Depth of field issue
Nov 8, 2015 06:28:17 #
Reality check time folks. The numbers for DOF being bounced around are not anything like what shows up in the picture! For a good reason too, or actually two.
First, lets get a good DOF calculator and learn how to use it. The one at http://www.tawbaware.com/maxlyons/calc.htm is very good.
Plug in the numbers: 10 feet focus distance, 400mm focal length, 6.3 aperture, and a CoC of 0.019, and it says the DOF is 0.5". That is twice what most folks are saying, and I suspect they aren't reading their calculators right.
However, it doesn't take much effort to look at the picture and realize the DOF has to be much greater than that! How can that be???
Easy, the OP said he thought the squirrel was about 10 feet away and that is what everyone is using. It just twar not so! The Exif data says 5.46 meters, or about 18 feet! Oooops. Back to the calculator, and we can see the DOF would be more like 1.6 inches, or six times as much as all the claims of 0.25 inches! And 1.6 inches fits what is visible in the picture too.
Incidentally, exiftool also calculates DOF and lists 0.04 meters for that image. That's 1.57 inches.
First, lets get a good DOF calculator and learn how to use it. The one at http://www.tawbaware.com/maxlyons/calc.htm is very good.
Plug in the numbers: 10 feet focus distance, 400mm focal length, 6.3 aperture, and a CoC of 0.019, and it says the DOF is 0.5". That is twice what most folks are saying, and I suspect they aren't reading their calculators right.
However, it doesn't take much effort to look at the picture and realize the DOF has to be much greater than that! How can that be???
Easy, the OP said he thought the squirrel was about 10 feet away and that is what everyone is using. It just twar not so! The Exif data says 5.46 meters, or about 18 feet! Oooops. Back to the calculator, and we can see the DOF would be more like 1.6 inches, or six times as much as all the claims of 0.25 inches! And 1.6 inches fits what is visible in the picture too.
Incidentally, exiftool also calculates DOF and lists 0.04 meters for that image. That's 1.57 inches.
Nov 8, 2015 12:39:42 #
Nov 8, 2015 13:13:00 #
amfoto1
Loc: San Jose, Calif. USA
Three things "rule" depth of field: lens aperture, lens focal length and distance to the object. Everything else is irrelevant, as far as DoF itself is concerned. Sensor and pixel site size aren't direct factors. The 7DII actually renders exactly the same DoF as your 5DIII (...see below).
However, once you are trying to "get the shot", other indirect factors such as shutter speed, lighting, ISO, diffraction, etc. start to come into play.
Smaller sensor cameras seem to have greater depth of field. They don't actually... If you shoot from the same distance with the same focal length and aperture, DoF will be exactly the same.
But, we don't do that. In order to frame a subject the same way with different sensor formats, we change focal length and/or our distance to the subject. In fact, in order to get that same shot with your 5DIII (assuming you were zoomed to 400mm on the 7DII) you would have had to use a 640mm lens. Or you would have had to move closer, say to about 7 foot distance. Or a little bit of both. And doing either of these would, in fact, decrease DoF. So, even though DoF is basically independent of sensor/film size, for all practical purposes due to the way we use them, smaller formats will render greater DoF. (Actually there's another factor that's related to sensor size and density, called the "circle of confusion"... which also relates to the size of the image as it will be viewed and even the distance from which it's viewed... but let's just put that aside for now, so as to not confuse the issue more than necessary!)
Here you're using a long focal length quite close with the lens aperture nearly wide open, so DoF is quite shallow. Using a DoF calculator, you'll find that it's .04 foot... or about 1/2 inch! (Here's an online calculator: http://www.dofmaster.com/dofjs.html).
Now, if you'd used a shutter speed half as fast (1/620), you could have closed the aperture down another stop (f9) to increase DoF a little bit... To .06 foot or roughly 3/4 inch. Still not very much to work with!
Increasing ISO a stop too, would have allowed you to use an even smaller aperture of f12.7, rendering a little more DoF... about .08 feet or roughly 1 inch. Better DoF, however you may not want to use such a high ISO and that small an aperture is going to start causing "diffraction", which will rob the image of some of it's fine detail. Like DoF, diffraction also isn't actually directly related to sensor size... But for practical purposes it is, because to create the same size final print from a full frame camera requires considerably less magnification of the image than it does from a crop sensor camera, and the less magnification that need be applied, the less obvious diffraction will be. So, diffraction can be a bigger concern with a crop sensor camera, than it is with FF.
Using a flash or other means of adding light might allow for slower shutter speed and an even smaller aperture, without having to use too high an ISO. But an even smaller aperture diffraction would then be a bigger problem. Plus a flash might be objectionable for other reasons (such as hot spots on the leaves of the tree that are closer to you... or a darker background in the distance behind the subject). And sometimes flash startles wildlife, too... Though, surprisingly, a lot will ignore the it (probably because they're accustomed to lightning). The sounds of the flash also may be a problem.
Other things that might help:
1. Look for a subject in better lighting, to allow you to increase DoF by stopping down a little more, without needing to use excessively high ISO or too slow a shutter speed. Just keep diffraction issues in mind.
2. Study up on "hyperfocal" focusing distances. Think of this as just simply making best use of what limited DoF you have with any given setup. For example, for this particular shot you might have had better results focusing on the critter's farther ear, instead of it's nose. That way the DoF that falls both nearer and farther from the point of focus might have been adequate for it's face and eyes, but encompassed more of it's body, too.
3. In post-processing, apply some sharpening to the slightly OOF areas. For example, a high pass filter sharpening method, applied carefully and selectively using a mask so that it only sharpens specific parts of the image, often works pretty well with furry and feathered critters.
4. It also sometimes might actually help to back off to take the shot... Or zoom to a shorter focal length... Or a little of both.... And then crop the image to the desired framing of the subject later in post-processing. More distance from the subject and/or a shorter focal length both will increase DoF. However, doing either or both of these, you'll have to weigh the loss of resolution against any gain in DoF.
5. Don't worry too much about it! The face and eye are the most critical point of focus anyway. Also, an image may have plenty of apparent sharpness when displayed in the final image size. It may be quite usable, though limited in how large you can print it.
A lot of people freak out reviewing their images at 100% on their computer monitors. Think about it. With most modern monitors and a camera with the resolution of the 7DII or your 5DIII, that's like making a five foot wide print and then judging it from 18" viewing distance... Well, of course it's going to look like crap!
While it's great to be able to zoom in to retouch images at 100% or more, most of us won't be making prints anywhere near that large or, even if we do, viewing them anywhere near that closely! So back off to 50% or even 33% or 25% for a much more realistic evaluation of sharpness and focus.
Finally, shooting with a set of gear over time, we learn what to expect simply through experience. Part of the "problem" here was that you were using a friend's camera and lens, stuff that you were less familiar with. It can be a long process to learn how things perform together through trial and error... So it might be very helpful to run a series of test shots and then study them, to get up to speed how things perform together. Maybe not with borrowed gear, but to ramp up with new stuff it can be helpful to shoot series of shots at different apertures, distances and focal lengths, just to get a feel for what DoF you might expect. This also can help you determine the optimal apertures of a lens, as well as the limitations of the less-than-optimal. Do test series with various ISOs, to study diffraction effects and more, too. In a sense, you're "calibrating" yourself so that there are fewer big surprises later.
However, once you are trying to "get the shot", other indirect factors such as shutter speed, lighting, ISO, diffraction, etc. start to come into play.
Smaller sensor cameras seem to have greater depth of field. They don't actually... If you shoot from the same distance with the same focal length and aperture, DoF will be exactly the same.
But, we don't do that. In order to frame a subject the same way with different sensor formats, we change focal length and/or our distance to the subject. In fact, in order to get that same shot with your 5DIII (assuming you were zoomed to 400mm on the 7DII) you would have had to use a 640mm lens. Or you would have had to move closer, say to about 7 foot distance. Or a little bit of both. And doing either of these would, in fact, decrease DoF. So, even though DoF is basically independent of sensor/film size, for all practical purposes due to the way we use them, smaller formats will render greater DoF. (Actually there's another factor that's related to sensor size and density, called the "circle of confusion"... which also relates to the size of the image as it will be viewed and even the distance from which it's viewed... but let's just put that aside for now, so as to not confuse the issue more than necessary!)
Here you're using a long focal length quite close with the lens aperture nearly wide open, so DoF is quite shallow. Using a DoF calculator, you'll find that it's .04 foot... or about 1/2 inch! (Here's an online calculator: http://www.dofmaster.com/dofjs.html).
Now, if you'd used a shutter speed half as fast (1/620), you could have closed the aperture down another stop (f9) to increase DoF a little bit... To .06 foot or roughly 3/4 inch. Still not very much to work with!
Increasing ISO a stop too, would have allowed you to use an even smaller aperture of f12.7, rendering a little more DoF... about .08 feet or roughly 1 inch. Better DoF, however you may not want to use such a high ISO and that small an aperture is going to start causing "diffraction", which will rob the image of some of it's fine detail. Like DoF, diffraction also isn't actually directly related to sensor size... But for practical purposes it is, because to create the same size final print from a full frame camera requires considerably less magnification of the image than it does from a crop sensor camera, and the less magnification that need be applied, the less obvious diffraction will be. So, diffraction can be a bigger concern with a crop sensor camera, than it is with FF.
Using a flash or other means of adding light might allow for slower shutter speed and an even smaller aperture, without having to use too high an ISO. But an even smaller aperture diffraction would then be a bigger problem. Plus a flash might be objectionable for other reasons (such as hot spots on the leaves of the tree that are closer to you... or a darker background in the distance behind the subject). And sometimes flash startles wildlife, too... Though, surprisingly, a lot will ignore the it (probably because they're accustomed to lightning). The sounds of the flash also may be a problem.
Other things that might help:
1. Look for a subject in better lighting, to allow you to increase DoF by stopping down a little more, without needing to use excessively high ISO or too slow a shutter speed. Just keep diffraction issues in mind.
2. Study up on "hyperfocal" focusing distances. Think of this as just simply making best use of what limited DoF you have with any given setup. For example, for this particular shot you might have had better results focusing on the critter's farther ear, instead of it's nose. That way the DoF that falls both nearer and farther from the point of focus might have been adequate for it's face and eyes, but encompassed more of it's body, too.
3. In post-processing, apply some sharpening to the slightly OOF areas. For example, a high pass filter sharpening method, applied carefully and selectively using a mask so that it only sharpens specific parts of the image, often works pretty well with furry and feathered critters.
4. It also sometimes might actually help to back off to take the shot... Or zoom to a shorter focal length... Or a little of both.... And then crop the image to the desired framing of the subject later in post-processing. More distance from the subject and/or a shorter focal length both will increase DoF. However, doing either or both of these, you'll have to weigh the loss of resolution against any gain in DoF.
5. Don't worry too much about it! The face and eye are the most critical point of focus anyway. Also, an image may have plenty of apparent sharpness when displayed in the final image size. It may be quite usable, though limited in how large you can print it.
A lot of people freak out reviewing their images at 100% on their computer monitors. Think about it. With most modern monitors and a camera with the resolution of the 7DII or your 5DIII, that's like making a five foot wide print and then judging it from 18" viewing distance... Well, of course it's going to look like crap!
While it's great to be able to zoom in to retouch images at 100% or more, most of us won't be making prints anywhere near that large or, even if we do, viewing them anywhere near that closely! So back off to 50% or even 33% or 25% for a much more realistic evaluation of sharpness and focus.
Finally, shooting with a set of gear over time, we learn what to expect simply through experience. Part of the "problem" here was that you were using a friend's camera and lens, stuff that you were less familiar with. It can be a long process to learn how things perform together through trial and error... So it might be very helpful to run a series of test shots and then study them, to get up to speed how things perform together. Maybe not with borrowed gear, but to ramp up with new stuff it can be helpful to shoot series of shots at different apertures, distances and focal lengths, just to get a feel for what DoF you might expect. This also can help you determine the optimal apertures of a lens, as well as the limitations of the less-than-optimal. Do test series with various ISOs, to study diffraction effects and more, too. In a sense, you're "calibrating" yourself so that there are fewer big surprises later.
Nov 8, 2015 13:44:56 #
Nov 8, 2015 15:12:45 #
boberic wrote:
At 400mm -f6.3 at 10 feet DOF is about 1/4 inch. BTW it's not a squirrel it's a chipmunk. Actually a good capture
OK so it's about 1/2 inch. I guessed at the 1/4. I was off by 1/4 inch I profoundly opologise. (I hope all my guesses are only off by 1/4 inch} Meanwhile I was right about the chipmunk, Maybe it was only 1/4 chipmunk
Nov 8, 2015 15:27:30 #
Nov 8, 2015 15:42:13 #
Apaflo wrote:
Reality check time folks. The numbers for DOF bei... (show quote)
Again profound apologies. I took the op at his word regarding 10 feet. I promise to never again take any one at their word.
Nov 8, 2015 16:05:25 #
Nov 8, 2015 20:43:47 #
Thanks to all who contributed here. I am humbled by what I haved learned.
1) The EXIF data is very useful. Many camera stores focal point and focus distance. (Thanks coj and apaflo)
2) There are esy to use DOF calculators on the PC and mobile applets.
3) Its of course a chipmunk, not a squirrel. FOCUS dude.
Thanks to all that liked the capture.
1) The EXIF data is very useful. Many camera stores focal point and focus distance. (Thanks coj and apaflo)
2) There are esy to use DOF calculators on the PC and mobile applets.
3) Its of course a chipmunk, not a squirrel. FOCUS dude.
Thanks to all that liked the capture.
Nov 8, 2015 20:58:40 #
Amfoto1: Thanks for your detailed e-mail. I will learn and try several of the improvements you recommend.
I have a question about DOF though. The on-line DOF calculator requires the camera type to be entered. If you switch between 7D and 5D, for example, you get a different result.
From what I understand that is because of the different 'Circle of Confusion size', which is a function of sensor size. Pixel (sensel) size does not matter.
Again, thanks for your detailed idea's on how to improve my technique. I do pixel peep too much. I am starting to print more of my pictures just to see how they "really" come out, compared to on the monitor.
I have a question about DOF though. The on-line DOF calculator requires the camera type to be entered. If you switch between 7D and 5D, for example, you get a different result.
From what I understand that is because of the different 'Circle of Confusion size', which is a function of sensor size. Pixel (sensel) size does not matter.
Again, thanks for your detailed idea's on how to improve my technique. I do pixel peep too much. I am starting to print more of my pictures just to see how they "really" come out, compared to on the monitor.
Nov 8, 2015 21:00:32 #
boberic wrote:
Again profound apologies. I took the op at his word regarding 10 feet. I promise to never again take any one at their word.
And I'll promise to look at my EXIF data before I make any public statements again.
Nov 8, 2015 21:48:47 #
Dec 19, 2015 11:32:13 #
Rongnongno wrote:
There is no problem here. While the DOF is shallow it serves the purpose.
Shot and killed four of the grey kind in three days, one this very morning. Way to many around my house - and do not feed them other than with led, trust me.
Shot and killed four of the grey kind in three days, one this very morning. Way to many around my house - and do not feed them other than with led, trust me.
:thumbdown:
Dec 20, 2015 00:36:31 #
greg vescuso
Loc: Ozark,Mo.
I think if you take the time to calibrate this lens to this camera you will find it is slightly front focusing. If that bothers you because it is a very minor amount.
Dec 20, 2015 19:01:29 #
idahoshooter wrote:
This shot is with a Canon 7D mII, Canon EF 100-400L II.
1/1250s, f/6.3, ISO 6400.
I was shooting fast because I was trying to get some BIF, then saw this squirrel munching on the tree leaves. He was only 10' away or so, but I'm surprised how little DOF I have. The forehead is in focus, but not the rest of the body.
a) small depth of field @400mm, relatively close
b) maybe I was slightly front focused to start with?
1/1250s, f/6.3, ISO 6400.
I was shooting fast because I was trying to get some BIF, then saw this squirrel munching on the tree leaves. He was only 10' away or so, but I'm surprised how little DOF I have. The forehead is in focus, but not the rest of the body.
a) small depth of field @400mm, relatively close
b) maybe I was slightly front focused to start with?
I just checked the DOF chart at these settings. At 10 feet the DOF is .04 ft or less than 1/10 inch. So- the shot of the chipmunk is about as good as can be. Correction- DOF at 10 ft is about 1/2 inch
If you want to reply, then register here. Registration is free and your account is created instantly, so you can post right away.