DOF and Cropping take 2

Started 10 months ago | Discussions thread
James O'Neill
Veteran MemberPosts: 3,939
Like?
If I might step in
In reply to awaldram, 10 months ago

awaldram wrote:

F2.8 - f2.8

iso 100 = iso100

OK lets see if we can get bits we agree on.

1. The bending power of the glass - it's focal length - doesn't change with depending on what is behind it, nor does the size of the aperture measured in mm, or the ratio of the aperture size to the focal length - the f/ number. We usually think of aperture as an f/ number rather than a distance.

2. There is a well known, widely used equation which states a lens of focal length d, focused at a distance d, with an f/ number of a will render a point at infinity as a circle whose diameter
c = f^2 / da . None of these terms change with sensor size either.

3. Depending on the format and any cropping applied different sizes of circle will be visible as circles rather than points. Having set the right size c for a given format the equation can be rearranged to give the distance at which the lens can be focused and still have a point at infinity look like a point : this is the hyperfocal distance and other d.o.f calculations depend on it.
H = f^2/ca  
; if you increase the sensor width you increase f and c by the same factor. Since F is squared, if you want the same d.o.f, you must increase the f/ number by the same factor. If you increase the f/ number you must use a longer shutter speed or higher ISO

4. There multiple ways to calculate ISO but they are all of the form: find the luminous flux (in Lux seconds) to give a level of exposure and multiply by a constant. For digital it's common to work from the lux seconds to get a pixel to report it's maximum value.  If two sensors have the same base ISO it means their pixels will both report "max" for the same number of photons per sq-mm hitting them. If one has smaller pixels then fewer photons will hit each one, and either the sensor must convert them to electrons more efficiently, or it must have a shallower well depth so it reports "max" with fewer pixels (but the same number per sq-mm)

etc etc

http://www.josephjamesphotography.com/equivalence/#quick

Joseph James is dangerous, because he talks about aperture diameters which people who aren't on their guard will read as "f/ number" it isn't : it is diameter of the hole in mm. He talks about "total light" which isn't illumination in lux or lux seconds but lux-seconds * area. And so on. He's not actually wrong, but people will argue about two lenses with "the same aperture" using him for support and both will understand different things by "aperture".   When we find ourselves at cross purposes with terms like "light" and "Aperture" we're in trouble !

The idea of equivalence is problematic, because you CAN say. If I put my APS-C camera on a tripod with a 50mm lens set to f/8, and expose for 1/4 second with camera set to ISO 100 that will give me the same image as a 35mm film camera loaded with ISO 100 film with a 75mm lens set to f/12 exposed for 1 second.  But the differences between film and digital rendering mean they aren't the same even if they are equivalent. By the same token when I compare two digital formats the sensor in the cameras are not identical in enough of their parameters to get the same DR, the Same noise, the same detail so it's massively unlikely that when if we can image equivalence we get the same results in the output.

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