Focus and DOF bracketing in macro photography

[lloyd_nakatani]

In macro photography, many of my discards are due to slight errors in

focus or inadequate DOF. I use manual focus (sometimes with the

assist

of AF confirmation) and DOF preview, but its hard to judge the

difference between right on and just slightly off, especially when the

DOF preview image is so dim. The problem is not my equipment or basic

technique, because I've taken macro shots that are extremely sharp. I

thought a solution might be a systematic approach to focus and DOF

bracketing. The two are obviously related, so bracketing taking both

into account is not straightforward. What is the best way to do focus

and DOF bracketing? Using the smallest aperture is an obvious but not

necessarily best answer, because diffraction effects reduce sharpness

and

the background might become too sharp and distracting. For the sake of

discussion, assume that there is some combination of focus setting and

aperture that will produce a keeper (everything that should be sharp

is

adequately sharp, and the background is sufficiently blurred and not

distracting). What is the best way to insure that at least one shot

will be close to that ideal combination?

[ron_shaw]

One thing you may want to do, is to focus 1/3 of the way into your subject, as opposed to the leading edge, for instance. That way you take advantage of the DOF increase you get on both sides of the plane of focus. Other than that, there isnt much you can do.

[lloyd_nakatani]

Let me give an example to make the problem clear. Suppose that I'm taking a macro shot in which I want a near object and a far object both sharp, and the far object is the primary subject. Suppose further that f/11 has sufficient DOF for the shot (a fact that I don't know for certain until I look at my slides under a loupe). If I knew that when I was taking the shot, I would focus halfway between the near and far objects (because the DOF is approximately symmetric in macro shots), and shoot at f/11. The problem is that I don't know that for sure *when* I'm taking the shot. So I might focus and DOF bracket as follows. At maximum aperture, I should focus on the far object, since that is the primary subject, and let the near object go soft. A minimum aperture, say f/32, I should focus on the near object to get both near and far objects sharp (thanks to the DOF) and to blur the background as much as possible. But at f/32, the background will be sharper than I want, and overall sharpness will be reduced because of diffraction effects. Somewhere between these extremes is the shot I actually want with near and far objects sharp, and maximum blurring of the background. Another factor favoring the largest aperture that will produce the desired picture is the frequent need for the fastest possible shutter speed to stop wind motion. So is there a systematic way to approach focus and DOF bracketing that will insure that one of the shots will be the one I want in terms of sharpness and blurring?

[clivecc]

Lloyd, what I usually do with my 100mm macro is find the optimum focal point (usually 1/3 between near and far subjects) and take several shots varying apperture from wherever I want to f 32. I also take several shots varying flash exposure and backround exposure and even focal point at times. This takes up quite a bit of film but if the shot is a killer I don't mind. Besides sometimes looking in the loupe the "just in case" shots (f 32 etc) are the ones I keep. (All this assuming a co operative subject)

 

I hear what you say about not being able to determine exact sharpness with DOF preview due to loss of light in viewfinder and I have the same problem but, short of getting out a calculator and ruler, taking many varied shots seems the best solution until you know your system so well that the number of rejects reduce.

 

I find that if backround blurr is not good with the 100mm, I use the 70-200mm with closeup lens and/or extension tubes/ TC's.

[bobatkins]

I don't know where this "1/3 distance rule " junk comes

from (probably Popular Photography),

but it's <b>wrong</b> for macro work.

DOF for macro work is <b>symetrical</b>.

You get the same DOF in front of the subject as behind it.

<p>

Sure there's some combination of lens focal length, aperture and subject distance

that gives you a 1/3 - 2/3 DOF split, but it's not macro. DOF varies from

a 1 to 1 split all the way out to a 1 to infinity split. The 1/3 is just a VERY rough

approximation of some undefined general condition which sometimes applies but often

doesn't. Not much of a rule really.

[lloyd_nakatani]

The risk of always focusing halfway between the near and far object and using f/32 for the safety shot is that even at f/32, the DOF may not be enough to get both sharp. In that case, the safety shot should focused so that at least the primary subject is sharp, and the secondary subject is as sharp as possible. Again, the dilemma is not knowing when the shot is taken that both won't be sharp at f/32. DOF preview is not of much help in determining this because the image is so dark at f/32.

[paulo_bizarro1]

Here's what I do, especially with flowers (3D objects). I focus where I want the subject to be in focus, and then take several shots at different apertures, up to the aperture where the background gets distracting. This means usually f/11; the background does not need to be sharp to become distracting. Up to f/11 or f/16, DOF preview works fine.

 

With two objects far apart, to get them both in focus is virtually impossible, so you have to choose which one gets the leading role and which one gets the supporting role.

[lloyd_nakatani]

But, Paulo, don't you then "waste" the DOF that's in front of the point focused on? That "wasted" DOF might be put to better use to get more sharpness on something more critical than the air in front of the subject while keeping the background from getting too distracting.

[zack_lau]

Work on improving your ability to see DOF previews.

I find I see them better in a dark room after I

wait for my eyes to adjust. Brighter lights,even if

only temporary, may also help. It may help to remove

any filters temporarily to make the image brighter.

[lloyd_nakatani]

I try to get as much as I can from DOF preview with the following techniques. I start wide and close down gradually to the working aperture, letting my eyes adapt on the way down. Also using a dark cloth (actually just a black T-shirt over my head and camera) is a big help in keeping stray light from washing out a dark image. I just got a angle viewfinder attachment which should eliminate the contortions I have to go through now to look through the viewfinder. The attachment should enable me to use the DOF preview more effectively. Removing filters, especially a polarizer, should help, but it is tedious. Besides, most of my shots are made without filters. Making the light brighter is a possibility indoors, but for macro shots outdoors of flowers in natural light, it's usually impractical.

 

The problem is that DOF preview is fundamentally inadequate. DOF preview requires us to judge sharpness in a small, dark image. This is often hard to do even under ideal conditions: slides on a bright light table with an 8X loupe. It's easy to see perfect sharpness, but much harder to see *adequate* sharpness even then. DOF preview is really about judging adequate sharpness under less than ideal conditions.

 

Exposure metering, no matter how carefully done, leaves some margin for error. Exposure bracketing is a way to address the residual uncertainty about getting the "perfect" exposure that can only be determined after looking at the slides. There is analogous uncertainty about getting "perfect" sharpness that should be addressable with a bracketing strategy. DOF preview, like an exposure meter, gets me in the ballpark, but often I still strike out because the margin for error is so small.

[paulo_bizarro1]

Lloyd,

 

I have found out that I get more keepers if I get the part of the subject I want focused to be in focus in the first place. I might be wasting some DOF as you say, but at least I am sure that the focus is where I want it to be.

 

I usually shoot flowers at f/3.5, f/8, and f/16, with the 180 macro. At f/3.5, I focus on the stamen (spelling?), and let the petals go out of focus. I like this for red poppies, because the out of focus petals give a pastel like effect.

 

At f/16, the background usually starts to get distracting, and I have to compose more carefully to place the other elements in a pleasing way. Nevertheless, I still place the focus on the stamen.

[lloyd_nakatani]

DOF is an extremely precious commodity in macro photography, because there is so little of it. I'll waste DOF for artistic effect if that is my intention, but I'd rather not waste it out of ignorance when I don't want to. Also, I'd rather not waste film on shots because the bracketing strategy is suboptimal. For example, exposure bracketing by 1/3 stop is a waste of film in most situations. Are there equivalent rules-of-thumb for focus and DOF bracketing?

 

The problem also arises in landscape photography when you want both foreground and background to be sharp in a wide-angle shot at an aperture that avoids degradation in sharpness due to diffraction effects and to wind-induced motion of, say, flowers in the foreground.

 

The collective wisdom of this forum can make a contribution to photographic technique by coming up with a strategy for "sharpness bracketing" (for want of a better term). This is problem that's pervasive in all kinds of photography. Macro photography is just when the problem is most acute.

[steve_smith1]

Lloyd: You're certainly right that DOF is a precious commodity in macrophotography, but it's hardly ever wasted. Misused, maybe.

 

I think the only way to do what you want would be to calculate the DOF for every aperture you want to use at various magnifications, and create a little slide rule. When you take a shot at F(x) with a magnification of (y) you know you have DOF of 5mm - focus in the the middle of the field, then (perhaps) on an object 1.25mm closer, then on one 1.25mm beyond the center point. It might work.

 

Paulo's comment hits the nail on the head: pick your subject (or sub-subject) and camera placement carefully to use the limited DOF as wisely as possible. I doubt there's a simple equivalent to auto-bracketing exposure (where there are a lot fewer variables), but if you find it please report back (who knows, maybe autofocus- bracketing will be a feature an the next high-$ Canikon, if they don't have it already!?!).

 

Steve

[lloyd_nakatani]

Steve: In theory, you have the answer. Measure the distances to the near and far subjects to be sharp, and compute the aperture that provides a sufficient DOF and set the focus accordingly. But it's hard to put that theory into practice. As you mentioned, distances have to be measured to a precision that's hard to realize in the field, and the use of a zoom lens with diopter for macro photography complicates the computation, assuming that it can be done conveniently at all. I don't relish the thought of taking a laptop along to do macro photography, and the kind of table you mentioned will be too big to carry for a zoom lens. So we need something that's practical and more seat-of-the-pants than precise science.

 

So how does one "use the limited DOF as wisely as possible?" That is the crux of the problem.

[lloyd_nakatani]

By the way, does anyone know if the automatic DOF feature of some EOS camera works for macro photograpy with a zoom lens fitted with diopter? If it does, then maybe the answer can be formulated in terms of how that feature might be used to solve the problem.

[andrewdawsongallery]

Just another thought to keep in mind re: using f/32. Although you will get more depth of field than with f/22, the *overall* sharpness will not be as good. At very small apertures there is a greater tendency for light scattering and other problems, which reduce overall sharpness. Technically, it probably starts to occur at f/16, but it doesn't become noticeable until f/32. (This applies to absolute f-stops, not the apparent ones that your camera may display when you're focused down tight, say around 1:1).

[bobatkins]

Most lenses (including macro lenses) peak in sharpness around f8 to f11. After that you trade overall sharpness loss for DOF. It's diffraction rather than scattering that causes the loss of sharpness. The drop in sharpness is quite evident with most lenses even when you stop down to f16 if you look closely. This is one reason why T/S lenses can be useful. You can tilt the lens to get an shift in focal plane and so effectively increase your DOF while still maintaining optimum apertures around f8.

[lloyd_nakatani]

In pondering the problem of sharpness bracketing, it just occurred to me that maybe I'm not using DOF preview effectively. What I've been doing is to focus halfway between the near and far subjects and then activate the DOF preview. Then stop down gradually to give my eye time to adapt to a darkening image --- sometimes with a dark cloth over my head and camera to block stray light ---- until I get to an aperture where the DOF looks adequate (near and far subjects sharp enough). As I mentioned in a earlier post, it's easier to see perfect sharpness than to judge adequate sharpness, so this process is fraught with sloppiness due to the difficulty of making such a judgment accurately.

 

The following might be a better way to do DOF preview. I'll do as above until I get to a aperture that looks good. Then I'll do the following. Release the DOF preview. Close down one more stop. Focus on the far subject. Activate the DOF preview, wait for my eyes to adapt to the darkened image. Then --- this the critical step --- while keeping my eye on the near subject, adjust the focus to make the near subject perfectly sharp. If I can't see the sharpness change noticeably while adjusting the focus, then the DOF must be *more* than sufficient with the extra stop down. Then I open up a stop to the shooting aperture with a DOF that should be *just* sufficient. With this process, now I'm looking for a change in the image rather than judging a static image, so the judgment about sharpness should be more accurate.

 

This new process makes a critical assumption: that closing down one stop doubles the DOF, or conversely, opening up one stop halves the DOF. Does anyone know whether this relationship between f/stop and DOF is true for distances in the range of macro photography. If not, what is the relationship?

 

Has anyone else tried this technique, or any other technique for making DOF preview more effective?

[bobatkins]

The answer is that DOF preview isn't very good. You just can't tell what's sharp no matter what you do. Looking sharp in a dim viewfinder and on an 11x14 print are different animals. For macro work a DOF table is probably more use.

<p>

To double your DOF you have to stop down by 2 stops. DOF is directly proportional to aperture in the macro regime.

 

<p>

Here's a table (total DOF in mm)

 

<pre>

1:1 1:2 1:3 1:4

 

f8 0.8 2.4 4.8 8.0

f11 1.1 3.3 6.6 11

f16 1.6 4.8 9.6 16

f22 2.2 6.6 13 22

 

</pre>

[lloyd_nakatani]

Thanks, Bob. I knew you'd have the answer and was hoping you'd respond.

 

Since a 2-stop change doubles or halves the DOF, that means that the technique I suggested will darken the image too much. So here's a modification of the technique that might work better.

 

Focus midway between near and far subjects. Activate DOF preview and gradually stop down until the image looks OK. Then grab the focus ring in such a way that you can change it but return back to the same focus setting (maybe with the fingers on both sides, or on top and bottom). Then while looking at the near subject, adjust the focus to make the near subject perfectly sharp. Return to the original focus setting. Then while looking at the far subject, adjust the focus to make the far subject perfectly sharp. If you can see the sharpness of either subject change as you adjust the focus, then the DOF is *not* sufficient. In that case, close down a stop and repeat the procedure.

You'll either find the "correct" aperture or find that no aperture is gives suffficient DOF. In that case, you'll have to compromise and use what DOF you have "artistically" or give up.

 

I agree with Bob that DOF preview is not as precise as we'd like, but it's all we have to work with when taking the picture. I don't think a table is really an alternative, because DOF depends on distance, focal length and aperture, and taking measurements accurately to a mm doesn't seem feasible to me in the field with 3D objects. The technique I've suggested is my stab at making the best of a bad situation. Can anyone suggest anything better?

[bobatkins]

DOF depends on aperture and magnification (as in the table). It's not too hard to estimate magnification most of the time. Some macro lenses have it marked on the barrel I think (though not modern AF lenses).

[lloyd_nakatani]

Thinking harder about the problem, the technique can be simplified further. It should suffice to just adjust the focus slightly in the direction of bringing the near subject into sharper focus while watching it to see if its sharpness changes noticeably. Then do the equivalent with the far object. In essence, you're rocking the focus back and forth and looking to see if the sharpness changes in the critical parts of the image. Since this reduces the problem to focusing (albeit on a dim image) rather than judging DOF, and since we can focus accurately, this technique should prove better for determining adequate DOF. After all, we can all focus accurately, right?

[zack_lau]

While a laptop and a cheap dial caliper may be impractical, perhaps you can develop a sense a "photographic vision," the ability to study a subject carefully with both eyes (you lose distance info looking through a viewfinder) and translating what you see to how it appears as a finished product. 1 hour or same day film processing is great for closing the feedback loop.

[lloyd_nakatani]

Last night, I tried the technique described. It works!!! With a dark cloth, I could even determine DOF at f/32. By rocking the focus ring back and forth around the midpoint focus setting, I could see the near and far subject either going from slightly blurred to sharp or remaining sharp despite the rocking. Because the DOF is so tiny in macro photography, it doesn't take much adjustment of the focus to cover the range from near to far.

 

Focusing at small f-stops is hard to two reasons. First, the image is dim. A dark cloth is a big help when the image is very dim. And second, the wider DOF makes it harder to determine the precise focus setting for a particular distance. The technique capitalizes on the second difficulty, and turns it into an advantage. Since we can all focus (otherwise, we could not take sharp pictures using manual focus), the technique reduces the problem of judging the adequancy of DOF into a focusing problem, which is much easier to solve visually.