a linear signal proportional to the detected signal from the charge in the photodiode potential well; deriving a compressed signal from the charge in the potential well of the floating diffusion node, while keeping the compressed signal separate from the linear signal, and the compressed signal being a non-linear function of the detected signal; and summing the linear signal and a linearized version of the compressed signal and performing a non-linear conversion on the summation signal to fit the imaging device's output range.