DaveW wrote:But if you increase the object distance with the same lens aren't you simply decreasing the magnification? If you blow the two images up to the same size having used the same f-number will they not both have the same DOF?
Well, no. Here's why...
Let's use "depth of field outside the box" analysis
as explained by Dick Lyon .
Think about the cone formed by all the light rays that come from a spot on the subject and pass through the aperture of the lens.
DOF depends on the thickness of that cone (its "cone angle"), and also on the total magnification (ratio of final image size to subject size).
At the same total magnification, a thinner cone (smaller angle) will give more DOF.
You can make the cone thinner by reducing the aperture diameter (stopping down).
You can also make the cone thinner by moving farther away from the subject, leaving the aperture diameter unchanged.
So, increasing the object distance with the same lens at the same f-number does
not simply decrease the magnification. It also makes the cone thinner, so you get more DOF even after enlarging the smaller sensor image to get the same total magnification.
As always, there are some awkward tradeoffs here. Backing away gives you a skinnier cone and more DOF at the same shutter speed, or even faster. But it also makes the sensor image smaller. The extra enlargement needed to keep same total magnification then exposes more noise and may also expose resolution problems due to lens aberrations and/or sensor pitch. I don't know any theory that accurately predicts which approach will produce a better picture in any particular circumstance.
Looking back at earlier posts in this topic, I see that I wrote an incorrect answer to your question the first time. Sorry about that. Hopefully this cone-angle description will help clear things up.
--Rik