Topics

Noise, Noise & Noise

Paul Curtis
 

On 6 Feb 2018, at 09:25, Colin Elves <colin@...> wrote:
Are we reinventing the wheel a bit here? Can’t RMS Noise and SNR be used to put an objective number on this?
Of course we are! That's what boys do, never follow the past...

Did you get those files. As i said i can do the highlight clip stuff today.

Seems you're investigating

1. DR Recovery in ProRes vs R3D
2. Noise across sensor crops

So for 1 then DR Recovery is very dependent on the tone curves and rolloff settings in RedCine, so i can shoot a test in ProRes and R3D and then produce a similar ProRes from that R3D to compare to the camera internal one? I suspect the R3D route would be substantially better.

On two i would imagine that the noise in the R3D would be very similar and really it's a debayer option how we treat it. I really doubt there is anything temporal going on though. The difference is that 8K -> 4K halves the size of the noise vs 4K straight capture.

Make sense?

cheers
Paul

Paul Curtis
 


On 6 Feb 2018, at 09:36, Paul Curtis <paul@...> wrote:
id you get those files. As i said i can do the highlight clip stuff today.

Seems you're investigating

1. DR Recovery in ProRes vs R3D
2. Noise across sensor crops
Sorry everyone, clearly that was meant for Colin directly...

Not enough coffee this morning...

Paul

Paul Curtis, VFX & Post | Canterbury, UK