EXIF Hyperfocal Distance etc.

[diplonics]

Hi,
  After many attempts to understand digital cameras and images I'm still unclear, so hopefully someone can help clear some of my confusion. I've been working with a Canon Powershot A85, assuming this is one of the simplest models, set on auto, to try and understand a lot of the basic image caracteristics i.e. Hyperfocal Distance, Depth of Field, Angle of View, Circle of Confusion. A sample image is linked here:



1 - I'm using the images EXIF data to acquire the parameters for the various formulas but am finding it difficult to have faith in their reliability. Using ExifTool I can get alot of detail from an image but find, in particular, the FocalPlane X and Y size calculations difficult. From what I can ascertain this camera has a sensor size of 1/2.7" which is 5.27X3.96mm according to - http://www.dpreview.com/reviews/specs/Canon/canon_a85.asp, (I also don't know how this figure is arrived at). Yet the EXIF data reports a FocalPlane size of 5.461X4.0894mm which is a conversion from the EXIF stored inches reading of 215X161. Clearly two different sets of parameters which affect all the associated formulas mentioned above. Which is correct? Should EXIF be trusted in general and should all parameters, example focal length (5.4 or 5.40625), COC (.005 or .004548295); be considered at its roundoff figures?

2 - Can an image have a Depth of Field and a Hyperfocal distance? The later is the only calculation that I can preform as the former needs a subject distance, which I don't have for any of the images I'm working with, as they are all tytpical of the one included i.e. landscapes with large panoramas. Can a Depth of Field be extrapolated without knowledge of the subjects distance, not even sure what the subject would be in the attached image? Can a Depth of Field be calculated for this image?

Any help, direction, explanations or advice would be very appreciated.
Thanks in advance.

[Bob Atkins]

EXIF data on things like sensor size is notoriously unreliable, especially when using 3rd party software to read it. While there are standards for the format of EXIF data, not everyone follows them. So if you use Canon software to read EXIF data from Canon cameras, the data should be good. However the Canon software may not read all the info in the EXIF file, so you may be foced to use some 3rd party software to try to access it. If the software is expecting format "A", but the data is in some proprietary Canon format "B", the 3rd party software may interpret it incorrectly. It's as though you asked a question in English and got a reply in French! Much of the data in EXIF files is coded, not plain text and so has to be interpreted.

My info is that the A85 does indeed have a 1/2.7" (5.27 x 3.96mm) sensor, whatever the EXIF data as interpreted by ExifTool appears to tell you. If not using Canon software, it's always possible that the interpretation of the EXIF data is wrong, especially for little used parameters.

Without a focus distance you can't calculate depth of field. To calculate hyperfocal distance you need only the focal length of the lens and the aperture (you have to pick a value for the circle of confusion though, which depends on the size of the sensor). For most small digicam sensors you can use a value of around 5 microns. COC is only an estimate based on a series of assumptions. No need for lots of decimal places.

See http://www.bobatkins.com/photography/technical/digitaldof.html
and http://www.bobatkins.com/photography/technical/depth_of_field_calc.html

[diplonics]

Thanks for the reply, really appreciate this. I'd seen your second link before but not the first, its very helpful.
One question you missed was asking about Hyperfocal Distance and Depth of Field and whether an image can have both properties. I'm assuming that Hyperfocal distance is a sub category of Depth of Field i.e. Depth of filed has a near and far property while the Hyperfocal Distance near property is half its value and the far property is infinity. Is this correct?
What I don't understand is why both Depth of Field and Hyperfocal Distance are usually shown together in tables of camera characteristics, i.e. the last table from your first link. Is this table saying that at X Focal Length etc. the Hyperfocal Distance is Y, however if the subject distance is known then your Depth of Field is near, far, etc. i.e. for a 10D image it can't have an acceptable sharpness for its Hyperfocal Distance from 3.3mtrs to infinity, and a Depth of Field from 2.85 to 20 meters?

[KeithB]

Hyperfocal distance is simply the DOF with the back end set at infinity.

It simply states that anything from that point to infinity will be in focus.