p.5 #1 · The main (theoretical) differences between APS-C and Full-Frame?
ruthenium wrote:
I can easily be mistaken in my thinking, but I was under the impression that "increased iso is done" during, (that is in the process of) and not "after the analog to digital conversion." Is this correct?
I could be wrong, but it looks like the iso gain is a combination of one stage of analog gain that is either on or off followed by digital gain to get the desired iso. You can certainly think of all of this process as the “analog to digital” conversion, after this it is written to the raw file.
Another possibility instead of digital multiplication is a variation in the delta-sigma adc bit depth as a function of iso. We know the bit depth of the adc is varied depending on the mode of the camera. 14 bit for uncompressed raw, I think it is 12 or 13 bit for lossy compressed raw and I believe 10 bit for video. These changes are visible in the scan times of the sensor and the modes are listed on the data sheet for the sensor.
The digital manipulation could be done in the adc and would not be observable except for a possible faster scan rate. Mathematicaly I don’t think it makes a difference. The results should be the same either way.
I don’t design cameras, but have designed other electronics. I typically get the signal to digital as fast as possible and then do additional manipulation in digital after that. My main reason for that is every analog stage will add some noise, but the digital manipulation does not.
The cameras are operating at a combination of bit depth and speed that is very tricky for electronics.
p.5 #2 · The main (theoretical) differences between APS-C and Full-Frame?
dclark wrote:
I have used the term "gain stage" which may be a bit misleading and have not said anything about the big step at ISO 500. The step change at ISO 500 is not due to amplifier gain, it is due to the reduction of the capacitance at the sense node where the photoelectrons are deposited to produce a voltage that is digitized.
Dave
Dave's explanation of the dual gain stage being a reduction in storage capacitance makes a ton of sense and would be cleaner than turning on an analog gain before adc. I would expect this to cause a slight increase in read noise as the ADC input capacitance will then be a larger fraction of the photosite storage capacitance. But as Dave mentions the photon shot noise is dominant until very high ISO. So my explanation of an analog gain stage before the ADC is incorrect.
With a 1.5X crop, IMO, the biggest advantage APS-C has is photographing small subjects like the tiny hummingbirds, which involves cropping a lot most of the time. Only a 56MP full frame sensor has 24MP in Crop Mode.
The latest software advancements are truly amazing! After Batch Processing RAW files, this is what I can get from a 9 years old 2014 A6000 without the latest, greatest anything! I prefer to show my claims - all EXIF data are intact:
