help - confused by dof v. sensor size

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Gunn
Posts: 10
Joined: Mon Aug 07, 2006 9:42 pm
Location: adelaide, south australia

help - confused by dof v. sensor size

Post by Gunn »

Hi guys

I chanced upon this depth of field 'calculator' and, having given it a go a couple of times, am now utterly confused by it. http://www.dofmaster.com/dofjs.html

You see, I've always been labouring under the impression that bigger image sensors, ceteris paribus, give lesser dof than smaller ones do. For example, the 1dsm2's full-frame sensor will give us lesser dof than would the 1.3x sensor of a 1dm2n, which in turn, would give us lesser dof than that given by the 1.6x sensor of a 20D. Assuming, of course, the focal length of the lens, the aperture employed and the camera-to-subject distance all remain the same. Say the ef-100 macro at f.8 and at 24 inches from the subject. (For these, the calculator gave 0.59 inches dof for the 5D and only 0.37 inches dof for the 20D.) Shouldn't the 20D produce more dof and, therefore, lesser background-blur than the 5D given the circumstances ?

I was trying to decide, in the context of getting MORE background-blur and macro-subject isolation, whether a new 5D would do better than my ageing 20D. I thought the bigger sensor size of the 5D would 'produce' less dof and, hence, more background blur. Apparently not. The website's calculator gave me exactly the opposite results. I'm totally confused right now, having had my long-held belief demolished so swiftly and unexpectedly!

Charles Krebs
Posts: 5865
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

The key is that a different COC (circle of confusion) is used for the 5D (0.03) and the 20D (0.019) calculations. This would be based on the contention that the larger the "format" (film or sensor) the less enlargement is needed to reach a given overall viewing magnification or print size.

For example, check out how the COC changes as you change formats, and notice how large it gets when you use formats like 4x5" or 8x10" film. The value you use for COC will have a dramatic effect on DOF calculations.

In practical close-up work, it will be easier to get "MORE background-blur and macro-subject isolation" with a 5D. This is because when photographing the same subject, the required magnification to "fill the frame" will be greater, and the higher the magnification the lower the DOF at the same effective f-number.

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

I have a feeling this discussion is going to get complicated, but I'll try to keep it simple as long as I can.

The simplest way I know to think about DOF is using the "outside the box" analysis promoted by Dick Lyon. See http://www.dicklyon.com/tech/Photograph ... d-Lyon.pdf for details. There is a section titled "Different Formats", starting at the bottom of page 7, that discusses your concerns. One particularly relevant sentence says:
Notice that changing to a smaller format can make your depth of field higher, lower, or unchanged, depending on what you keep fixed.
Quickly summarizing "outside the box" theory: If the total magnification is fixed, then DOF is determined entirely by the angular diameter of the aperture, as seen by the subject.

Sensor size by itself makes no difference. However, larger sensors go along with longer lenses, which have larger linear diameters at the same f-number. If (and only if) that larger linear diameter turns into a larger angular diameter, as seen by the subject, then (and only then) you get shallower DOF.

What's written in Gunn's post is this: "Assuming, of course, the focal length of the lens, the aperture employed and the camera-to-subject distance all remain the same."

Under those assumptions, and also assuming the same total magnification, DOF will be exactly the same. To analyze via outside-the-box, just notice that from the subject's standpoint, nothing has changed -- you're using an aperture of the same linear diameter at the same distance, so the subject sees the aperture subtending the same angle. Same angle aperture, same magnification, gives same DOF.

Classic DOF calculations give the same result if you work the details correctly, but it's easy to get them wrong. Let's walk through this case. Since you're using the same lens at the same distance, the image of the subject on the sensor will be the same size. But the 5D's sensor is bigger, so the 5D is seeing a bigger field. If you enlarge both images to the same size, full-frame, the subject will be smaller in the image from the 5D. That corresponds to a smaller total magnification, which produces larger DOF, as shown by DOFmaster. To make the subject be the same size in the final images, you have to enlarge the 5D's full-frame image to be larger than the 20D's, in proportion to the sensor size, and crop the subject out of the middle. When you do that, and adjust COC at the sensor to be constant in terms of the final image (which in this case also means the same COC at the sensor!), then DOF becomes the same for both sensors.

Now, let the assumptions change a bit. Since the 5D has a bigger sensor, you need to get closer to fill the frame with the same subject using the same lens. So do that. Keep the aperture setting constant, and as you move closer, the subject sees the aperture getting bigger. So in this case, larger sensor gives bigger aperture, gives less DOF. If you plug the shorter distance and the appropriate COC into DOFmaster, you get the same result. (The appropriate COC in this case is the one that DOFmaster wants to use, which is larger than for the 20D.)

If you really want minimum DOF, and you're working at low magnifications, then larger sensors win hands-down because they permit bigger apertures as seen by the subject.

If you go for maximum DOF, by stopping down until diffraction just starts to be a problem, then it turns out that sensor size has close to no effect except that you have to set different f-stops depending on the sensor size. See http://www.photomacrography1.net/forum/ ... 6&start=27 and surrounding posts for more discussion of that.

Everything that Charlie says is right, but you need to be careful exactly what the words mean. The magnification that he's talking about is only onto the sensor. That's different from the "total magnification" that relates to showing the same subject, the same size, in the final images.

Great confusion surrounds DOF. :?

"Outside the box" analysis helps a lot to clear things up. :idea: :D

I hope this posting helps a bit too.

--Rik

Epidic
Posts: 137
Joined: Fri Aug 04, 2006 10:06 pm
Location: Maine

Post by Epidic »

To add to the two excellent posts above, there may not be a great perceived difference in DOF between the 35mm and the APS-C size sensors. But as Rik pointed out, there are lots of variables that can affect the results so you need to think about what is happening to judge particular results. But sensor size is a real factor.
Will

Post Reply Previous topicNext topic