I understand that an amount of GPU processing is done when an image is captured using the official pi camera boards, and that in fact most of the image processing from the board is processed via the GPU e.g. when generating a jpg output from the sensor data.
I also understand that the GPU changes between RPi versions (correct me if I'm wrong here though), presumably getting better (whatever that might mean) each time.
My question is:
Does the GPU materially effect the quality of the image output in any way, and therefore do different GPUs lead to different capture results?
Or is it that the code for GPU processing remains constant and therefore the only improvement if any is speed of image generation?
As a case in point if I took an image with a v1.3 camera on an RPi 1 and an RPi 4 under exactly the same conditions, would I/should I notice a difference on any level?
I should make it clear that I have not seen any difference, it's purely a questions born out of not understanding how the GPU takes action on the data in hardware.