The Ultimate Depth of Field Skinny

[Michael Chen]

<<<-- Originally posted by Jeff Donald : It depends on how important the object size is. In their examples, notice how the size of the persons head increases in size. If you can work with your subject size being larger, then changing the focal length will decrease DOF. If the subject size must remain the same size (News Anchors head size) then zooming in (decreasing DOF) will be offset by changing distance to subject. The key is if the subject must remain the same size or not.

DV Creators is correct in a limited way. They do not explain the consequences of changing subject size and the Law of Reciprocity. They are repeating the same misconceptions most people have with DOF. -->>>


So , from what I understand, is this. To achieve a shallow DoF, we either zoom in, or we put the camcorder closer to the subject. Both will have the same result as long as the object size remain constant rite?

Which also means, the bigger the object size, the shallower the DoF since both method, zooming in and putting the camcorder closer will increase the object size, rite?

[Wayne Orr]

I don't think you are being argumentative John, but I do think you are in error, or perhaps making a semantic argument that went way over my head.

In regards to chip size and depth of field, may I offer the following from tvtechnology.com:


The Elusive Film Look by Randy Hoffner 4/7/03

http://www.tvtechnology.com/features/Tech-Corner/f_rf_technology_corner.shtml


"You might recall that the depth of field is the range of object distances from the camera lens over which everything appears to be in focus. As lens aperture or f number decreases, depth of field increases, and, if the f number is held constant, as the size of the image decreases, depth of field also increases. The result is that for the same f number, a 2/3-inch CCD camera will have about 2.2 times the depth of field that a 35mm camera will have. As many 35mm movie lenses are designed to function optimally at maximum aperture, which may be as much as f/1.6, there is no chance that a 2/3-inch video camera will achieve the same depth of field as a movie camera. So while it is technically possible to use a 35mm movie lens with a 2/3-inch video camera, the angle of view and depth of field will not mimic 35mm film at all. One solution is to mount the 35mm lens so that its image is generated on a ground-glass screen outside the camera (increasing the image size to that of a 35mm film frame), and using relay optics to route the image to the CCD sensors. This would increase the size, complexity and price of the video camera system."


If a 2/3inch video camera has 2.2 times the depth of field of a 35mm camera, how much greater is the depth of field of a 1/3 or 1/4 inch camera?

Additionally I disagree with your contention that distance to subject affects depth of field. Only in so far as you may use a longer focal length for subjects that are farther away, as in the case of Charles' duck, er, goose. Only focal length and aperture affect dof.

In comparing the chart from the super 8 site with the Panavision guide, I find them quite compatible. (Using the 1/4 inch video on the Panavision) No conflict here.

This is an interesting discussion, but I am afraid it only serves to add to the confusion for the newbies, who are probably figuring, "Hey, if these guys can't agree, there must be some way to get blurred backgrounds for my movie with my GL1 (or whatever)." But unfortunately there is not, unless you want to shoot your entire movie outside at the longest focal length you have with a wide open aperture.

[Ken Tanaka]

Wayne: This is an interesting discussion, but I am afraid it only serves to add to the confusion for the newbies, who are probably figuring, "Hey, if these guys can't agree, there must be some way to get blurred backgrounds for my movie with my GL1 (or whatever)." But unfortunately there is not, unless you want to shoot your entire movie outside at the longest focal length you have with a wide open aperture.

Indeed. Well I think there is one statement we can all agree on: getting shallow DOF with video cameras, even big ones, is hard without making practical and/or compositional compromises.

And sometimes you have to step in goose poop.

[Jeff Donald]

The formula for determining DOF must have distance from camera to subject. The Panavision site requires it, as well as any other DOF chart or scale. The formula from my article is directly from the eighth edition of the American Cinematographers Manual, pages 698 and 699. Without the distance from camera to subject you can only determine hyperfocal distance (which is required for DOF). Let's not reinvent physics here, it's required, end of subject.

DOF does not always extend 1/3 in front and 2/3 behind the subject. It varies with the hyperfocal distance and many times is 1/2 and 1/2.

There are five factors that effect DOF and not three. They are stated in my article.

CCD size or target size effects DOF as stated in my article. If your subject must stay a certain size (News anchors head on TV) then CCD (or negative size) will effect DOF. Why? Because you must change one of the variables to keep target size the same. When you change one of the variables, you change DOF, end of subject.

If you have questions, read the article. If there is something you don't understand, post back. All questions are welcome.

[Jason Balich]

I have a huge page to print out now. THis weekend i'll give it a go with the new GL2. I'll get back to you all with what works and what doesn't. Thank-you all for the insight ( no pun intended ). I love this forum. Everyone is so into filming here!

Chat soon

[Wayne Orr]

I don't quite know how to respond to Jeff's last post, as he seems to be getting a bit tite about all this. I have visited his tutorial, and while there is nothing there to rabidly disagree with, I feel that it is way more information than is necessary to resolve the dilemma of depth of field. For instance, Circle of Confusion is a rather arcane subject that is not often brought up when calculating depth of field on the set. CoC varies with different lenses of similar focal lengths, from different manufacturers. It really is not going to affect depth of field calculations of a fixed lens on a mini dv camcorder.

Frame rate will also affect dof, but that is hardly an issue for video cameras, except the DVX100, and once you set it, it becomes a constant.

Filtration will affect depth of field, if you use an ND. But this is not mentioned in most discussions of dof, because what you are really doing is changing the aperture.

Distance to subject must be known to calculate dof, but then it is constant. If you move to a different distance to the subject, you must re-calculate the dof.

1) The longer the focal length, the narrower the depth of field. The wider the focal length, the greater the depth of field.
2)The wider the aperture setting, the narrower the depth of field. The narrower the aperture setting, the greater the depth of field.

Here are a few pictures to illustrate these points: (Charles Newcomb, this duck's for you)

http://www.digitalprods.com/Ducks.jpg

Here are some publicity comments on the unveiling of the Arricam:
"Camera assistants will become “system’s managers” as they master the Lens Data System (LDS) that “shows relevant information from the lens in use, such as focus and iris settings, the focal length, as well as the resulting depth of field, all on a convenient display.”

Note that "relevant information" is; focus (distance to subject) plus iris setting plus focal length equal depth of field. No need for a long discussion, or, as Jeff likes to put it, "end of subject."

Please go to the following link for more excellent information on depth of field:
http://octopus.drama.bris.ac.uk/Onli.../FocusPulling/

Finally, in no way is my intention to denigrate the hard work Jeff Donald has put in on his tutorial on depth of field, but I do feel that for the person who may be new to this concept, and is trying to apply it in a practical situation, that maybe he has gone a bit overboard with the information.

In the Army, they say there are two kinds of information; Need to know, and nice to know. It is my humble opinion that in the discussion of depth of field, items 1) and 2) above are "need to know" and all else is "nice to know."

[Jeff Donald]

Here are some publicity comments on the unveiling of the Arricam:
"Camera assistants will become ?system?s managers? as they master the Lens Data System (LDS) that ?shows relevant information from the lens in use, such as focus and iris settings, the focal length, as well as the resulting depth of field, all on a convenient display.?

Note that "relevant information" is; focus (distance to subject) plus iris setting plus focal length equal depth of field. No need for a long discussion, or, as Jeff likes to put it, "end of subject."

Do you have a link to the article, I'd like to read it? I suspect it does not show Coc because it would be fixed data and would not change, why clutter the display with information that would not change.

Circle of Confusion
The diameter of a circle formed by a lens imaging a true point. The largest circle which the eye will perceive as a point, without producing perceptible unsharpness (lack of focus). Is the primary factor in determining image sharpness to the viewer.

As the aperture is stopped down, the Coc is reduced in diameter. However, spherical aberration causes the plane of sharpest focus to shift or move along the optical axis, toward the CCD.

The smallest area of intersection for crossing light rays produces not a point, but a circle. This is the origin of the term Circle of Confusion. If the circle has a diameter of 1/3000 of the viewing distance, or less, it is perceived by the eye to be a point (therefore sharp and in focus). As the viewing distance increases, physically larger diameter circles are perceived by the eye to still be a point.

The diameter of the Coc is not a fixed number, nor is it's size agreed upon as an industry standard. The 1/3000 figure is the most commonly accepted. It translates to a size of .01 inches or .25mm. For many applications 1/300 of an inch is the standard. The bottom line is sharpness, or the lack there of, is in the eye of the beholder. What appears sharp to me may not appear sharp to you, or vice versa.

Why do I dwell on accuracy in discussing things of a technical nature? Well, to put it quite simply, I teach photography. Each term I have students take my classes that want to reinvent DOF or Coc. They have their own limited knowledge and personal experience with trying to deal with it and learn it. There are several threads here where members have tried to invent the formula for DOF from their experimentation around the kitchen table. It doesn't work that way. We need to have agreed upon standards and definitions so that an intelligent discussion can take place.

What happens if people only have the information they "need to know." Well, in science they make false and/or inaccurate assumptions. Complete and accurate information is required for intelligent discussion. I don't think the army wants to encourage too much intelligent discussion.

But this information might be too hard for me? I doubt it. I've had students from 18 to 80 take my classes and have no problem understanding DOF or Coc. The biggest problem is many of my students have "learned" misinformation or inaccurate information about DOF. They've been to web sites, read articles, talked to so-called experts and been misinformed or received half truths. I feel it is our responsibility to see that does not occur at DV Info.

[Wayne Orr]

<<<-- Originally posted by Jeff Donald : Do you have a link to the article, I'd like to read it? I suspect it does not show Coc because it would be fixed data and would not change, why clutter the display with information that would not change. -->>>

http://www.cameraguild.com/interviews/chat_alsobrook/alsobrook_machines5.htm

Gee, Jeff, that's kind of my point. Why get into discussions about CoC when it really does not enter into determing depth of field for a camera with a built in lens? And even with a camera with interchangeable lenses, such as the Canon XL1s, how would the user determine the CoC and apply that information to his set-up? Coc is of interest when your choice is between a Cooke Prime and your Panavision Zoom, but I have a hard time seeing the value to a quarter inch chip camera, and just how does this apply to Jason's original question:

"I have been trying to get the camera to focus on people a few feet away and then blur the background.

What is the trick to this with the GL2?"

Again, my point is that narrow depth of field is impossible to achieve with small chip cameras under normal conditions, so rather than fighting it, work with it and use other methods to give separation to your background. And very often the added depth of field will work in your favor. I hear people complain that they can't get a remote focus for their DVX100 to use when the camera is mounted on a jib. Hey, when you are shooting full wide at f/4, everything from 2 feet to infinity is in focus, so what are you worrying about focus? See, every cloud has a silver tinted lining.

[John Jay]

<<<-- Originally posted by Wayne Orr : Again, my point is that narrow depth of field is impossible to achieve with small chip cameras under normal conditions, so rather than fighting it, work with it and use other methods to give separation to your background. -->>>

I wholeheartedly disagree with this. My experience with the small camera format goes from Canon 814 & Nizo 2056 Super8, right through Video8 and Hi8 and now to DV with VX2k , 950, 900.


Follow my advice given earlier and slim DOF (blurry foreground or background ) is a cinch.

[Wayne Orr]

Then John, I am sure you will be happy to post a clip or two for our examination. Earlier you wrote:

"so using this for the background blur effect you need - arrange your camera to subject distance to be approx 1/4 (or less) of the distance of the camera to the background. Then the aperture will act as a background blur control (small f number - more blur)"

So, if we are shooting an actor that is ten feet from the lens, to get the blur effect the background must be an additional thirty feet away? That's handy.

[John Jay]

Wayne

check back tomorrow and the shots are yours its a bit dark over here at the moment

[Jeff Donald]

Wayne,

Gee, Jeff, that's kind of my point. Why get into discussions about CoC when it really does not enter into determining depth of field for a camera with a built in lens?

I didn't bring up Coc in this discussion. It was brought up by you and Jay. I brought it up in my original thread because it is the "Ultimate DOF Skinny". I need to define DOF because it had not been defined for the members here. As one of the factors determining DOF, Coc was mentioned in that article. However, since you and Jay brought it up I felt it necessary to accurately define it for members and guests not familiar with the term.

You ask why discuss Coc when it does not enter into determining DOF? It's very simple. You can not determine DOF without having a value (size) for Coc. Every chart, scale, guide, Palm program will allow you to add a value or uses a predefined one. The Panavision site uses predetermined Coc in their equations. You can not define sharpness without Coc. Most people have a curiosity about those things. If all this is old hat, I'm sorry. Gloss over my instructions, but understand, many members and guests want to learn as much as they can (I've already received 3 emails on the subject).

[Bob Harotunian]

Chris,
"Any other DoF image examples, anybody?"

I have a couple of DOF frames but need some advice on how to post them.

Thanks,
Bob

[Wayne Orr]

Jeff writes, "I didn't bring up Coc in this discussion. It was brought up by you and Jay."
Excuse me, Jeff, but the first mention of Circle of Confusion (aptly named for this discussion) was brought up by you in an earlier response to my recommendation of the Panavision site. You wrote:
"The Panavision site uses non-standard values for CoC."

A few other comments. You did not respond to my quote on the Arricam:
"...the Lens Data System (LDS) that “shows relevant information from the lens in use, such as focus and iris settings, the focal length, as well as the resulting depth of field..."

No mention of the two missing factors from your "5 factors for determining depth of field," although later you did say;
"I suspect it (the LDS) does not show Coc because it would be fixed data and would not change, why clutter the display with information that would not change."

But later you said,
"The diameter of the Coc is not a fixed number, nor is it's size agreed upon as an industry standard."
How can it be "fixed data" one moment, and then "not a fixed number, nor its size agreed upon," the next? Later, I asked, "just how does this apply to Jason's original question:" which was

"I have been trying to get the camera to focus on people a few feet away and then blur the background.

What is the trick to this with the GL2?" No reply from J.D.

You write: "You can not determine DOF without having a value (size) for Coc. Every chart, scale, guide, Palm program will allow you to add a value or uses a predefined one."

Oh? The "Kelly-Wheel" which is used by film camera assistants to determine depth of field does not list the Circle of Confusion.

One final reference:
"The depth of field in an image is controlled by three factors: the distance to the subject, the focal length, and the aperture used to capture the image."

"For users of compact digital cameras, depth of field is a subject of special interest becuase depth of field is more difficult to control with a compact digital camera than with earlier film cameras. The small imaging sensors of compact cameras require the use of short focal lengths, and this in turn gives these cameras an unusually long depth of field when compared to 35mm cameras. Thus, intentionally getting a shallow depth of field is more difficult."
http://www.megapixel.net/cgi-bin/fs_...ticle-dof.html (this is an excellent site with good graphics to explain depth of field)

For John Jay, who writes, "Follow my advice given earlier and slim DOF (blurry foreground or background ) is a cinch."

"The bad news is that it is much more difficult, using a digital camera, to blow the background out of focus, which is a pleasing effect in portrait and nature photography. You will have to use the longest possible focal length, and keep your lens wide open. Well, there is no free lunch. I'm not retiring my 35 mm SLRs yet. (2002 note: I'm lying! In the last year I went through just two rolls of film.)

In close-up photography, the greatly increased depth of field is a lifesaver. I never had so good, sharp tabletop pictures as I have now, in the digital domain."
http://www.wrotniak.net/photo/dof/
Andrzej (Un-jay) Wrotniak

Wayne Orr
Caveat Emptor