Why does raising ISO make image quality appear lower even though my measurements suggest it shouldn't?

Question

I know this may be rehashing one of the most discussed topics, but I would like to take it a slightly different direction and solicit input. First of all, I am familiar with the electrical engineering and mechanisms associated with changing the camera gain, normally described in the DSLR community as ISO. I understand that it is changing the capacity of the electron well, which leads to a faster sensor for each pixel/photodiode. However, what I am trying to rationalize is theory vs. reality.

I have done due diligence with my Canon 70d and 6d and measured the read noise, thermal noise as a function of time for various stagnant ambient temperatures, measured the gain in electrons per ADU at most ISO's, determined the electron well size in photons per ADU and most importantly, made analogous measurements of the signal to noise ratio for various exposure times (in a dark frame) at various ISOs. In my measurements, as theory would predict, while the noise does increase as a function of ISO, the signal to noise ratio increases.

This would lead one to believe that for properly exposed high ISO images in post processing, while you may loose some signal effects by stacking, subtracting bias and dark-thermal frames and adjusting the luminance channel, that you will still come out with a superior image due to the fact you started out with much more signal-effect than noise. However, reality has consistently presented a different situation, where as my ISO increases the best image I can achieve after post processing is in many cases inferior to what was produced at a lower ISO, especially if shooting time-lapse where stacking and frame subtraction are much more impractical. For crop frames I have found that I can never get above ISO 1000 and for full frame cameras, I have a hard time getting above ISO 2500 without inducing granularity that looks un-natural, despite the fact I see many people shooting at 25,000 and higher. I would love to hear comments are explanations from any interested people describing why the higher signal to noise ratio rarely translates to a better image, or rebuttals to this idea if anyone has one.

Answer 1

I believe it may be a flaw in your testing. Raising ISO should not be getting you a higher signal to noise ratio (or usable dynamic range).

You are raising the noise floor, without any corresponding increase in dynamic range at the bright end. Thus you should be left with overall lower dynamic range. It would not be possible to go the other way.

Is it possible that your testing did not test full dynamic range (between noise floor and lightest point without clipping) for each ISO, or that you made false assumptions such as not compensating for the gain caused by raising ISO? For instance, taking a photo with higher ISO but with exactly the same lighting, shutter and aperture will raise the intensity of the image by nominally the same amount as the intensity of the noise which may make it appear that you have lost no detail to noise, yet at the other end it will also mean the brightest intensities that were not clipped are now clipped, because the overall exposure is higher as well.

In a real world situation when you decide to raise ISO you'll effectively then be able to use a faster shutter speed in order to equalise the exposure. When doing so, ie when keeping the final exposure the same, you'll find that you lose detail in the noise floor while preserving the same clipping cut off at the top end, resulting in a net decrease to usable dynamic range.

Answer 2

In this answer ISO, shutter speed and aperture are independent parameters.

1) Yes, setting ISO to a higher number will almost always lead to better SNR - because additional analog amplification occurs. There are exceptions to this rule, though: Canon 6D gives better SNR at ISO51200 than at ISO102400.

Say, you have 100 minutes to record some dark scene and you may do 10x10 min exposures at ISO100 or 100x1min exposures at ISO1000. Almost every available digital sensor will export better material at ISO1000 in this case.

2) Setting ISO to a higher number in S or A mode will always yield to worse SNR because less light reaches the sensor.

3) People only shoot at ISO25600 because they resize for web or use noise suppression.

4) Regarding your experience with time lapse: it is not exactly clear what are you comparing to what. If you keep the exposure/(F^2) same then there you sure will have better SNR.