In the olden days, as they say – that’s way back last century, photographers were taught to stop down the lens for best sharpness and depth of field. With 35mm cameras that was modified a little – apertures no longer went all the way to f/64. 
Smallest apertures were only f/16 on some lenses as seen in my photo here of a fifty year old lens.
Still, stopping down remained a good rule. Lenses sported aperture scales with the f-numbers, distance scales that were pretty accurate, and depth-of-field marks, as seen here, that made the relationship between aperture and depth of field pretty clear. In the photo here, the lens is set to the hyper-focal distance for f/16, about 8 feet. You can tell that the scale shows that objects from 4 feet to infinity would be sharp.
That brings us to “sharp”. The “sharpness” of objects in a photograph are always dependent on a number of different things. The resolution capability of the recording medium being just one factor. How a photo is reproduced and viewed is another. What looks tack sharp in a small print might be unacceptably fuzzy in a wall-sized enlargement.
Film for a long time was a limiting item, especially as cameras moved to smaller formats. One big problem lurking in the background always was “diffraction”. Yes, camera lenses are diffraction limited optics at the smaller aperture sizes. The aperture is that “hole” through which the light reaches the recording medium, nowadays, the sensor.
The distance from that aperture to where the light bundle converges to a point for an infinitely far object on the other side, is called the “focal length”. The focal length divided by the diameter of the aperture is called the “f-number”. The light actually never “converges to a point”. This is where the laws of physics wag a big finger. Diffraction happens, it is the cantankerous, ornery, way of light bending slightly around corners instead of proceeding in a straight line. The smaller the light bundle, the more light won’t converge to a point. We will let the physicists worry about “Airy disks”, the point here is that light is smeared out and small detail becomes fuzzy, the larger the f-number, the worse it gets.
As digital cameras get ever smaller, this problem becomes more bothersome. The individual sensor elements, “pixels” to all of us, are already smaller that what the optics can resolve. As the aperture is set to a smaller hole, the “spread” of the light increases. So the minimum aperture for some camera lenses is limited. I have seen f/11 as the smallest aperture on some cameras. Smartphones, which have absolutely tiny sensors, go to the extreme: No stopping down of the lens. They work at full aperture all the time.
I ran an interesting experiment. 
I mounted my camera firmly on a tripod a good distance from my subject. This subject was a barren tree some two hundred feet from the camera. I was careful to exclude any foreground. I turned off auto-focus and vibration reduction. Then I took a series of photos at apertures from the widest, f/5.6, to the smallest, my camera goes to f/36. The first indication of how image quality is affected can be seen from the file sizes of the images. The seven images were taken at f/5.6, f/8, f11, f/16, f/22, f/32, and f/36. I let the camera choose the shutter speed.
Notice that file size gets larger, reaches a maximum for f/11, then declines. When the raw images were translated to JPG format (at 100% quality), the differences became even more noticeable. Photo file formats reduce the file size by eliminating repeating values. The simplest way to explain this is to say that when a pixel has the same value as the previous one, instead of recording that value, a “ditto” is recorded. So if three pixels have the same value, the file says effectively “three times xxx”. As a consequence photos with less detail produce smaller file sizes since there are fewer different values. This is a gross oversimplification, of course.
Can you tell the difference in the pictures? You sure can!
Here are small cutouts from the images. These cutouts are from the JPG images since this is how most photos get distributed and viewed.
Here, in a larger view the extremes, the f/5.6 image, the f/11, and the f/36:
I was far enough from the tree so that the small twigs were narrower in the image than the pixels of the camera sensor. The first couple of photos show improvement in the image as the lens aberrations decrease as the lens is stopped down. Then, after f/11 (third image in the top row, center image of the larger views) diffraction and some other effects take over.
The moral of this story is this: For best image quality, do not go to small apertures, large f-numbers.
For my camera the deterioration in this test starts at about f/11. This is a DX format Nikon D-60. For full-size sensor cameras, the sensors, and the pixel elements are larger thus the diffraction effects will be less. For smaller cameras, especially the point-and-shoot pocket cameras, image quality will decline starting at even larger apertures.
To give you an idea of the size of the areas shown above, here is the full image with the section indicated.
Camera manufacturers either help out, or cover up, depending on your point of view. You don’t see aperture scales on lenses any more. Distance scales are only on professional lenses. Cameras are programmed to do the best they can, and this means working at the largest aperture possible, not the smallest one.
.:.
© 2011 Ludwig Keck
