ISO Sensitivity
Feb 19, 2019 21:59:30 #
TriX wrote:
Nikon D7200 (~24MP) Pixel pitch=3.92 microns low light ISO=2252 (9.49 stops)
Nikon D850 (~46 MP) Pixel Pitch=4.36 microns low light ISO=4115 (10.36 stops)
Difference = .87 stops (as expected)
Source: http://photonstophotos.net/Charts/PDR.htm
Nikon D850 (~46 MP) Pixel Pitch=4.36 microns low light ISO=4115 (10.36 stops)
Difference = .87 stops (as expected)
Source: http://photonstophotos.net/Charts/PDR.htm
That is what i mean by "generally" & "approximately the same size".
Full frame usually has twice the pixel size with 200% more area. But the D850 has more pixel than the average full frame, hence its pixel area is only .25%bigger. Since we are talking microns here, that is really a very small difference that as you say, it is not even a full stop away.
Feb 19, 2019 22:30:53 #
Wallen wrote:
That is what i mean by "generally" & "approximately the same size".
Full frame usually has twice the pixel size with 200% more area. But the D850 has more pixel than the average full frame, hence its pixel area is only .25%bigger. Since we are talking microns here, that is really a very small difference that as you say, it is not even a full stop away.
Full frame usually has twice the pixel size with 200% more area. But the D850 has more pixel than the average full frame, hence its pixel area is only .25%bigger. Since we are talking microns here, that is really a very small difference that as you say, it is not even a full stop away.
No disagreement on the “approximate” pixel pitch - I was actually responding to your previous post concerning a 3 stop difference between crop and FF... ( “Generally speaking, all APS-C cameras have a working limit of about ISO 12800 and full frames about 3 stops more”). In general, I think you’ll find there is typically approximately a 1 stop difference in high ISO/low light performance between the two, given a similar generation of sensors, as shown in the above example.
Feb 19, 2019 22:34:38 #
LWW wrote:
Godzilla Facepalm
You break my heart bro LOL.
The pixel size of a D850 is .0043 x .0043
The Pixel size of a D7200 is .0039 X .0039
That is a difference of .0004 X .0004
That my friend is a very very small difference, hence my comment "APPROXIMATELY the same size".
Its pixel size being slightly larger, the D850 is a tiny bit better, but never the less it is still less than a full stop better than the D7200.
Do you know whats the real facepalm? In Digital cameras, ISO does not exist. 1. Because manufacturers do not adhere to a standard and 2. Because Camera ISO in not about sensor sensitivity but the amount of gain done to the signal.
We can not change the sensitivity of the sensor. What we adjust as ISO in the camera is signal gain. In essence they can claim as much ISO as they want because they can boost the signal as much as they want. The limiting factor which i for reference reason call "working limit ISO/optimal sensitivity" Is the amount of gain we can apply which the signal to noise ratio is acceptable. Which is "generally" at present times, a setting of ISO12800.
Beyond a certain amount of boost, the signal to noise ratio becomes too much that we do not have a good image anymore.
Feb 19, 2019 23:04:22 #
TriX wrote:
No disagreement on the “approximate” pixel pitch - I was actually responding to your previous post concerning a 3 stop difference between crop and FF... ( “Generally speaking, all APS-C cameras have a working limit of about ISO 12800 and full frames about 3 stops more”). In general, I think you’ll find there is typically approximately a 1 stop difference in high ISO/low light performance between the two, given a similar generation of sensors, as shown in the above example.
You are correct technically since all changes are done in squares, on apples to apples comparison it should be a stop better. But I was speaking about a theoretical "working limit" which is based on (my own) acceptable image quality.
Feb 19, 2019 23:26:09 #
Wallen wrote:
You are correct. Technically since all changes are done in squares, on apples to apples comparison would be a stop better. I was on the other hand speaking about a theoretical "working limit" which is based on (my) acceptable image quality.
I agree. I think it’s interesting that the D850 is still almost a stop better than the D7200 in spite of similar pixel pitch. My guess is that’s due to the fact that the D850 is a newer camera with a later generation sensor. While you can find the sensor manufacturer, and occasionally the model used in a particular camera, it’s very difficult to find an actual data sheet on the device - I’ve looked.
Feb 19, 2019 23:45:56 #
TriX wrote:
I agree. I think it’s interesting that the D850 is still almost a stop better than the D7200 in spite of similar pixel pitch. My guess is that’s due to the fact that the D850 is a newer camera with a later generation sensor. While you can find the sensor manufacturer, and occasionally the model used in a particular camera, it’s very difficult to find an actual data sheet on the device - I’ve looked.
Technology is always a factor. As for the spec sheets, It may be deliberate on the manufacturers part to avoid having those specs in the open.
If you want to reply, then register here. Registration is free and your account is created instantly, so you can post right away.