Dean Hansen wrote to me today saying:
> Hi Chuck, You recently posted on the OM list: "That's OK but you
> should realize that, at f/13, your 10MP camera is producing only
> about 3MP images." Guess I know more about insects than I do about
> digital cameras. Could you explain this, or direct me to a web page
> that explains this? Many thanks! It will be interesting what Moose
> thinks of his OM-D. Best wishes, Dean
So I wrote a response explaining it as best I could to which Dean
replied that he thought it was a well written response and that I should
post that response to the OM list. He thought perhaps he wasn't the
only one who didn't fully understand what I said. So here's what I told
him:
-------------------------------------------------------------------
It's actually fairly simple. First consider a point source of light
such as a star. Despite their actual size, stars (except for the sun)
are so far away that they are essentially a mathematical point. No
telescope in the world is powerful enough to image a star as anything
but a point.
However, when an optical system attempts to image that point source of
light various abberations will conspire to make that point source into
something larger and measurable. One of those aberrations is
diffraction which is the bending of a light beam as it passes by an
edge. Diffraction is present in ALL optical systems as they all have
edges... the edges of lens cells, the edges of aperture blades, etc. The
bending of the beam distorts it and causes the image of the star to grow
larger.
Finally, the control factor for diffraction is focal ratio. The smaller
the aperture (higher the f-number) the worse the diffraction. In the
case of a film camera, once the diffraction controlled size of a point
source of light exceeds the diameter of a film grain it takes more than
one grain to capture that point. In the case of a digital camera, once
diffraction has swelled the size of that mathematical point to something
larger than a pixel it takes more than one pixel to record that point.
The smaller the size of the pixels the more prone they are to the
effects of small aperture. In the case of Jim's E-510 with its 10 MP
sensor (note: very tiny pixels on a 10MP 4/3 sensor) when the camera is
set to f/13 diffraction has swelled the size of those point sources to
where they require more than 3 pixels to record a single point source.
Therefore, although the camera has 10 MP, only 3 MP of data can actually
be resolved. On the other hand, my Canon 5D with 12.7 MP does not run
into diffraction limitations until it's past f/11 because the individual
pixels on that large sensor are much larger diameter. The same effect
is seen on 35mm film vs. medium and large format film. Large format
especially often uses very small apertures like f/32 or f/64. But the
film is very large and the final image does not have to be magnified
very much.
Now, having said all that, something out of focus is blurring the image
in a fashion somewhat like diffraction. Although diffraction might be
limiting the E-510 to a 3MP image at f/13, f/13 might be producing
greater depth of field and still make a pleasing image. Even so, the
total resolution of the final image can't exceed 3 MP.
If you'd like to dig deeper see this Luminous Landscape article. In
particular pay attention to Table 3 which lists the maximum resolution
obtainable at various apertures on different sized sensors. Note that
each sensor size has 3 columns of data: Since diffraction is dependent
not only on focal ratio but also on wave length the table gives
different resolution values for red, green and blue light (with
wavelengths specified). Blue light, having the highest frequency and
energy is the least affected by diffraction. Red the most.
The actual subject of the article is lens resolution but as limited by
diffraction and pixel dimensions. Note that the diffraction values are
theoretical for a perfect lens. Real world resolution will be somewhat
less since no lens is perfect.
<http://www.luminous-landscape.com/tutorials/resolution.shtml>
Chuck
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|