Why is it fine to subtract average dark current, but not possible to do something similar for read noise?
jackcsullivan
How would you know the kind of pixel noise besides inspect it visually? Is the solution to always take multiple frames, average and correct by subtraction if there is still noise?
pgujjula
I wonder which kind of noise is the one associated with very high ISO levels? I mean the kind of noise you see in low-light imagery, and the one I have heard photographers refer to as "grain"
jenzou
To RichardChen9: On the previous slide 85 https://cs184.eecs.berkeley.edu/sp19/lecture/21-85/image-sensors, there is no fixed pattern for read noise, as it is thermal noise in readout circuitry. Therefore we cannot correct by subtraction and need to use cooling method.
jenzou
To pgujjula: Noise from high ISO levels in low-light imagery ("grain") is from digital noise, or "randomness caused by your camera sensor and internal electronics." "ISO has no effect whatsoever on shot noise" because shot noise is the randomness of photons and doesn't depend upon camera settings. https://photographylife.com/what-is-noise-in-photography
jenzou
To pgujjula: From https://www.cambridgeincolour.com/tutorials/image-noise.htm: "Random noise is characterized by intensity and color fluctuations above and below the actual image intensity. There will always be some random noise at any exposure length and it is most influenced by ISO speed. The pattern of random noise changes even if the exposure settings are identical." So noise from high ISO levels doesn't originate primarily from hot pixels and fixed pattern noise.
orkun1675
How is longer exposure less noisy? Wouldn't this increase lambda and this the dB of the SNR?
michellebrier
@orkun1675 I think this is because there's a sqrt(lambda) variance with lambda photons and when lambda is smaller (i.e. the photo is darker), sqrt(lambda) is closer to lambda, causing significant noise. Increasing exposure increases lambda and also increases sqrt(lambda) but the relative noisiness is decreased.
Why is it fine to subtract average dark current, but not possible to do something similar for read noise?
How would you know the kind of pixel noise besides inspect it visually? Is the solution to always take multiple frames, average and correct by subtraction if there is still noise?
I wonder which kind of noise is the one associated with very high ISO levels? I mean the kind of noise you see in low-light imagery, and the one I have heard photographers refer to as "grain"
To RichardChen9: On the previous slide 85 https://cs184.eecs.berkeley.edu/sp19/lecture/21-85/image-sensors, there is no fixed pattern for read noise, as it is thermal noise in readout circuitry. Therefore we cannot correct by subtraction and need to use cooling method.
To pgujjula: Noise from high ISO levels in low-light imagery ("grain") is from digital noise, or "randomness caused by your camera sensor and internal electronics." "ISO has no effect whatsoever on shot noise" because shot noise is the randomness of photons and doesn't depend upon camera settings. https://photographylife.com/what-is-noise-in-photography
To pgujjula: From https://www.cambridgeincolour.com/tutorials/image-noise.htm: "Random noise is characterized by intensity and color fluctuations above and below the actual image intensity. There will always be some random noise at any exposure length and it is most influenced by ISO speed. The pattern of random noise changes even if the exposure settings are identical." So noise from high ISO levels doesn't originate primarily from hot pixels and fixed pattern noise.
How is longer exposure less noisy? Wouldn't this increase lambda and this the dB of the SNR?
@orkun1675 I think this is because there's a sqrt(lambda) variance with lambda photons and when lambda is smaller (i.e. the photo is darker), sqrt(lambda) is closer to lambda, causing significant noise. Increasing exposure increases lambda and also increases sqrt(lambda) but the relative noisiness is decreased.