Home of Professional Cinematography since 1996

>
Lens Extender/Doubler & Depth of Field

Published : 27th October 2004

>Greetings everyone,

>If I have to calculate the depth of field on a lens with an extender or double, what is my base to calculate from? Let's say a 2 x extender.

>Appreciate your replies, and thanks in advance.

>Regards
Emmanuel from Munich


>Greetings again,

>Some rephrasing.

>What I actually wanted to know is :

>Let's imagine I use an Ultraprime 50 with a 2x extender. Do I get the exact same depth of field as the Ultraprime 100mm, taking into account that I compensate two stops for the use of the 2x extender?

>Hope this makes it clearer?

>Regards
Emmanuel from Munich


>Emmanuel from Munich writes :

class="Paragraph">>What I actually wanted to know is :
>Let's imagine I use an Ultraprime 50 with a 2x extender.
>Do I get the exact same depth of field as the Ultraprime 100mm, taking >into account that I compensate two stops for the use of the 2x extender?

>If I think I know what you're asking, then the answer is yes.

>If your f/stop with the 50mm and 2X is the same as the 100mm would be without an extender and the camera to subject distance is the same, then the DOF is identical.

>The two stop compensation is only for exposure. What determines DOF in this situation, is the actual f/stop setting on the lens.

>At the risk of re-opening a can of worms, the major determinants of DOF are f/stop and image size, not focal length.

>If you make the subject the same size on the film, and use the same f/stop, the DOF will be the same no matter
what the focal length of the lens is.

>Brian Heller
IA 600 DP


>Brian Heller wrote :

class="Paragraph">>At the risk of re-opening a can of worms, the major determinants of DOF >are f/stop and image size, not focal length.

>A fact Mr. Heller reminds me of repeatedly...and of course he's right.

>Jeff Kreines


>Brian Heller writes:

class="Paragraph">>At the risk of re-opening a can of worms, the major determinants of DOF >are f/stop and image size, not focal length.

>Yes, as long as "image size" means *the absolute size of the projected image behind the lens* and not *the field of view vis-a-vis a given chip or gate size*.

>Dan "pass the can opener, will ya?" Drasin
Producer/DP
Marin County, CA


>Brian Heller wrote:

class="Paragraph">>...At the risk of re-opening a can of worms, the major determinants of >DOF are f/stop and image size, not focal length.

>Well, yes and no. It depends upon the relationship of the focused distance to the hyperfocal distance of the lens and f/stop combination being used. If the HD is more than around 10x the subject distance this is true.

>But if the HD is considerably LESS than 10x the subject distance it isn't true, and you can actually gain DOF by moving in with the shorter lens.

>For example, a 100mm lens vs. a 25mm lens :

>If the 100mm lens is at f/11 (HD = 117.42 feet) and the subject is 20 feet away, the DOF is from 17 feet, 1 inch to 24 feet, 1 inch, a total DOF of 7 feet. Switching to a 25mm lens (HD at f/11 = 7.34 feet) and moving in to 5 feet to cover the same area, the DOF is now from 3 feet to 16 feet, a total DOF of 13 feet. By switching to the shorter lens and moving in you have gained almost twice the DOF.

>In practical terms, this advantage works with relatively short lenses at near distances and small f/stops.

>Wade "the worms crawl in, the worms crawl out" Ramsey


>I wrote:

class="Paragraph">>If you make the subject the same size on the film, and use the same >f/stop, the DOF will be the same no matter what the focal length of the >lens is.

>Wade Ramsey writes:

class="Paragraph">>Well, yes and no. It depends upon the relationship of the focused >distance to the hyperfocal distance of the lens and f/stop combination >being used. If the HD is more than around 10x the subject distance this >is true.

class="Paragraph">>But if the HD is considerably LESS than 10x the subject distance it isn't >true, and you can actually gain DOF by moving in with the shorter lens.

>Only if you do not consider the subject size. If you maintain the identical subject size on the imaging plane, then the DOF will be identical.

>>For example, a 100mm lens vs. a 25mm lens : If the 100mm lens is at >f/11 (HD = 117.42 feet) and the subject is 20 feet away, the DOF is from >17 feet, 1 inch to 24 feet, 1 inch, a total DOF of 7 feet.

Of course, but this is the size or area of the field at a given Hyperfocal Distance, not the actual subject size on the image plane. The computed field size may be the same, but because of the angle of view the subject (on the imaging plane) will not be the same size. Wide angle lenses have more "apparent" DOF, simply because the subject becomes smaller.

>If you try your example with an actual subject, you will find that the subject is considerably smaller with the 25mm lens than with the 100mm. When you move the 25mm in toward the subject to achieve the same image size (on the imaging plane) -- not field of view -- you will have the exact same DOF.

>Brian "There is no free lunch, even if it's worms." Heller
IA 600 DP


>Brian Heller wrote:

class="Paragraph">>...If you try your example with an actual subject, you will find that the >subject is considerably smaller with the 25mm lens than with the >100mm.

>Please go back and re-read my examples. I'm not talking about a "size or area of field at a given hyperfocal distance", the HD's were totally different. The subject size IS the same on the imaging plane when you switch from a 100mm at 20 feet to a 25mm at 5 feet. That's the point.

>This was explained in detail, with illustrations, by L. Andrew Mannheim in Modern Photography, July 1970. He was dealing with still camera lenses. I simply changed the examples to focal lengths (and C of C) used in filming, referring to the ASC Manual DOF tables for the specifics. You can look them up.

>As I said, this works only with shorter lenses, small stops, and relatively close focuses, the kind of scenario miniatures present.

>Wade "Diet of Worms, here I stand!" Ramsey


>Wade "Channelling for Martin Luther" writes :

class="Paragraph">>The subject size IS the same on the imaging plane when you switch >from a 100mm at 20 feet to a 25mm at 5 feet. That's the point.

>If the subject is the same size and if the f/stop is the same, then the DOF is be the same, regardless of focal length.

class="Paragraph">>This was explained in detail, with illustrations, by L. Andrew Mannheim >in Modern Photography, July 1970. He was dealing with still camera >lenses.

>I would very much like to see the article. Is it on line somewhere? I can't really comment on something I haven't seen. So I'd like to hold off on Mr Mannheim’s article until I read it. However, arguments from authority didn't carry much weight with Martin Luther either...

>If you would like to study up on DOF, I recommend Stanley Blacker's Applied Depth of Field. It's become very hard to find. I've been told that most library copies have been stolen by optical engineering students.

class="Paragraph">>As I said, this works only with shorter lenses, small stops, and >relatively close focuses, the kind of scenario miniatures present.

>My experience with miniatures and table top work has proven time and time again that the only way to get more DOF is to stop down -- increase the light level or shoot with faster film or shoot with a longer exposure time.

>Brian "Doubting Thomas" Heller
IA 600 DP


>Brian Heller wrote:

class="Paragraph">>If the subject is the same size and if the f/stop is the same, then the >DOF is be the same, regardless of focal length...

>Go to the ASC Manual, Ninth edition, p. 757 (100mm) f/11, 20 feet; and p. 738 (25mm) f/11, 5 feet (numbers slightly different, I used the Eighth Edition, but substantially the same.) You don't need Mannheim or Blacker, the numbers are boldly self evident.

class="Paragraph">>...I would very much like to see the article. Is it on line somewhere? I >can't really comment on something I haven't seen. So I'd like to hold off >on Mr Mannheim’s article until I read it.

>I'm sure it isn't on line, Modern Photography was absorbed by Popular Photography years ago. And yes, Martin Luther also rejected the popular wisdom of the day in favor of the Book.

>Wade "do the math" Ramsey


>Wade Ramsey writes:

class="Paragraph">>Go to the ASC Manual, Ninth edition, p. 757 (100mm) f/11, 20 feet; and >p.738 (25mm) f/11, 5 feet (numbers slightly different, I used the Eighth >Edition, but substantially the same.)

>I have looked at the tables and you have quoted them correctly, and as far as I can tell, there's nothing wrong with the book. However, you seem to be missing the point. Nowhere in the tables is there any mention of subject size on the image plane. The tables are describing DOF in the subject space. The 100 mm lens is subtending a narrower angle, so the subject will be larger. If you make your subject the same size, the DOF will be the same. (At the same f/stop).

>Try this : set up a 6" cube so that one vertical edge is facing the lens, so that the two adjacent surfaces are receding from the camera. Attach a tape measure to the top edge of one of the receding side. With the 25mm lens position the camera so that the vertical edge of the cube fills the frame -- frame line to frame line. Focus on the leading edge. Set whatever stop you like. A wider stop makes it easier to see.

>Repeat the process with the 100mm, that is so the leading edge fill the frame from top to bottom. Use the same stop. The depth of field for the cube with both lenses will be identical.

>Brian Heller
IA 600 DP


>Brian wrote:

class="Paragraph">>I have looked at the tables and you have quoted them correctly And as >far as I can tell, there's nothing wrong with the book.

>You choose the distance to provide the image size. A 100mm lens at 20 feet produces the same image size of a given object as does the 25mm lens at 5 feet. But at f/11 the 25mm is providing almost twice the DOF. This won't happen in every comparison, as I mentioned before. It occurs with relatively short lenses at stops where the HD is considerably less than 10x the subject distance.

>This is what Mannheim claims and the DOF tables bear it out.

>I reckon we've worn this subject out

>Wade Ramsey


>Wade writes:

class="Paragraph">>I reckon we've worn this subject out

>Not exactly. I think you need to have another look at the tables.

>We tested Mannheim's assertion this afternoon on two 435s set up with a Zeiss 25mm on one and a 100mm Zeiss prime on the other. Both were set up on Century lens charts to the dimensions you gave. Both instances produced identical DOF.

>What's more, if you take a look at the DOF tables in the 7th Edition of the ASC Manual -- the last edition which also gives the actual size of the field of view -- you will see that when the 25mm is focused at 5' and the 100mm is focused at 20' the field of view are, as you said -- and as we confirmed today -- virtually identical.

>However, so is the DOF which is shown as 14.5" at f:2 for the 25mm vs. 15" @ f:2 for the 100mm. In the 8th Edition both are given as 15", which is almost exactly what we read off the charts this afternoon.

>Depth of field is entirely a function of f/stop and image (on the image plane) size, not focal length. Q.E. D.

>Brian Heller
IA 600 DP


>Brian writes:

class="Paragraph">>However, so is the DOF which is shown as 14.5" at f:2 for the 25mm vs. >15" @ f:2 for the 100mm. In the 8th Edition both are given as 15", which >is almost exactly what we read off the charts this afternoon.

>But, the hyperfocal distance for the 100mm lens in this setup is 646'2", well more than 10x the subject distance, so the test doesn't meet the requirements Wade mentions. (The HD for the 25mm in this situation is 40'5", which is indeed less than 10x the subject distance, but perhaps not "considerably".)

>I'd love to know why this phenomenon occurs, if indeed it does occur. Someone must have a copy of that article around!

>Chris Freilich
Cinematographer
Princeton, NJ
http://www.virtuosofilms.com


>Hi Guys,

>Brian "The Infidel" Heller wrote :

class="Paragraph">>However, so is the DOF which is shown as 14.5" at f:2 for the 25mm vs. >15" @ f:2 for the 100mm.

>Wade "The Believer" Ramsey previously wrote :

class="Paragraph">>A 100mm lens at 20 feet produces the same image size of a given >object as does the 25mm lens at 5 feet. But at f/11 the 25mm is >providing almost twice the DOF.

>So Brian, your example is good and correct at f/2 but ignores Wade's point. If you stop down to f/11 as Wade suggests the 100 will have a DoF of about 7' and the 25 will have a DoF of about 13'. Almost twice the DoF as Wade suggests. Keep stopping down and/or moving back and the wide lens will hit it's hyperfocal distance much sooner than the long.

>For all practical purposes, in a studio or interior location you will probably never reach a point where this comes into play. So depth of field is most commonly a function of image size and stop, but not entirely.

>I would also make a loose suggestion that perspective, or at least optical distortion has an effect on apparent DoF, but that's a vague and hard to argue point..

>Best Regards,

>Anders Uhl
Cinematographer
The DoP Shop
http://www.thedopshop.com


>Anders writes :

class="Paragraph">>So Brian, your example is good and correct at f/2 but ignores Wade's >point. If you stop down to f/11 as Wade suggests the 100 will have a >DoF of about 7' and the 25 will have a DoF of about 13'. Almost twice the >DoF as Wade suggests.

>In practice this is not the case. And there are many factors.

>Modern high quality lenses need individual DOF charts. In our tests the set-up with 100mm actually seemed to have more DOF than the 25mm at smaller stops. This may be due to optical variations in the lenses but without setting up a bore sight microscope it was difficult to eyeball much of anything past f:8

>Also the fields of view were off by a couple of inches.

>When you eyeball it for yourself, you may also come to a different conclusion.

>Brian "infidel?" Heller
IA 600 DP


>Brian writes:

class="Paragraph">>When you eyeball it for yourself, you may also come to a different >conclusion.

>That's why we get paid the big bucks

class="Paragraph">> Brian "infidel?" Heller

>Sorry, I was trying to follow a previously established theme.

>Best Regards,

>Anders Uhl
Cinematographer
ICG, New York


>Brian wrote:

class="Paragraph">>...However, so is the DOF which is shown as 14.5" at f:2 for the 25mm >vs. 15" @ f:2 for the 100mm

>Brian, you're absolutely right--at f/2. But slide over the charts to f/11, as I specified, and the DOF for the 100mm focused at 20' is 7', while the 25mm focused at 5' has a DOF of 13', almost twice as much.

>This situation only obtains at small stops, where the HD is considerably less than 10x the distance focused upon.

>Wade K. Ramsey, DP
Dept. of Cinema & Video Production
Bob Jones University
Greenville, SC 29614


>Chris Freilich writes :

class="Paragraph">>But, the hyperfocal distance for the 100mm lens in this setup is 646'2", >well more than 10x the subject distance, so the test doesn't meet the >requirements Wade mentions.

>We used the example Wade gave. Presumably because those settings and focal lengths gave the closest matches in terms of fields of view.

>The DOF tables are compromises based on simple lenses. They are accurate enough for what they were intended, movie production. Macro work for instance requires very different DOF standards, as do very high quality lenses -- S4s, UltraPrimes, Primos, etc.

class="Paragraph">>I'd love to know why this phenomenon occurs, if indeed it does occur. >Someone must have a copy of that article around!

>I'd like to see it as well. My experience with table top work has proven many times that focal length is not a factor in DOF. However diffraction, interference, and quality of the optics can be significant factors.

>Brian Heller
IA 600 DP


class="Paragraph">>Does the image size for the two lenses match at these two distances?

>Yep! Every time you double the focal length you double the distance to get the same image size of the subject at the plane of focus. So for 25mm to 100mm, you have to quadruple the distance, 5 feet to 20 feet.

>Of course, the sizes of objects nearer to or farther from the camera than the subject focused upon are going to change drastically because of perspective. That's one big reason we like to use different focal lengths.

>Wade K. Ramsey, DP
Dept. of Cinema & Video Production
Bob Jones University
Greenville, SC 29614


>Brian Heller wrote :

class="Paragraph">> In practice this is not the case. And there are many factors.

>I'm very interested in these other factors. I certainly respect your experience and knowledge, but I'm not understanding how hyperfocal distance as it relates to depth of field is a wash in practice.

class="Paragraph">>Modern high quality lenses need individual DOF charts.

>Agreed. You get similar results with individual charts from Cooke and Zeiss.

>At the same distances previously stated (5' and 20') and the same stop of f/11 the Cooke S4 and Zeiss UltraPrimes charts say the following :

>Cooke S4

>25mm, f/11, 5' - DoF= 3'2.4" to 12'11.5" or roughly 10'
100mm, f/11, 20' - DoF= 17'3.2" to 23'9.5" or roughly 6'6"

>Zeiss UP

>24mm, f/11, 5' - DoF= 3'4 1/4" to 26"1/2' or roughly 23' - which seems like a misprint of about 10' so lets knock 10' off and call it 13' (granted it is wider than the 25mm and image size will be slightly different) 100mm, f/11, 20' - DoF= 17'4 1/2" to 23'11" or roughly 6'6"

>So again, with these numbers you have almost twice the DoF with the wide lenses as was originally suggested. And you will hit the hyperfocal distance sooner on the wide lens, do you not agree with that?

>Best Regards,

>Anders Uhl
Cinematographer, NY
The dop shop
http://www.thedopshop.com