22MP - 36MP how much more resolution does it equal

[Flake]

A question because I can't work it out! I do know that you need an awful lot more pixels to double the resolution (Square of?) , and that a jump of 14MP might look impressive on paper, but I have this funny feeling that it doesn't actually translate to all that much more in terms of a percentage gain in resolution, all of course at the cost of extra noise, loss of dynamic range, and early onset of diffraction issues.

Please would some one do the math & let me know just what kind of percentage advantage 36MP is over 22MP.

 

[Mt Spokane Photography]

There is sensor resolution, which is merely the number of pixels.

Then there is the linear resolution, vertical and horizontal, which is probably a more important number.

http://www.dpreview.com/learn/?/Glossary/Digital_Imaging/Resolution_01.htm

 

[Fleetie]

The linear resolution factor you want is:

√(36/22)

~= 1.279

i.e. a 27.9% increase in linear resolution.

 

[LetTheRightLensIn]

A question because I can't work it out! I do know that you need an awful lot more pixels to double the resolution (Square of?) , and that a jump of 14MP might look impressive on paper, but I have this funny feeling that it doesn't actually translate to all that much more in terms of a percentage gain in resolution, all of course at the cost of extra noise, loss of dynamic range, and early onset of diffraction issues.

Please would some one do the math & let me know just what kind of percentage advantage 36MP is over 22MP.

as said already, sqrt(36/22) more (linear) resolution

36/22 times more resolution if you use the definition of resolution used for computer graphics/tv screens

and 36/22 times more total information/detail captured

 

[Fleetie]

And to reinforce the point: If you want double the resolution of the 22MP sensor, you need an 88MP stonker! (Neglecting other effects like diffraction, which may also start to work against you.)

 

[dtaylor]

A question because I can't work it out!

It's a 31% increase in linear resolution over the 5D2, or 27-28% over the 5D3 depending on how Canon rounded to 22 MP.

a jump of 14MP might look impressive on paper, but I have this funny feeling that it doesn't actually translate to all that much more in terms of a percentage gain in resolution,

It's the difference between printing at 24-30" (5D2; landscape; critical review) and printing at 32-40" (same conditions). If you relax the conditions you can print larger with either, but the D800 will always be able to go about 1/3rd larger assuming no other limiting factors. In some situations it may be even more than 1/3rd. If foliage at a certain distance is well rendered at 36 MP, but mushy at 21 or 22 MP, then you might be able to push the D800 image a bit more, maybe to 48" yet still remain critically sharp and detailed.

For a landscape or studio photographer, this is significant.

all of course at the cost of extra noise, loss of dynamic range, and early onset of diffraction issues.

I highly doubt D800 images will have more noise for a given view/print size. Only pixel peepers will see more noise. Likewise diffraction will not impact D800 images any more than 5D2/5D3 images for a given view/print size.

DR may be lower due to smaller pixels, but this may also prove insignificant. The difference in pixel size is much greater between the 5D2 and 7D, yet the DR gain is only one stop.

 

[Actionpix]

And little off topic, but often seen wrong stated, crop compared to FF has a factor 2.6 and not 1.6 as you get 1.6 times more pixels in width and height as well. (crop pixel count x 1.6 x 1.6= 2.56 x crop pixel count for FF) 36MP FF thus translates to 14MP crop.

 

[wrack_of_lamb]

And little off topic, but often seen wrong stated, crop compared to FF has a factor 2.6 and not 1.6 as you get 1.6 times more pixels in width and height as well. (crop pixel count x 1.6 x 1.6= 2.56 x crop pixel count for FF) 36MP FF thus translates to 14MP crop.

just to be clear that is the Canon crop factor, if you are comparing the D800 to a Canon camera. Nikon's crop factor is 1.5, so it would be a 2.25 factor. i read somewhere (not sure if true) that the D800 has a similar sensor/pixel pitch as the D7000 (16MP, or 36/2.25) and may be a scaled up FF sensor from that. if so, it will probably have incredible DR as the D7000 has a very high DR.

 

[psolberg]

A question because I can't work it out!

It's a 31% increase in linear resolution over the 5D2, or 27-28% over the 5D3 depending on how Canon rounded to 22 MP.

a jump of 14MP might look impressive on paper, but I have this funny feeling that it doesn't actually translate to all that much more in terms of a percentage gain in resolution,

It's the difference between printing at 24-30" (5D2; landscape; critical review) and printing at 32-40" (same conditions). If you relax the conditions you can print larger with either, but the D800 will always be able to go about 1/3rd larger assuming no other limiting factors. In some situations it may be even more than 1/3rd. If foliage at a certain distance is well rendered at 36 MP, but mushy at 21 or 22 MP, then you might be able to push the D800 image a bit more, maybe to 48" yet still remain critically sharp and detailed.

For a landscape or studio photographer, this is significant.

all of course at the cost of extra noise, loss of dynamic range, and early onset of diffraction issues.

I highly doubt D800 images will have more noise for a given view/print size. Only pixel peepers will see more noise. Likewise diffraction will not impact D800 images any more than 5D2/5D3 images for a given view/print size.

DR may be lower due to smaller pixels, but this may also prove insignificant. The difference in pixel size is much greater between the 5D2 and 7D, yet the DR gain is only one stop.

thanks for the excellent summary. Looking at the D800 images, there is no question in my mind that the difference is very noticeable even when if not printed that large. I expected this because the linear resolution difference of 12MP vs 21MP was similar and it was noticeable.

 

[Mt Spokane Photography]

And little off topic, but often seen wrong stated, crop compared to FF has a factor 2.6 and not 1.6 as you get 1.6 times more pixels in width and height as well. (crop pixel count x 1.6 x 1.6= 2.56 x crop pixel count for FF) 36MP FF thus translates to 14MP crop.

In digital photography, a crop factor is related to the ratio of the dimensions of a camera's imaging area compared to a reference format; most often, this term is applied to digital cameras, relative to 35 mm film format as a reference. In the case of digital cameras, the imaging device would be a digital sensor. The most commonly used definition of crop factor is the ratio of a 35 mm frame's diagonal (43.3 mm) to the diagonal of the image sensor in question; that is, CF=diag35mm / diagsensor. Given the same 3:2 aspect ratio as 35mm's 36mm x 24mm area, this is equivalent to the ratio of heights or ratio of widths; the ratio of sensor areas is the square of the crop factor.