Serena,

We are both saying the same thing.

But I just wanted to clarify that since the CoC spec is always given as a measurement on the sensor, it may be confusing to say the "same diameter of the circle of confusion" applies in both cases.

All in the effort to prevent a circle of confusion around this topic, for those who may have a more casual interest in DoF than we both clearly do!

Isn't the CoC a result of the resolution through the whole process.
For example an image from a 2/3" SD camera would produce a larger CoC than one from a 2/3" HD sensor if they were both viewed at a large viewing angle. Conversely at a small enough viewing angle there'll be no difference in the CoC as our eyesight becomes the limiting factor.

From what I've noticed also the amount of EE in video cameras also creates a differently perceived DOF compared to film. With video the image appears sharp within a certain range and then 'snaps' out of focus ouside that range.

Bob,

The CoC spec is an attempt to apply a number to the question "Does it look in focus to you?" There really isn't a threshold when things pop out of focus, just a threshold when they are perceived to be out of focus.

There are many factors that determine this. Even subject matter plays a huge role in this perception.

In 35mm film the sharpest standard generally used is .001" even though film resolution is considered to be four times as sharp as that.

The phenomenon you describe in video may exist to at the pixel level (never witnessed it myself) but generally defocusing is a gradual affair.

What you said is correct, but to rephrase it slightly, in most cases the acceptable CoC on an SD camera would be larger than the acceptable CoC on an HD camera. And the farther we get from the viewing screen, the more tolerant we are of a larger circle of confusion.

But other many factors come into play. Add some camera movement (with or without motion blur) and at larger viewing sizes I would bet you would want to see a more similar CoC spec between HD and SD.