Depth of Field - measure from where?

[michael_fox]

This seems to be a stupid question but I can't find the answer

anywhere.

 

When measuring such things as hyperfocal distance or near/far

distances related to depth of field, where is it measured from?

(The lens nodal point, the film plane...?)

 

Thanks,

Michael

[neal_shields]

I think that this link will answer your question and even if it doesn't you will know so much about DOF when you finish reading it, you head will be too full to hold anything else.

 

http://www.trenholm.org/hmmerk/HMArtls.html

[stephen hazelton]

I think should be from lens nodal point, but in reality the distances are approximate enough that it doesn't really matter.

[michael_briggs2]

Hyperfocal and near/far distances are object distances, and so should be measured from the front principle plane (aka nodal) of the lens. But really depth of field quantities aren't precisely enough defined for this to matter. The calculation of depth of field is based upon the diameter of the circle of confusion, which is the largest blurring that an observer considers acceptable. This is a subjective opinion and can't be defined to 0.01%, so where on the lens to measure the hyperfocal distance isn't really important.

[michael_fox]

Thanks Michael. I hear you RE where on the lens not being significant. I guess my concern was when using a long, non-telephoto lens and a close subject, the bellows can extended far (could be 24" or so) and, therefore, depending on how close the subject is, could be a significant percentage of the distance in question. So lens nodal point vs. film plane could be a significant difference.

[dan_fromm2]

Um, er, ah, Michael Briggs, isn't it possible to calculate depth of field about the plane of best focus? And wouldn't this address Michael Fox' concern about DoF in closeup work? I mean, he'll know where the plane of best focus is, won't he?

 

Cheers,

[michael_briggs2]

The original question was for "hyperfocal distance or near/far distances". So if the best focus is 20 meters, the near focus might be 19 meters and the far focus 22 meters (completely made up numbers, not from equations). These are object distances. If you want to do depth of field as deltas, as differences from the distance of best focus, then of couse the deltas are measured from the best focus position, and in this example would be 1 meter and 2 meters. Whichever way is easiest....

[timarmes]

They are actually measured from the film plane.

 

The link that Neil posted needs to be read...

 

Tim

[leonard_evens]

As others have pointed out, for distant objects (beyond 10 times the focal length), it doesn't really matter from where you measure since the the errors are well within the margin of error for the formulas.

 

If you are talking about close-ups, you should not be using distance to the subject. It is better to use magnification, which is defined as the ratio of image size to subject size. Often the image size can be measured directly on the ground glass, and then compared to the subject size to find the magnification. Alternately, you can use the following method. Focus on infinity and note the position of the standard on the rail or bed. Then focus where you want your plane of exact focus to be, and measure the distance along the rail or bed from the position of infinity focus. Divide that distance by the focal length to get the magnification. (Of course, use the same units for both.) This method avoids having to worry about the position of the nodal point for a lens of telephoto design.

 

Once you have the magnification, use the following formua

 

F-number times coc divided by square of magnification and all this multiplied by one plus the magnification. (Nc/M^2)(1+M)

 

This will give you the distance in adequate focus on either side of the plane of exact focus. It is an approximation, but it is close enough for practical work. It is also independent of focal length.

 

Let me do an example. Suppose the magnification ratio is 1:2 so the magnification is 0.5. Suppose you use a coc of diameter 0.1 mm (a common choice for 4 x 5). Suppose your aperture is f/22. then

the formula gives

 

(22 x 0.1/0.5^2) x (1 + .5) = (2.2/0.25) x 1.5 = 13.2

 

So 13.2 mm will be in focus on either side of the plane of exact focus. The actual distance of the plane of exact focus from the front nodal point in this case will be three times the focal length, but you don't need to know that.

[michael_briggs2]

Tim, I don't know where you get from Merklinger's work that DOF distances should be measured from the film plane. Look at Figure 2 of "DOF of Field Revisited", http://www.trenholm.org/hmmerk/DOFR.html. He shows D measured from the lens and the depth of field E as deltas about D. This figure doesn't answer Michael's original question because it is a schematic with a simple (one element) lens. (Merlinger's approach is not the traditional one, and will give different results than the usual equations.)

 

For a detailed and careful derivation of the traditional equations, see Applied Photographic Optics, by Sidney F. Ray, section 22.2 of the third edition. It's more complicated than any of us were saying. He measures the focused distance from the front principal plane, but the near and far distances of acceptable focus from the entrance pupil. This leads to rather complicated equations, so usually one makes the approximation that the pupils and principal planes have the same location, which results in the usual equations for the near and far distances. Under this approximation, the near and far distances are measured from the front principal plane.

[dan_fromm2]

Michael Briggs, last night I took a look in Cox and Al Blaker's Applied Depth of Field. I was wrong; in the formulae presented DoF and focused distance are measured from the front principal plane.

 

That said, except for closeup work, in practice one can measure from anywhere convenient in the general vicinity of the camera. In practice, as opposed to on paper, everything is pretty, um, approximate.

 

Regards,

[michael_fox]

Thanks all, this has been informative.

 

Can you all provide your recommendations for a "reading list" of references that cover related subjects such as depth of field, hyperfocal, light fall-off from bellow extension, etc.

[michael_briggs2]

Applied Photographic Optics by Sidney F. Ray is a superb book, covering a broad range of optical and lens topics. It uses equations at the level of algebra and geometry. If this is the level you want, I recommend the book highly. Probably most photographers will find the book excessive. It is expensive.

[dan_fromm2]

Michael Fox, if you want a book that's not too expensive and very accessible, try A. A. Blaker's book Field Photography. Long out of print, available use via, in alphabetical order, www.abebooks.com, www.addall.com, www.amazon.com, and, sometimes, eBay. I think its the best basic text on photography around, few other people agree with me.

 

Cheers,

[timarmes]

Michael,

 

The link in question doesn't discuss this. However, his book, the INS and OUTs of FOCUS does (available by PDF if you follow some links from the link in question).

 

It really depends what the origianl poster is asking. I'd asumed that he was talking about the DOF scales on lenses, but I conceed that this may not be the case.

 

At several points the book above mentions the discrepency between the distances as mesured from the lens from nodal point and those marked on the lens. For example:

 

"One more correction: remember the distances we have been working in are measured from the front of the lens whereas the standard distances shown on camera lenses are measured from the film."

 

So, it all depands on what you're using.

 

Tim

[michael_briggs2]

Tim, this question was asked in the LF Photography Forum, so I wasn't thinking in terms of DOF scales on lenses, which LF lenses don't have. The equations for DOF that are typically derived in books have the distance measured from the principal plane.

 

Manufacturers of small cameras frequently mark the film plane on their cameras and refer distances to this mark. To do this for a depth-of-field table, they will have to take the usual DOF equations for near or far distance, then add the image distance. For full accuracy, they would have to add in the nodal point separation. positive or negative. This last step would make the result specific to a particular lens. The nodal point correction would rarely matter, only making a difference for closeups, and for telephotos and retro-focus lenses.

[michael_fox]

Michael Briggs was correct - the question was with respect to LF lenses.

 

Thanks for everyone's contributions. I've ordered the books mentioned above and plan to curl up next to the fireplace on a cold night this fall for an "exciting" night of reading. :-) ZZZZZZZZ!

 

Michael