The fact that the depth of field varies depending on focal length seems pretty straightforward, doesn’t it? Matt Granger however says that wide angle lenses don’t necessarily have a smaller depth of field when compared to longer telephoto lenses:
To understand any of this, you have to know what depth of field is: (Yes, this is very basic) Depth of field is basically the depth of your image that is in sharp focus, it is usually about 1/3 in front of your focus and 2/3 behind it.
In his video, Matt conducts a test to prove his point: He takes the same shot with the same framing and only changes the focal length and the position of the camera. The aperture was kept the same – f/2.8 – throughout the shoot. Of course when changing the focal length of your lens you’ll have to physically move the camera if you want your final result to have the same crop.
Here are the results:
As you can see, there’s not a lot of change in the area that is in focus, the background and the perspective change a lot though. As Matt points out the depth that is in focus is exactly the same from 85-200mm with 6.45cm with the only exception being 30mm with 6.47cm. Two millimeters however is just a very slight different that’s hardly noticeable.
Note that this only applies if you maintain the same framing (or crop), if you conducted the test without moving the camera, the results would be completely different.
[via MattGranger on YouTube, images via screencaps]
The test is invalid. He changed the camera position and aperture diameter. The f/number is a ratio of aperture diameter divided by focal length. So an f/2 10mm lens has the same DIAMETER aperture as a 100mm lens at f/20. You can only change one thing in a A/B test.
A constant f/number lens is a variable aperture diameter lens. The 1/3 and 2/3 thing only applies at a single focal distance which is different for each focal length. Take a look at any older lens with a depth of field scale on the lens.
This is wrong. Depth of field is not defined as how out of focus objects are. It is the distance from the focal plane where objects appear to be in focus.
Any depth of field calculator can easily disprove this “test.” Not only does the total depth of field change, but the distribution around the focal distance changes.
This phenomenon is less known and is true, and the two previous replies are both wrong.
Actually is a physics law, as common as the law of Ohm in electricity, just ignored by most photographers, too lazy to “zoom by the feet”.
DOF depends on aperture only, not on focal, if the framing is the same.
What change is the perspective and this fool people who don’t know the difference, being wrong perceived as DOF.
There are three things that affect DOF:
* Aperture
* Focal length
* Focus distance
The OP changed both focal length and focus distance, and then, if I understand him correctly, tries to imply that changing focal length does not change DOF. This is, of course, not strictly true, even in his special case where he also changes focus distance. There ARE differences that can be seen in DOF in these images.
I guess what he is trying to demonstrate that the effect of changing focal length and focus distance come close to canceling themselves out, which is not exactly a revelation, but is probably a decent working rule of thumb. At least with his lens at his aperture on his camera.
An even more relevant demonstration of the same thing:
http://photography.worth1000.com/tutorials/162260/theory-depth-of-field-vs-focal-length
Once again, no more look at background when you judge, DOF this is about subject.
Also… “…so that for a given magnification, DOF is essentially independent of focal length…”
http://en.wikipedia.org/wiki/Depth_of_field#DOF_vs._focal_length
In the film days, not using zooms I walk to get the desired framing, today lazy people just zoom without thinking that short focal lens are for a purpose and longer focal lens are for other… not to make the framing easier.
Fortunately, my 20 years of film still lead me, even using a zoom I always have it set at 45mm as a starting point, and zoom only according to scene, not to fill frame with the subject.