Mavic Pro Native ISO and Sensor Noise

[JMcFly]

Has anyone tested the Mavic Pro's Dynamic Range or done a test to see what ISO produces the least noise. I doubt this sensors native ISO is 100, if its 400 you could use like an ND64 or 128 and bring it down to shoot at 400 producing less noise. I just can't seem to find the native ISO.

 

[Mauri59]

I doubt is 400, but much less (140?).
Nearly sure, you won't notice more noise shooting @ 100.

Has anyone tested the Mavic Pro's Dynamic Range or done a test to see what ISO produces the least noise. I doubt this sensors native ISO is 100, if its 400 you could use like an ND64 or 128 and bring it down to shoot at 400 producing less noise. I just can't seem to find the native ISO.

 

[Brojon]

Dynamic range isn't codependent on noise.
Think of the receptors as little buckets capable of holding (x) bits - that's your DR. So say we have a 12bit bucket = 12 stop DR (since each successive bit represents doubling the value - same as fStops).
in that range 0 is ultimate black 12 bits all on is brightest. Each receptor is given a finite time to charge up to the quantity of light hitting it. In low light there's simply not many photons to do the job so the charge is very low - meaning that the brightest thing in your image might only be 4 bits instead of all 12 available - meaning that you just lost 8 stops of exposure.
The ISO boost simply jacks up what minuscule charge is in that bucket - think "stretching" the bits so the brightest thing looks like the brightest thing and not something buried in mud.
There's always inherent noise in the sensor and it's being amplified along with actual signal.So now noise that was essentially at the zero point might get amplified up that 8 "missing" stops - to bring up the noise to visibility - considerably so.
Hope my analogies helped explain. I have no idea what the DR is of the Mavic sensor, nor the native ISO - but I suspect the ISO is the lowest so 100. I don't see banding in subtle gradients so I'd peg the sensor at least as a 12bit.