>>3545445[cont.]
Sorry anon, meant to quote
>>3545219 .
Anyway, from the above formula it's absolutely clear that for the same generation (=same electronics= same σ^2_read and same σ^2_dsnu and same i_dc), the only way to increase SNR is either:
1. increase t_int, i.e. exposure time
2. increase i_ph, i.e. photocurrent, i.e. photons/second captured, i.e. photosite size.
All this explanation to state the obvious, that you don't get better SNR just because you have a bigger sensor, if you don't increase the photosite size (thus sacrificing the resolution advantage). Or alternatively, if you didn't increase the photosite size, smartly downscale the image from the bigger sensor by averaging nearby pixels, simulating the effect of larger photosites. Again, sacrificing the resolution advantage.
This has been completely obvious for centuries.
Using a an 6x7 frame of medium format film (assume it's 4x the size of a 35mm frame), you can print at *either* 4 times the area at the same granularity as 35mm film, *or* at the same area at better granularity than 35mm, but *not both*.
The bigger piece of film doesn't magically give you better SNR.
If I intend to keep the resolution advantage, i.e. print 4 times larger, i.e. use the same magnification on the enlarger, the granularity of the final print will be the SAME.
Not magically better just because "used bigger film, captured more light lel".
Don't believe me?
Here's a free SNR calculator by a manufacturer of industrial camera for machine vision, where SNR plays a huge role:
>https://camera.hamamatsu.com/jp/en/technical_guides/relative_snr/index.htmlDo you see any area in picrelated to input total sensor size? Cause I sure as fuck don't.
>BIGGER SENSOR = BETTER SNRYeah mate, take your misleading broscience shit and shove it, cunt.
