DOF Confusion for group shots - Is focus a plane or fixed distance to sensor?

<p>Hi all,<br>

Here's the reason I'm asking this question: Say you have a group of 5-10 people, and for argument's sake let's just say DOF is not a lot w/ the settings you're shooting at. Would you line up the group in a straight line parallel to the camera's sensor, or would you bring the people standing at the outer edge a little closer to the camera to get all of them in focus (since the people at the outer edge are farther away from the camera distance-wise)?<br>

x x x x x x x x x x</p>

<p>-----Camera-----</p>

<p> OR</p>

<p>-x x x x x x x x-<br>

x---------------x</p>

<p>-----Camera-----</p>

<p>I'm trying to find out if DOF is a fixed plane at the focus point, or if it is based on straight up distance to the sensor. I was pretty sure it's a plane, but after reading an example from a VERY reliable source (book) which suggested placing the outer people closer, it made me second guess myself.<br>

Thoughts?</p>

<p>The point of sharp focus is different from DOF. Your first example shows the point of sharp focus - parallel to the sensor and at a pre-determined distance (the AF point, if you will). If you shoot with large apertures, then that's how you would want to organize the group - all in one row - boring, I know, but at f/1.8 or even f/2 there is not much DOF.</p>

<p>If on the other hand you can shoot with a small aperture, then you can use the hyperfocal distance for your lens, say the HFD is 6 feet to 20 feet, then everything from 6 feet to the 20 feet mark will be in acceptable focus</p>

<p>Unless you are specifically buying a flat field lens, when you buy your lens, the focal plane of your lens is not straight across in front of you, it is curved, (around you).<br>

Arrange your group in a curve and set your aperature small, say F4 to F8, for a group of that size.<br>

For larger groups than that, F8 to F11 is even better.<br>

For groups, you want a lot of depth of field so everyone is in focus.</p>

<p>I thought the in-focus zone of a typical lens <em>was</em> a plane. Maybe not as perfect as the mentioned "flat field" lens, but pretty close. For example, you should be able to photograph something like a bookcase, focus in the middle, and have the titles in the corners still reasonably sharp.</p>

<p>The flatness of the curve will depend on the optical design. Flat field lenses have the flattest curve, single meniscus lenses have a very curved one. The only way to know other than lots of testing is if the maker published the information. If the lens design uses internal focusing or is a zoom, the flatness of the curve could vary substantially over the focus or zoom range.</p>

 

<p>Run your own test on your target lenses:</p>

<p>D D<br /> AXXXXXXXXXXX*MID*XXXXXXXXXXXA<br /> B B<br /> C C<br /> ^<br /> *****************camera</p>

<p>One can use chairs; rows of desks at as school; vertical shovels/poles in the yard; filing cards or bricks on the yard; etc. <br /> <br /> If you focus on *mid* and the two A's are sharper than B, C or D; its a flat field lens. Some lenses might have B abit sharper; the field is curved inward. IF A,B C is a yardstick; one can see the offset distance.<br /> One thing that can mess up the test is if you are not parallel to the X's; or ones film plane is machined abit off. I oncd had an Exakta VX500 that was like that; it tested with wonky results; more like the best would be B left and D right.<br /> <br /> There is no real answer to your question; the lens is unknown.<br /> A good lens typically is designed to be sort of a flat field lens. In simple cameras suchs as box cameras the film plane is sometimes curved.</p>

<blockquote>

<p>Unless you are specifically buying a flat field lens, when you buy your lens, the focal plane of your lens is not straight across in front of you, it is curved, (around you).</p>

</blockquote>

<p>Hang on a minute... this is <strong>totally </strong>untrue. To a first approximation (and even to a second approximation) all lenses are flat field lenses. Clues: a) the lens equation (pre-high-school physics), b) university physics c) the term "focal plane", d) the focus-and-recompose-error which wouldn't be an issue unless the focal plane was, well, a flat plane, e) the scheimpflug principle, and so on. Divergence from this ideal behaviour takes the form of spherical aberation, and coma, which lens designers do their best to minimize - but these errors are very far from giving focus to things that curve around the lens.<br /> Group portraits: in a straight line and parallel to the film plane unless you're using a camera with "movements".</p>

<p>Wow! I'm not even going to argue with that mumbo-jumbo.<br>

Best thing, photograph your group at F8.</p>

<p><em>"Here's </em><strong><em>the reason I'm asking this question</em></strong><em>: Say you have a group of 5-10 people </em><strong><em>would you bring the people standing at the outer edge a little closer to the camera</em></strong><em> to get all of them in focus (since the people at the outer edge are farther away from the camera distance-wise)?"</em></p>

<p>No. </p>

<p>There is no need to bring the people standing at the outer edge, closer to the camera to ensure they are within the DoF.<br>

Mine is merely a practical answer. I have no argument with the fact that lenses are not perfect and that the actual PLANE of sharp focus is indeed not a perfect plane. But with most modern lenses within the FL ranges one would be using at Weddings and Functions for group shots it would be safe to assume the PLANE of SHARP FOCUS is indeed a Plane and right angles to the Focal Plane of the camera. </p>

<p>Other practical considerations:</p>

<p>> Straight lines are not usually tasteful composition though.<br>

> And we might consider that people at the edges might suffer barrel distortion and be affected by other lens issues more adversely than those standing in the centre, like Chromatic Aberration for example.<br>

> All lenses tend to be softer at the edges than in the centre - and this is usually worse when the lens is used wide open. A contributing factor to this is the inherent design of the lens to be practical (economical) to produce - so some of that edge softness is because the actual Plane of Sharp Focus is indeed not a Perfect Plane - so it seems I am arguing against myself . . . well I am not really - I am merely attempting to separate what is practically useful and where theory takes over.</p>

<p>As an example - if you do the experiment as described by Kelly and you are using a 24 to 70, you might find comparatively "worse" results for the lens when at the 24mm end than when it is used at 50mm or 70mm.<br>

<br>

WW</p>

 

<p>There are a couple of things involved. 1. Plane of focus relative to the distance (literally) from the sensor. 2. Curvature of the field of focus, due to the lens.</p>

<p>1. this means you don't have to take into account the actual distance from cam to subject, but because of #2, the lens will (in essence) be designed to bring those outer edges into focus at the same plane as the center. Its dependent on the lens used (wider being worse), but the general principle is that you should not have to compensate with your "line" of subjects. In reality however, you do have to compensate relative to the way a particular lens renders the curvature of the field. Confusing somewhat, but once you know the way your lens behaves at a particular focal distance, and the distance to subject, you will know how much to compensate.</p>

<p>As a rule, using a 35 - 50mm FL will mean you can keep a relatively straight line (of 10 people) provided you are at least 15' from the line with your camera position.</p>

<p>Hope that helps.</p>

<p>Alec, the human eye lens seems to differ from your statement regarding ALL lenses, as do many books on optical theory. Have a look at chapter 6 of<br>

The manual of photography: photographic and digital imaging By R. E. Jacobson</p>

<p>Aesthetic considerations aside, please let the OP be in absolutely no doubt whatsoever that with any camera where the film or sensor is flat, the plane of focus of any photographic lens is a flat plane which lies parallel to that film/sensor plane (unless lens tilt or shift is employed, in which case it it still a flat plane, angled as per the 'hinge' and 'scheimpflug' rules).</p>

<p>If there be any residual doubt I'm happy to refer to any number of wide-angle group portraits that I have been forced to shoot - dozens a week - at f/2.8 because of atrocious lighting, and all of which demonstrate an entirely flat plane of focus, to within the acuity of the human eye to distinguish.</p>

<blockquote>

<p>Have a look at chapter 6 of<br />The manual of photography: photographic and digital imaging By R. E. Jacobson</p>

</blockquote>

<p>Bob - thanks, I will. But cameras use spherical lenses for which a good approximation is the "lens-makers's formula" - and where aspherical lenses are used it is to correct for aberations and create a flatter plane of focus.<br>

EDIT: ok then - you're right, I guess, in principle:<br>

<a href="http://toothwalker.org/optics/astigmatism.html">http://toothwalker.org/optics/astigmatism.html</a></p>

<p> </p>

<p>Seems like the answer is, "it depends". The practical answer is to have a margin of error on your DOF whenever possible, whether the group is U shaped or not, and know where to put the point of focus after estimating the group's depth.</p>

<p>I always heard that you should 'not' purposefully arrange groups with the ends curved. I think this is hard to do in some cases, however, because people tend to do this naturally, I find. I am always telling people on the ends to go backward. If one is pressed for time, well--sometimes I just let it go. So back to having a margin of error on DOF and not shaving it so closely.</p>

<p>Don't confuse DOF with hyperfocal distance. If you calculate the hyperfocal distance using say F8 or F11, you will get all your group in acceptable focus.</p>

<p>because people tend to do this naturally, I find. I am always telling people on the ends to go backward.</p>

<blockquote>

<p>This is certainly true!</p>

</blockquote>

<blockquote>

<p>Seems like the answer is, "it depends".</p>

</blockquote>

<p>Nadine - I think it is misleading to leave the OP with the impression that the answer to his original question is "<em>it depends.</em>" He asked:</p>

<blockquote>

<p>Is focus a plane or fixed distance to sensor?</p>

</blockquote>

<p>To which the answer is "it is absolutely <strong>not </strong>a fixed distance to the sensor." </p>

<p>There may be some residual curvature of field away from a flat plane but this is of the kind of order that interests optical designers and people who like to mount their lenses on calibrated test-beds to measure and obsess about MTF charts. Any lens where the focus even approached the stage where groups needed to be curved around the camera would be laughed off the market as unusable. </p>

<p >I think a couple of things are happening here. There was not just one question asked by Alen: also we might easily extrapolate that other question(s) were implied and it is noted that other contributions have developed the conversation beyond those questions asked an implied.</p>

<p > </p>

<p >I am not answering for Nadine, but I read her answer to mean:</p>

<p > </p>

<p >"It depends . . . whether or not we might have (or allow) the subjects make a slight bend the ends of the line of the group of subjects</p>

<p >OR</p>

<p >"It depends . . . whether or not we might squish the ends in a little bit from the edge of the frame.</p>

<p > </p>

<p >I read either or both of these meanings to be in regard to getting the best possible image and it being the sharpest at the edges - with the underling premise being that we have enough DoF so to do (i.e that the DoF be at least equal to, or a bit more than, the depth of the group) and then it is important that we accurately assess where to focus, so that the DoF does indeed cover the group’s depth. </p>

<p > </p>

<p >WW </p>

<p > </p>

 

<p>William -<br>

Points well made, and all taken. No offence intended by me to anyone - and I hope none taken.</p>

<blockquote>

<p>Run your own test on your target lenses</p>

 

</blockquote>

<p>Well said! Back before the internet supplied answers to all our questions, this is how we managed ;)</p>

<p>Alec--no worries. I can see how you would think about my answer. I was actually answering the following.</p>

<p><strong>Would you line up the group in a straight line parallel to the camera's sensor, or would you bring the people standing at the outer edge a little closer to the camera to get all of them in focus (since the people at the outer edge are farther away from the camera distance-wise)?<br /></strong><strong></strong><br>

So William W. answered for me perfectly (thank you!). What I really wanted to point out was that even if we don't know the answer to the whether the focal plane is a fixed distance, etc., we can still shoot group shots with a technique we know will work, and that other factors also come into play. And, I am a very practical person, so while I must have read somewhere that focus is a plane, so don't curve groups, I only remembered the latter part, because that's the part I would use in practice.</p>

<p>***Really what matters is if the *answer* is going to help you shoot better group images.<br /> <br /> There is really no mention of the target lens; thus the *answer* is just a crapshoot.<br /> <br /> In REAL engineering one avoids all this guessing; marketing fluff and runs an actual test; or finds a test for the specific device.<br>

<br /> If one is afraid to do an actual test; or one seeks a group concensus; just assume the lens is perfectly a flat field device and stop down alot so there is *no problem*; ie DOF quashes any field curvature.<br>

<br /> ****This is the most practical thing to do IF one has no time to test; or it is not worth the trouble.<br /> How folks brains are wired varies. Here I have shot images for 50 years; have 2 engineering degrees; and have worked in optical design. I am always testing stuff; it is in my DNA. Testing is not in everybodys DNA. Some would rather ask folks on photo.net if a scratch is an issue on a lens and get a concensus. Here I test it; I prefer an actual test to a groups "feelings".<br /> <br /> If one looks back on older photo books of the 1920's and 1930's; some DID have one moving in folks at the edges of a groups shot. This was based on a simple design of lens; often a single one or triplet.<br /> <br /> Running an actual test is really no big deal; whether one is seeing if the field is curved; or ones strobe/flash really has a wide angle coverage; or ones shutter speeds.<br /> <br /> As a practical matter; understanding ones tools limits and flaws helps one do a better job.<br /> Even our old process camera that shoots a 24x36" negative of flat objects on a len board has a curved field; and it is a so called flat field lens; a process lens. Both film and map are parallel; one HAS to stop down to F22 to get the corners sharp.</p>

<p><img src="http://d6d2h4gfvy8t8.cloudfront.net/1238404-lg.jpg" alt="" /></p>

<p><img src="http://d6d2h4gfvy8t8.cloudfront.net/1238354-lg.jpg" alt="" width="467" height="417" /></p>

<p>Even with a test of a lens like this Konica Auto S2 at F5.6 of USAF charts on a brick wall; one can get a slightly better corner performance/results/finer bars resolved; IF one purposely does NOT have the focus spot on in the center. NO lens is perfect; all lenses have some curvature; even great ones.</p>

<p><img src="http://d6d2h4gfvy8t8.cloudfront.net/1471363-lg.jpg" alt="" /></p>

<p>Instead of folks quoting simpleton Physics books; one should either:<br /> <br /> (1) just assume your lens is perfect; and then stop down some to cover ones bum with DOF.<br /> <br /> (2) Run an actual test : IF one wants to see what is really happening. It is not all that hard to do; just run a test and see where the best point of focus really is. One can place Barbie/Ken/GiJoe dolls on a garage floor in rows; use old brick in rows; stick dowels in the yard; whatever.</p>

<p>The bizzare thing about this question is that NO mention of the actual lens was made; and a sweeping general answer is wanted. It is like asking if a dog is good at fetching; with no mention of breed.:)<br /> <br /> Assuming you lens is flat field is a decent hip shoot if you know nothing.<br /> IF one really wants to explore using a faster fstop than another person; running an actual test is warranted. Tests like these are common in the movie business; one may want two actors to be in focus with a wide open lens; one on axis; one to the side. Real test data matters more than talk.<br /> <br /> If one runs a Circuit Camera that revolves; one place the folks in a constant distance from the camera. Folks also do this too soemtimes when one shoots several frames and stitches together a group shot. A stitched group can be problematical; folks move!.<br /> <br /> ****In summary a practical answer is to NOT move in folks at the corners; UNLESS one has some real test data on your lens; at the specific distance.<br /> <br /> Here If I want to push the limits; or learn something; I run a test; where many others would rather use a consensus on some unknown lens. There is really nothing wrong with either stance. One has other issues with testing; one ones up the Pandoras box of focus shift while stopping down; focus errors; having a faulty test settup. A poor test can lead you down the wrong path.</p>

<p>A question like this is like asking if ones floor is flat.<br>

One person may just walk around the room an *sense* if the room is like a circus fun house. Another might see if a ball rolls. Another may note that a spilled glass of liquid travels East. Another might bring out a torpedo level; or 6 foot level. Another might compare its flatness to a NBS standard surface plate; or reference 1/20 wave optical flat. One might be in a basement floor;and the whole floor is purposely made with a slant to aid draining/cleaning; ie water flow to the sump pump. At some point the quest goes from practical to an absurd worry!</p>

<p><br /> One must be carefull about getting bogged down in worrying about what doesnt matter; or creating other issues.<br>

With a group shot the chances of getting no blinks; looking down/away grows and the group size grows; thus one has to shoot more shots to get a keeper.<br>

If too low a shutter speed is used; suger filled kids will be blured; and asleep grandpa in the chair might be tack sharp; but with closed eyes.</p>

<p>Just wanted to point out that the lens everyone is talking about is not a fish-eye!<br>

Also, from the internet: Flat field lenses generally are used in macro photography or copy work.... They are corrected and balanced mostly for close up work at miniscule levels of depth of field.</p>

<p>:)<br>

One of the pleasures here for me, is reading this "Engineer's" prose . . .<br>

<em>"If too low a shutter speed is used; suger filled kids will be blured; and asleep grandpa in the chair might be tack sharp; but with closed eyes."</em><br>

that sentence is an absolute cracker! <br>

WW</p>

<p > </p>

 

<p>Hey Kelly,</p>

<p>What's with the money on the bellows???</p>