Why a Poor Depth of Field?

[Michele0513]

Hello, and thanks in advance to any who answer

I have the Canon Rebel T3i 1100D with the 18-55mm kit lens, and for some reason am getting a really shallow depth of field in all my pictures. I have taken close to 1000 shots in every mode from ADEP to full auto in more "test" situations than would be practical to list, but the photos are still bad.
Here's one I shot this morning, and yes I know that it's not a great picture in terms of composition and all. I just wanted to show a quick example of the D.O.F. issue.

* The pup's head is approximately 3" wide, the ear-to-ear distance maybe 6".
* Distance to subject was around 2.5' (maybe a little more)
* Taken in Aperature Priority mode with a setting of 1/16
* Focal Length 37mm
According to the Depth of Field calculator I use, I should have been able to get 6.2" of focused depth at only 2 foot away (it showed a near/far measurement of 1'9.3"/2'3.5"), so it would be a little deeper for the extra inches away the pup was at. But as you see, barely the 3" head width is in focus, and that's spotty.

A screen shot of Canon's DPP software screen shows the active focal points in red, and most (though not all) of what's within those points is pretty much in focus, but *where* is the 6"+ depth of field?

 

[happysnapper64]

Michele0513 wrote:

I have read the manual, both when I got the camera in December and then in spots as needed since then. There is no such preview button on the 1100D unless I've REALLY missed something.

Michele. In your original post you refered to your camera as a T3i - 1100D. The T3i or 600D as it's known as in Europe, does have a DOF button. It's on the body near the lens release button. Some posters may think you have the T3i.

 

[Dareshooter]

You won't see anything happen until you stop the lens down to a smaller aperture F8-22 e.g. Look through the viewfinder ,push the DOF button in and change the aperture from say F 5.6 to F16 and you'll see the viewfinder image go noticeably darker and the zone of sharp focus becomes greater.Best try it out in bright conditions because the darkening of the viewfinder makes assessing DOF rather difficult even for those with good vision.

 

[Dareshooter]

Michele0513 wrote:

I have read the manual, both when I got the camera in December and then in spots as needed since then. There is no such preview button on the 1100D unless I've REALLY missed something.

You haven't missed anything Michele,the 1100D/T3 does not have a DOF preview button at the side of the lens throat.That's the bad news. :-( Good news is that you can assign the set button to DOF preview via the custom settings menu

 

[happysnapper64]

Dareshooter wrote:

You won't see anything happen until you stop the lens down to a smaller aperture F8-22 e.g. Look through the viewfinder ,push the DOF button in and change the aperture from say F 5.6 to F16 and you'll see the viewfinder image go noticeably darker and the zone of sharp focus becomes greater.Best try it out in bright conditions because the darkening of the viewfinder makes assessing DOF rather difficult even for those with good vision.

Thank's Dareshooter. Maybe my eyes are not that bad after all.:-D

 

[dsjtecserv]

Michele0513 wrote:

Well shoot, I can't see it changing one bit! Just to make sure I wasn't supposed to hold down any buttons or anything: all I did was put it in Manual and then use the menu to click/scroll across the aperture readings (while staring "down the barrel" of course). Was there anything else I should have done?

If not, then I guess the EXIF data is showing that I'm shooting at different f-stops, but the lens isn't actually doing it. Any fix for that?

The iris will not close down until you actually take a picture, or unless you press the depth of field preview button. To do what was suggested, you need to find the preview button, turn the camera so you can look down the lens (good light helps), press the preview button and turn the main dial to change the aperture.

Honestly, if the iris were not closing at all, you would have poorly exposed pictures, which you don't. So I think this is an issue.

Dave

 

[Guidenet]

dsjtecserv wrote:

Michele0513 wrote:

Well shoot, I can't see it changing one bit! Just to make sure I wasn't supposed to hold down any buttons or anything: all I did was put it in Manual and then use the menu to click/scroll across the aperture readings (while staring "down the barrel" of course). Was there anything else I should have done?

If not, then I guess the EXIF data is showing that I'm shooting at different f-stops, but the lens isn't actually doing it. Any fix for that?

The iris will not close down until you actually take a picture, or unless you press the depth of field preview button. To do what was suggested, you need to find the preview button, turn the camera so you can look down the lens (good light helps), press the preview button and turn the main dial to change the aperture.

Honestly, if the iris were not closing at all, you would have poorly exposed pictures, which you don't. So I think this is an issue.

Dave

 

[Brev00]

Let's forget about depth of field calculations for now and just look at the reality of this one shot. What I see looks perfectly normal for a dslr shot given that you are so close to your subject. Several factors influence depth of field (as you know) beyond just the aperture. Anytime you shoot a portrait at close distance and focus on the subject, you will blur the background. Even at f 16. At f 5.6, less will be in the dof window. To eliminate lens issues, you should shoot the exact same scene from a variety of apertures.

To maximize your depth of field, you have a few options. Shoot with a 4/3 camera or one with an even smaller sensor. Compacts are great at getting the dof you want. Moving to a dx or full frame dslr is often initially frustrating because of the naturally larger amount of blur that results. To improve the dof of this one particular shot with your camera, you still have some more choices. I would try shooting from further back. The closer you are, the smaller the dof. I might also use a wider angle. The wider the angle, the more dof. Now, the dog will be a smaller part of your frame, but you can crop to get the right proportions and you will have more dof. It is a trick but narrow dof is a cruel master and sometimes you have to outwit it.

 

[AnthonyL]

Michele0513 wrote:

I have read the manual, both when I got the camera in December and then in spots as needed since then. There is no such preview button on the 1100D unless I've REALLY missed something.

Well you've missed something but not REALLY missed something. :-D

http://www.imaging-resource.com/PRODS/T3/T3A.HTM

There's no longer a dedicated depth-of-field preview button on the T3; something that Canon probably rightly assumes most consumer users wouldn't use anyway. You can, however, configure the Set button on the rear panel to serve as a depth-of-field preview button.

Personally I agree with getting rid of it. It never worked well for me. You can easily see if aperture is changing by changing the aperture in Live View. Note that Live View contrast detection focus is the most accurate you can get (just referring to something you said somewhere else).

Canon T3 Review - Imaging Resource

by Shawn Barnett, Dave Etchells, Zig Weidelich, and Mike Tomkins Hands-on Preview Posted: 02/07/2011 Initial Test: 08/07/2011 Review Posted: 09/23/2011 A major revamp to their previous Rebel XS entry-level SLR, the new Canon T3 is significantly more advanced than its predecessor, and looks to...

www.imaging-resource.com

 

[dsjtecserv]

Michele0513 wrote:

Hello, and thanks in advance to any who answer

I have the Canon Rebel T3i 1100D with the 18-55mm kit lens, and for some reason am getting a really shallow depth of field in all my pictures. I have taken close to 1000 shots in every mode from ADEP to full auto in more "test" situations than would be practical to list, but the photos are still bad.
Here's one I shot this morning, and yes I know that it's not a great picture in terms of composition and all. I just wanted to show a quick example of the D.O.F. issue.

* The pup's head is approximately 3" wide, the ear-to-ear distance maybe 6".
* Distance to subject was around 2.5' (maybe a little more)
* Taken in Aperature Priority mode with a setting of 1/16
* Focal Length 37mm
According to the Depth of Field calculator I use, I should have been able to get 6.2" of focused depth at only 2 foot away (it showed a near/far measurement of 1'9.3"/2'3.5"), so it would be a little deeper for the extra inches away the pup was at. But as you see, barely the 3" head width is in focus, and that's spotty.

Having now read through all the replies, I'm surprised that no one has given the response I'm about to give: there is likely nothing wrong with your camera or with your technique, but your expectations may be skewed a little. A key element of the concept of depth of field is that only objects that are at the precise focus distance from the camera are in "perfect" focus; objects nearer and farther are increasingly blurred the further they are from this exact plane of focus. Here how this likely played out in your pciture:

I get the same depth of field results that you did using my calculator, so I agree that you had about 2.5 inches in front of the focus point and about 3.5 inches in back. But remember that these are relative to the actual point that the camera focused on, which was the dog's cheek to the right of the eye. Remember too that the actual sharpness of the image decreases progressively in front of and behind this point. If you look closely and critically, you can see that about 1.5 inches of the dog's near ear flap is "very sharp" to "pretty sharp", but declines noticeably beyond that. Similarly, about 3 inches of the top of the dog's head appear to be reasonably sharp, but fade to increasing blur beyond that distance.

I'd say that pretty consistent with what the DoF calculator indicates. You need to remember that some of your depth of field is in front of the actual focus point, so it can't help with getting things behind the focus point in focus. And even when something is considered "in focus" according to a calculator, that doesn't mean is is perfectly sharp; it actually means that it is just acceptably sharp, and things that are near the close and far limits will be just barely acceptable.

What this means is that your picture is just about optimum for this subject. For the same framing of the subject, it doesn't matter what focal length or subject distance you use; the depth of field will be about the same. You really can't stop down further than f/16 without increasing the blur due to diffraction throughout the picture. You could try to focus just a little further away (up the dog's forehead a bit) in order to move the zone of acceptable focus further back, although this would make the near ear more blurry. But frankly, I like it just the way it is; I think the balance of near and far focus is just about optimum and the focus follows the line of his/her muzzle just about perfectly. I think you should declare victory!

Dave

 

[Draek]

DoF calculations are based around the concept of "circle of confusion", which is roughly the largest blur circle that can be called "reasonably sharp" on a normal-sized print viewed at a normal viewing distance. In other words, it is not meant for pixel-level 100% viewing since at that scale, as you've already seen DoF is much, much less.

Fortunately you're in luck: the solution doesn't require you to spend much money on equipment, and certainly not to replace your camera and lens outright. It's simply... well, get a tripod, and stop down even further, f/22 or even f/32. That or you could use focus stacking, though be aware that also necessitates a tripod so the images align correctly; it's what most professional product photographers do, and though it's considerably less convenient than just stopping down further, the quality of the end-product is generally superior as it doesn't suffer from diffraction as much.

 

[JosephScha]

I almost hate to say this. If you want more depth in good focus, get a camera with a smaller sensor. I'm not sure what DSLR you have, but if it's FF then:

APSc - crop factor of 1.5 or 1.6. So you'd use a 36mm lens instead of a 55mm lens.

4/3 (as in micro four thirds cameras) - crop factor of 2. So you might use a 25mm lens.

Various point and shoots: I used to have an FZ7 which had a 6 to 72mm lens, which was the equivalent of 36mm to 432mm on full frame. That means a p&s or superzoom like the FZ7, with a 1/2.5" sensor, has a "crop factor" of 6 (although no one talks about it that way), so instead of a 55mm lens you would use its lens set to 9mm.

The shorter the focal length, the more DOF. Let me assure you that with a 9mm focal length you will get much more range in good focus. On top of that f/8 or so on such a tiny focal length is really a small opening (about 1mm), close to a pinhole, which also increases DOF.

But I don't expect you to actually take this advice. I think if I were you, the next thing I'd do is investigate the focus accuracy of the lens and camera that you use. Most DSLRs allow you to set focus "micro adjust" to compensate for possible focus errors. I suspect that the in-focus region is shifted, so only part of it is on your subject. This is a known possibility when you have separate phase detect focus sensors, but usually it's not a significant issue. There's even a commercial tool (although I suspect you could make one yourself) to check for it:

http://spyder.datacolor.com/portfolio-view/spyderlenscal/

 

[Michele0513]

Hi all!
Ok, I spent the day experimenting, and it turns out that pretty much all of you are right. First off, with the lack of a DOF preview button (seriously guys, there isn't one on this model) and my eyes' inability to see anything in there moving no matter what method I tried, I finally just took the advice about taking identical test shots. From f/3.5 to f/16, there are differences in the depth of field of the shots, but from f/16 to f/32 there is no increase in DOF at all and the overall shots seem blurry. This was consistent for ones taken on a tripod outdoors with adequate light, at distances ranging from 3 feet to 15 feet from the subject, as well as a set taken in a home made ligthbox from a distance of about 2 feet. Sooo, maybe something's wrong with the lens to a point, but f/3.5 to f/16 is still a pretty decent range to work with so that leaves me with good old "User's Lack of Skill" here.


To the people who said I may have the wrong expectations from the equipment, yes, absolutely. I really wasn't prepared to see the smaller areas of focus in my pictures, but when I browsed some of the Galleries on here earlier I realized that the pictures I liked most were the exact same way. Time to readjust my way of thinking in regards to my own photos, I guess.
So basically, I've printed out all the photo-taking suggestions you all have offered, and I'm going to be doing a lot more experimenting to see how well I can learn to work with what this camera & lens can do.


One thing happened today with the set I mentioned taking in the light box that I'd like to be able to better understand. A picture (in this case 2) will explain better than I can, so please forgive me for muddling up the forum with yet another...

These were taken using a tripod, and I was focusing on the number 21 on the ruler. Since I was just trying to compare aperture settings instead of going for any great exposure or lighting, I had the camera in AP mode and let it pick the ISO and shutter speed. To use the high/low comparison, what I noticed is...


In the shot taken at f/5.6:
Only a middle area of the ruler's marks and numbers were in reasonable focus, with a few being pretty sharp (like 20-22). None of the stones were much in focus, even the one next to the sharply focused numbers/marks.


In the f/16 shot:
Almost all of the marks & numbers were in reasonable focus, but none were as sharp as the few that stood out in the 5.6 shot (again, 20-22). ALL of the stones were in better focus.


I can work with this info "as is" for future shots, but I'd really like to have a better understanding of why they turned out like that. Because of what happened with the stones, I wonder is it normal to sacrifice a slight amount of sharpness in a photo when you work with smaller apertures? I noticed the same thing in the outdoor shots... the smaller apertures I used, there was a little less clarity. Also, why would the stones be better focused in one shot than the numbers were, but the exact opposite happen in the other shot? Is this because the stones were taller and thus closer to the camera?

And on a different subject... are there any adapters available that would let me use any of my old school lenses with this new camera? It's a real bummer to not be able to afford any other lenses for it when there's everything from a basic 2x multiplier on up to a pricey 200mm telephoto packed away in the closet!

 

[Sante Patate]

Michele0513 wrote:

In the shot taken at f/5.6:
Only a middle area of the ruler's marks and numbers were in reasonable focus, with a few being pretty sharp (like 20-22). None of the stones were much in focus, even the one next to the sharply focused numbers/marks.

That isn't what I see. In the f/5.6 shot the reflection of the window frame on the middle rock is at least as sharp as the corresponding window frame at f/16.

In the f/16 shot:
Almost all of the marks & numbers were in reasonable focus, but none were as sharp as the few that stood out in the 5.6 shot (again, 20-22). ALL of the stones were in better focus.

The depth of field is greater at f/16, so more of the numbers are in focus. But at f/16 the sharpness is reduced by diffraction. Resolution is greatest when aperture is largest, and as the aperture gets smaller resolution gets worse. But ... a number of the lens aberrations that degrade the image are worse at large apertures. So there is a race as the aperture gets smaller between aberrations getting better and resolution getting worse. The best aperture for lenses is often around 2 stops less than the biggest.

 

[Michele0513]

Oh sorry, my bad

 

[Michele0513]

Thanks Dave, that makes me feel a lot better

 

[MarkInSF]

Michele0513 wrote:

Hi all!
Ok, I spent the day experimenting, and it turns out that pretty much all of you are right. First off, with the lack of a DOF preview button (seriously guys, there isn't one on this model) and my eyes' inability to see anything in there moving no matter what method I tried, I finally just took the advice about taking identical test shots. From f/3.5 to f/16, there are differences in the depth of field of the shots, but from f/16 to f/32 there is no increase in DOF at all and the overall shots seem blurry. This was consistent for ones taken on a tripod outdoors with adequate light, at distances ranging from 3 feet to 15 feet from the subject, as well as a set taken in a home made ligthbox from a distance of about 2 feet. Sooo, maybe something's wrong with the lens to a point, but f/3.5 to f/16 is still a pretty decent range to work with so that leaves me with good old "User's Lack of Skill" here.

To the people who said I may have the wrong expectations from the equipment, yes, absolutely. I really wasn't prepared to see the smaller areas of focus in my pictures, but when I browsed some of the Galleries on here earlier I realized that the pictures I liked most were the exact same way. Time to readjust my way of thinking in regards to my own photos, I guess.
So basically, I've printed out all the photo-taking suggestions you all have offered, and I'm going to be doing a lot more experimenting to see how well I can learn to work with what this camera & lens can do.

One thing happened today with the set I mentioned taking in the light box that I'd like to be able to better understand. A picture (in this case 2) will explain better than I can, so please forgive me for muddling up the forum with yet another...

These were taken using a tripod, and I was focusing on the number 21 on the ruler. Since I was just trying to compare aperture settings instead of going for any great exposure or lighting, I had the camera in AP mode and let it pick the ISO and shutter speed. To use the high/low comparison, what I noticed is...

In the shot taken at f/5.6:
Only a middle area of the ruler's marks and numbers were in reasonable focus, with a few being pretty sharp (like 20-22). None of the stones were much in focus, even the one next to the sharply focused numbers/marks.

In the f/16 shot:
Almost all of the marks & numbers were in reasonable focus, but none were as sharp as the few that stood out in the 5.6 shot (again, 20-22). ALL of the stones were in better focus.

I can work with this info "as is" for future shots, but I'd really like to have a better understanding of why they turned out like that. Because of what happened with the stones, I wonder is it normal to sacrifice a slight amount of sharpness in a photo when you work with smaller apertures? I noticed the same thing in the outdoor shots... the smaller apertures I used, there was a little less clarity. Also, why would the stones be better focused in one shot than the numbers were, but the exact opposite happen in the other shot? Is this because the stones were taller and thus closer to the camera?

And on a different subject... are there any adapters available that would let me use any of my old school lenses with this new camera? It's a real bummer to not be able to afford any other lenses for it when there's everything from a basic 2x multiplier on up to a pricey 200mm telephoto packed away in the closet!

Yes, it is absolutely normal for very small apertures to cause increasing blurriness. This is due to diffraction. Cameras with smaller sensors (smaller photosites, more specifically) start showing the effects of diffractions sooner. On my Nikon 1, diffraction has already started blurring the image (very slightly) even at f/8, but that camera has a smaller sensor than yours does. You'll be getting some detectable softening at f/16 and a lot at the smaller apertures. There is a reason why many lenses top out at f/22. Sometimes choosing a small aperture is necessary to get a lot of depth of field and you just have to suffer the consequences. Pinhole lenses are an example. Their tiny apertures give massive depth of field but the images are very soft.

Your rocks are three dimensional. They aren't in focus in the first shot because they are actually quite a bit closer than the part of the ruler that's in focus. In the second shot more are sort-of in focus because you're using a smaller aperture, but diffraction is making everything soft.

Way back at the beginning you were also concerned that areas out at the edges of your images were blurry and you thought that was a problem. It is a problem, but one common to almost all lenses, and zooms more than primes. Lenses are almost always sharpest in the middle, reasonably sharp most of the way out towards the sides, but soft at the edges and sometimes very soft in the corners.

However, this is also affected by aperture. Almost all lenses are softer at their maximum (largest) aperture (like f/2 or something) , and that's even more true in the corners. Stopping down a couple of stops will get many lenses to their sharpest aperture, so an f/2.8 lens (maximum) will be sharper at f/5.6, in the center somewhat, but a lot in the corners. This is independent of depth of field and has nothing to do with diffraction, it's just one of the design tradeoffs lensmakers have to consider. They can't make a lens that's perfectly sharp in all situations. Lenses that are super sharp in very specific circumstances are expensive and specialized, sold to people with unusual needs.

For you, the photographer, it's also a matter of tradeoffs. Use a large aperture, get a faster shutter speed, get shallower dof, but give up sharpness, especially in the corners. Stop down a couple of stops, improve the sharpness, especially in corners, but dof increases (which may be either good or bad) and shutter speed is slower, not good for action and harder to hand hold. Stop down even further and dof increases more, exposures get longer yet, and diffraction starts to be an issue. It's always a balancing act. Some people like to start in the middle, with the lens stopped down a couple of stops to where it is sharpest, then work from there, making changes for specific reasons only, like wanting more or less dof.

 

[Michele0513]

Ok, I see. I'd read about diffraction before, but now that can see exactly what it does in my shots, it makes more sense. Thank you

 

[Daisy AU]

Dareshooter wrote:

Michele0513 wrote:

I have read the manual, both when I got the camera in December and then in spots as needed since then. There is no such preview button on the 1100D unless I've REALLY missed something.

You haven't missed anything Michele,the 1100D/T3 does not have a DOF preview button at the side of the lens throat.That's the bad news. :-( Good news is that you can assign the set button to DOF preview via the custom settings menu

Yes,

Have a look at how to do it here:

http://arnowelzel.de/wiki/en/blog/20120720-01-eos1100d-blendenvorschau

 

[Michele0513]

Thank you, I really appreciate all the info. You can't imagine how nice it is to feel like I'm actually understanding all of this rather than to just be going "Umm... ok I guess I get it. Kinda." (lol)

Somebody earlier said that this lens is at it's best at around f/8, so combined with what you've said I think I'll just try to work my pictures loosely around that f-stop for a while to get a better feel for it in different situations. That might be easier for me for now than going willy nilly through all the different modes and every extreme in settings all the time.

 

[Michele0513]

Oooh thank you, Daisy! You've saved me some Googling! I wonder though if that will disable my Set button when I'm in Q mode so it won't be able to be used to change the aperture and all?