Field of view/focal point

L

lionofjudah

Suspended / Banned
Messages
204
Edit My Images
No
My question is if you focus at distance A and want the background blurred which you have done is everything on the plane at distance A in focus or do you get a circle of clarity tapering off into blur outside it.
I'm not talking of blur in front of behind and image but to the sides. Say I'm taking a photo of a group of 20 people in a straight line and focus on the middle person how can I be sure that person 1 and 20 will be clear also yet try to keep background/foreground relatively blurry.

I hope that makes sense
 
The zone of sharpness is a flat plane running parallel to the sensor. It extends horizontally and vertically to the edges of the image. Depth of field extends in front and behind that plane and can be varied. If you photograph a group of people standing in a straight line along that plane, they should all be sharp.

However, that is theory. In practise, all lenses are naturally sharper in the centre than around the edges. The best lenses reduce this to a minimum so it's not noticeable, but not all lenses are like that, particularly if you use them at their lowest f/number - they will be better towards the higher end of the range. Some lower quality wide-angle zooms are not always as sharp as they could be. Wide-angles will also tend to distort towards the edges, pulling people out of shape and making them look fatter. This is not a fault.

Rule of thumb is, when shooting groups, keep people away from the edges and corners. Image quality issues that might otherwise pass unnoticed, show up because we're naturally drawn to look closely at faces.
 
lionofjudah said:
My question is if you focus at distance A and want the background blurred which you have done is everything on the plane at distance A in focus or do you get a circle of clarity tapering off into blur outside it.
I'm not talking of blur in front of behind and image but to the sides. Say I'm taking a photo of a group of 20 people in a straight line and focus on the middle person how can I be sure that person 1 and 20 will be clear also yet try to keep background/foreground relatively blurry.

I hope that makes sense
Click to expand...
It depends on the specific lens design. Some have a circular field of focus, some have a straight line, many something in between. Note too that many lenses have sharper central focus than the edges can ever achieve. If you want an even edge to edge sharpness you may have to stop down to an aperture where the centre has lost enough sharpness due to diffraction to match the best sharpness the edges can come up with.

I'm afraid the simplest answer to your question will be for you to spend some time photographing lines of sticks pushed into the grass of a convenient lawn. Short grass gives you a good indication of where the focus has gone when it's not where you expect it to be.

Note that if you're using a wide rectilinear lens to photograph a group that the people on the edges will look fatter. That's a natural consequence of the rectilinear perspective projection used by those lenses. They present straight lines in the world as straight lines in the image. An inevitable side effect of that is to widen subtended visual angles near the edges of the image. Photographs which make people look fatter are not popular. You can get round that problem by using a cylindrical perspective projection, which is what panoramas do, and is a perspective adjustment some image editors (such as DXO) can do, or by using a spherical perspective lens, aka fisheye. With a fisheye lens you can arrange your large group radially equidistant around the camera and they'll all appear in correct proportions, because that perspective projection preserves subtended angles at the cost of curving straight lines near the edges.
 
Short focal length lenses are not Ideal for taking large groups because of the factors that people have mentioned.
It is far better to move your viewpoint further away and use a standard or longer focal length lens.
such a lens and distance naturally produces less distortion of all kinds and a narrower depth of field.
The alternative is to use a cirkut camera as used for school photographs. For which you place the group in a curve with a radius centred on the camera.
However neither solution allows for much of a blur of backgrounds
 
Last edited:
portrait was just an example I was giving. but your answers make a lot of sense. normally I tend to use my kit lens of 18-105 a lot.
I was just thinking of the question as sometimes ive looked a photo and while one face looks clear the other one is not 100% sharp, so was just thinking of why that could have been and what I could do to improve.
thanks for all your replies....off to find some sticks :-)
 
lionofjudah said:
portrait was just an example I was giving. but your answers make a lot of sense. normally I tend to use my kit lens of 18-105 a lot.
I was just thinking of the question as sometimes ive looked a photo and while one face looks clear the other one is not 100% sharp, so was just thinking of why that could have been and what I could do to improve.
thanks for all your replies....off to find some sticks :)
Click to expand...
AFAIK, only some of the high end macro lenses have truly "flat field focus" w/ the same depth of focus across the entire image. But I highly doubt that is a significant factor in your images... the most likely issues are too little DOF, or too slow of a SS.
 
As Richard said, the design aim for a camera lens is to have a plane of focus that's a straight line.

In reality though, it's rarely that precise, as it's not really a significant factor in most photography.

Macro lenses are an exception as they generally have a very flat plane of focus making them ideal for copy work.

But as above, it's rarely an important factor, and in general use if I'm photographing a couple and want both of them sharp with as small a DoF as possible, the technique is to simply ensure that both pairs of eyes are the same distance from the lens.

No matter how large the DoF of an image, there is always one plane of focus, and so if you want 2 people 'in focus' they both need to be the same distance, there will always be a discernible OoF result to anything not in the plane of focus.
 
Phil is talking about field curvature, when the plane of focus is not flat as intended, but slightly curved - usually inwards but can also be slightly wavy. It's something talked about in lens reviews but, rarely seen to any significant extent (beyond very close distances) in a modern lens.* It's possible to misinterpret Phil's post when he says to keep subjects the same distance from the lens - that would position a group of people in a curved line. They should be in a straight line, and those at the ends of the line would therefore be further from the camera.

sk66 suggests too little depth of field, or too slow a shutter speed. It can't be too little DoF if the line is parallel to the camera, and in any case it would be hard to get too little DoF with a long line of people with the OP's lens. Too slow a shutter speed would blur the whole image, not just the edges. My guess is it's just a modest quality lens, not used at its best settings.

*If anyone has a clear example of field curvature taken at normal shooting distance, I'd be very interested to see it. That's not a challenge - I really would :)
 
Yes, please read 'same distance from lens' as 'same distance from
The plane that the sensor occupies.

Btw, that's what the symbol means on top of your camera, the circle with the line running through it, that line is the plane of the sensor.
 
HoppyUK said:
Too slow a shutter speed would blur the whole image, not just the edges.
Click to expand...
The situation was an image of multiple people; they move...
HoppyUK said:
*If anyone has a clear example of field curvature taken at normal shooting distance, I'd be very interested to see it. That's not a challenge - I really would :)
Click to expand...
Pretty much any WA/UWA lens will show it to some extent. The typical "unsharp corners" comment is normally due to field curvature, and not due to an inherent inability to focus at the edges.
 
sk66 said:
The situation was an image of multiple people; they move...
Click to expand...

:)

Pretty much any WA/UWA lens will show it to some extent. The typical "unsharp corners" comment is normally due to field curvature, and not due to an inherent inability to focus at the edges.
Click to expand...

Disagree. If that was the case, you would be able to refocus and get sharp edges, or stop down a bit more so DoF covered it - but you can't...

Edit: the field curvature test I use is a large area of block-paving where there is a nice regular pattern right across the frame. Easy to check different distances, focal lengths etc, and you can see both fore and aft of the sharp zone. I've yet the see anything where I can definitely say, that is field curvature and it's a problem.
 
Last edited:
HoppyUK said:
*If anyone has a clear example of field curvature taken at normal shooting distance, I'd be very interested to see it. That's not a challenge - I really would :)
Click to expand...

When I had DSLR's I had a Sigma 30mm f1.4 and a very nice lens it was :D I read criticism of this lens and its field curvature on internet forums but I can honestly say that I never ever saw anything that made me point and say "Ah Ha! There's that field curvature!" Not once, not even at f1.4.

Personally I wouldn't even consider field curvature.
 
Last edited:
HoppyUK said:
:)



Disagree. If that was the case, you would be able to refocus and get sharp edges, or stop down a bit more so DoF covered it - but you can't...

Edit: the field curvature test I use is a large area of block-paving where there is a nice regular pattern right across the frame. Easy to check different distances, focal lengths etc, and you can see both fore and aft of the sharp zone. I've yet the see anything where I can definitely say, that is field curvature and it's a problem.
Click to expand...
http://phototechmag.com/field-curvature/
 
HoppyUK said:
:)

Disagree. If that was the case, you would be able to refocus and get sharp edges, or stop down a bit more so DoF covered it - but you can't...
Click to expand...
Maybe you can't, but I've got at least one lens where I can. I'll have to try and remember which one it is. :-)
Edit: the field curvature test I use is a large area of block-paving where there is a nice regular pattern right across the frame. Easy to check different distances, focal lengths etc, and you can see both fore and aft of the sharp zone. I've yet the see anything where I can definitely say, that is field curvature and it's a problem.
Click to expand...

I use a lawn. The sharpness of the grass blades describes the straightness (or otherwise) of the "plane" of focus.
 
Graham W said:
And I thought that I was the only one outside of the film crowd that knew that! :)
Click to expand...

Here I though it was symbol for TFL.

great discussion. Thanks guys.
Might have been too slow a shutter speed possible to pick up mild movement from 1 person hence inducing mild blur. I'll have to find those images and check.
 
sk66 said:
The situation was an image of multiple people; they move...

Pretty much any WA/UWA lens will show it to some extent. The typical "unsharp corners" comment is normally due to field curvature, and not due to an inherent inability to focus at the edges.
Click to expand...

sk66 said:
http://phototechmag.com/field-curvature/
Click to expand...

I'm not suggesting for a moment that field curvature doesn't exist, but it's very rare for it to be any kind of significant problem at normal shooting distances and I disagree that unsharp corners are normally down to field curvature. If it was that common, and that noticeable, then I think we'd all be quite rightly up in arms and it'd be top of every lens reviewer's check list. I'm aware of the Lloyd Chambers article, but find it unconvincing (and his suggested test method with a sheet of newspaper is ill-considered). Whenever I've googled this, articles like that come up with a couple of sample images that conveniently seem to fit the theory in retrospect and make a good blog post. They often involve old, or extreme or obscure lenses, but without any conclusive proof.

If field curvature is a real and common problem suffered by most wide-angles as you say, then it should be easy to come up with some comparison images that show a) a lack of sharpness where it should be sharp, and then b) the same image refocused with the sharp and unsharp zones reversed. That would be proof, but I don't think I've seen it done. I've tried myself without success.
 
HoppyUK said:
If field curvature is a real and common problem suffered by most wide-angles as you say, then it should be easy to come up with some comparison images that show a) a lack of sharpness where it should be sharp, and then b) the same image refocused with the sharp and unsharp zones reversed. That would be proof, but I don't think I've seen it done. I've tried myself without success.
Click to expand...
Definitely agree it's not typically problematic... modern lenses are generally so well corrected that I've never even checked a lens for it, and I've never noticed an issue w/ it other than when I had a lens defect. But I don't normally care about edges/corners much, and in the rare occasion when I do I'm probably using specialty lenses (Macro/TS).

Another reason why it is hard to detect is because lenses can seldom focus quite as sharp at the edges, so when you focus for maximum sharpness at the edges, instead of "reversed" you end up with a slightly softer image overall. It would take pretty major curvature to be able to "reverse" the focus... apparently there are lenses that have it that bad, but I don't own one (or if I do, I don't want to know).
 
sk66 said:
Definitely agree it's not typically problematic... modern lenses are generally so well corrected that I've never even checked a lens for it, and I've never noticed an issue w/ it other than when I had a lens defect. But I don't normally care about edges/corners much, and in the rare occasion when I do I'm probably using specialty lenses (Macro/TS).

Another reason why it is hard to detect is because lenses can seldom focus quite as sharp at the edges, so when you focus for maximum sharpness at the edges, instead of "reversed" you end up with a slightly softer image overall. It would take pretty major curvature to be able to "reverse" the focus... apparently there are lenses that have it that bad, but I don't own one (or if I do, I don't want to know).
Click to expand...

That's a more measured view :)

The reason I think field curvature sometimes gets overplayed as a problem, is not because it doesn't exist but because it's usually very mild, often the least of the problems (as you describe) and doesn't matter. We often want and expect some areas of the subject to be sharp and unsharp, it's a key creative technique, and it's usually best when the sharp zone is reasonably flat, but if it's a bit curved or slightly wavy then that's fine because the things we photograph are also irregular. It passes unnoticed.

It only becomes a problem when lens test methods involve a small flat target, never more than six feet wide. That means an artificially close shooting distance with wide-angles like 1m where a) some lenses do not perform at their best, b) field curvature is usually worst, and c) because it's flat, it shows up slight field curvature errors as a significant loss of sharpness. It's nonsense, but the problem is the test method, not the lens. My own MTF lens test procedure uses a wide-angle test target that is effectively 20ft wide and all test areas of the frame are focused and shot individually. It's very laborious, but allows all lenses to show what they can really do, especially on a 105mp test camera.
 
sk66 said:
I don't think I ever said it was likely a significant issue...

BTW, this article shows a lens that has it bad enough to reverse the focus a bit... but it doesn't say what lens it is (and it could be faked FAIK).
Click to expand...

Maybe that's just how I read it ;)

Link not working for me Steven.
 
HoppyUK said:
*If anyone has a clear example of field curvature taken at normal shooting distance, I'd be very interested to see it. That's not a challenge - I really would :)
Click to expand...

Shift lenses (TS-E, PC-E etc) have field curvature out of necessity so any image from an un-shifted lens will exhibit this. The subject in centre frame un-shifted is closer to the lens than the subject which becomes centre frame after the shift and the field curvature removes the need to refocus between shifts as refocusing would potentially change the subject-image magnification.

Bob
 
You must log in or register to reply here.
Back
Top