What really makes big sensors produce more appealing images? *Serious*

Started Mar 13, 2014 | Discussions thread
bobn2 Forum Pro • Posts: 56,011
Re: What really makes big sensors produce more appealing images? *Serious*

Ontario Gone wrote:

bobn2 wrote:

A pixel on a D800 is IDENTICAL to a pixel on a 16mp apsc camera. It simply has less gain and less DOF for a given scene.

There is no change in gain. A D800 pixel will have the same conversion gain as a D7000 pixel (assuming they have essentially the same design) and the voltage gain applied before the ADC will be the same. Overall the function between the number of photons counted (which should be the same at the same exposure) and the output image tone will be just the same. Most of the rest of what you said was right, but the gain thing is wrong.

-- hide signature --


I will take your word bob as i know you know your stuff, but how does this work then. Same ISO, same SS, same F stop, FF ends up with less noise.

More photons, simple as that.

We know that the actual aperture is larger for the FF than a MFT, regardless of what the F stop says. So an F2 MFT lens lets the same amount of light through as an F4 FF lens. So setting them both at F4 means the MFT camera is getting two stops light less, and if that is true, how else do they end up with equal SS besides a gain increase per ISO stop?

No need for 'gain' - you just set the processing to produce the tone values you want. That's not 'agin' it's just mapping some photon count to a grey scale. The actual 'gain' (amount of light in original image to amount of light hitting your eyes when you view the print) in the end depends on the light power of the viewing device, so what's going on in the camera isn't about 'gain' or changing the amount of 'gain'.

Im curious to this one aspect here. How does a camera that is missing 2 stops of photons get the same SS without increasing gain? There are only two real pillars of exposure, aperture and SS, gain is the "fake" third pillar since it's not actually affecting photon count. Am i wrong in this understanding?

Wrong about the nature of 'gain' - it's just using a different mapping. The point is that the noise doesn't come from the 'gain' - it's in the light (or at least in the statistics of the collection of light).

-- hide signature --


Post (hide subjects) Posted by
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow