Looking at DoF.

woof woof

woof woof

I like a nice Chianti
Messages
43,551
Name
Alan
Edit My Images
No
After yesterdays discussion about DoF I wanted to take similar shots with an APS-C DSLR, a MFT Panasonic GF1 and a LX2 to try and see if sensor size affects DoF. Unfortunately the weather is pretty poor today so my photo day out is postponed and instead of shooting something outside I gathered things up and shot them on the bedside cabinet. Actually this is probably better even if the subject is more boring.

This is just for fun and argument and also to maybe learn a bit. I'm pretty convinced that zoom length doesn't affect DoF and I've done my own little tests that seem to prove this. I'm also convinced that bigger prints make a lack of DoF more obvious and that smaller prints can appear to have more DoF just because they're small, once blown up they'll have the same DoF as the larger print. That's my theory. So off I went to look at the effect of sensor size...

The shots were taken with a 20D + 50mm set to f2.8, GF1 + 20mm set to f2.8 and the LX2 set to f2.8. Moving forward or backwards a few inches to match the framing made next to no difference to the DoF and you'll have to trust me on this otherwise I'd have to post twice as many pictures.

So...the first shot is the whole scene so you know what you're looking at and the following three shots are 100% crops taken with the LX2, GF1 and then the 20D.

The scene.
c821597c.jpg


LX2, f2.8.
1-1-7c.jpg


GF1, f2.8.
2-1-7c.jpg


20D, f2.8.
3-1-7c.jpg


To me this shows that the LX2 has the most DoF and that the 20D has a teeny tiny bit less DoF than the GF1 but really there's little difference between the two but there's a much bigger difference between them and the LX2.

There is a slight difference in the size of the things within the frames but making them all exactly the same size wouldn't change things a great deal and I'm sure that it would still be clear that the LX2 has the most DoF.

For this last shot I had to move the LX2 much closer. I haven't bothered doing a 100% crop here as you can easily see that the background is now out of focus, so it can actually be done.

4.jpg


Can anyone explain why sensor size should affect DoF and maybe put some numbers to it?
 
We have been over it in the other thread! Larger sensors give less DOF, because with any given focal length, you have to move closer to the subject, to give it the same framing as you would with a smaller sensor. The closer you move to a subject, the less DOF you inherently have. That's all there is to it. Subject to camera distance. The other factors are just that - factors.

The LX2 is a compact I believe, hence it delivering the most DOF. The APS-C DSLR will only have a little less DOF than the GF1 as its sensor isn't that much bigger than the GF1.
 
Last edited:
Well if you don't wanna be part of the thread...leave.

I posted this because I said I would yesterday...As I said, moving forward or backward a few inches made b****r all differece. Not enough to account for the difference.
 
Thats a really confusing post, I read it through a few times lol.

I am pretty that focal length does affect DOF as to does distance from the lens to the subect and again the distance of the background and of course the aperture. I was also of the belief that a larger sensor in acamera allowed you to control the DOF more finely
 
Last edited:
Not sure what you're trying to prove here.

If you frame things the same, the bigger sensor will have less DoF as you just end up using a (35mm equivalent) shorter focal length to get the same field of view and the shorter focal length will have the bigger DoF for a given focal distance.
 
woof woof said:
Well if you don't wanna be part of the thread...leave.

I posted this because I said I would yesterday...As I said, moving forward or backward a few inches made b****r all differece. Not enough to account for the difference.
Click to expand...

My point was we've been over this in the previous thread, and numerous people including myself told you the reasons for exactly what you are seeing here. HoppyUK and Musicman even provided you with facts and figures to bear out these assumptions, and a link to DOFmaster.com.

I am not going to spend the time and effort reposting information which was just ignored in the first place; instead I'll provide a link:

http://www.talkphotography.co.uk/forums/showthread.php?t=265346

Take particular note of posts 18, 22, 23 and 29. All of the answers to the questions you are asking here, are in that thread.
 
Last edited:
arad, I'm not trying to prove anything, I thought I'd been clear about that. What I'm trying to do is understand why compacts have or appear to have more DoF. What I'm convinced of regarding focal length I've said.

Donki, It is a bit confusing but concentrating on the constants helps. I hope. They're all at f2 and at similar zoom lengths and obviously the subjects didn't move. Swaying backwards or forwards a few inches to get the framing as exact as possible didn't make a lot of difference, hardly any, so I didn't post those pictures.

I might have another go at this another day just for my own benefit as I don't truly understand what's going on with compacts and their larger DoF despite trenches certainty.
 
trench, you often come across as being aggressive. I don't know if you mean to be but that's just how I interpret your posts and as I wouldn't want you to waste another single second on me please hit the ignore button, I hope there is one.
 
"and the shorter focal length will have the bigger DoF for a given focal distance."

I don't think so. I'm pretty convinced that zoom length doesn't affect DoF.
 
woof woof said:
I don't think so. I'm pretty convinced that zoom length doesn't affect DoF.
Click to expand...

This isn't what we're disputing. Subject to camera distance affects DOF. That is it. The other factors that come into play as a result of that are exactly that - other factors. But having a bigger sensor, or a longer zoom in itself is not what is influencing the DOF.
 
But moving a few inches (and that's all it was) forward or backward to get the framing pretty much spot on didn't change things that significantly not enough to account for the much greater LX2 DoF. That's what I'm trying to understand.

You say that the framing is everything and accounts for the difference in DoF but so far I haven't been able to prove that to myself. I'll have another go another day. Maybe the f numbers aren't being reported correctly, I don't know, but this test didn't answer all my questions.
 
I just read the other thread and it is pretty clear and what trench says is correct, its the distance to the subject that I thougth it was which effects the DOF. The crop factor influences focal lenth and therefore you must change the distance you are standing from the subject depending on what one of your cameras you are using to get the same shot. I "think" understand it in my head but I dunno if I explained it very well, am i along the right lines trench or should I grab my coat?
 
"I just read the other thread and it is pretty clear and what trench says is correct, its the distance to the subject that I thougth it was which effects the DOF."

Yes. I accept the theory, but I'm trying to see it in practice and as I keep saying, swaying backwards and forwards a few inches to get the framing as exact as I could didn't make the DoF equal. The LX2 still had more.
 
to calculate DOF you use the formula on this page

If you used a 100mm lens on a FF body and crop body all the parameters would stay the same apart from C (or figures derived from using c) - is the circle of confusion which is worked out by

CoC = (CoC for 35mm format) / (Digital camera lens focal length multiplier)

and focal length multiplier is directly linked to sensor size
 
woof woof said:
I don't think so. I'm pretty convinced that zoom length doesn't affect DoF.
Click to expand...
Oh but it does...

Back to the online calculator here: http://www.dofmaster.com/dofjs.html

Dropping in 10ft to subject, with 5Dm2 at F2 and 15mm F/L gives:

Depth of field
Near limit 5.53 ft
Far limit 52.3 ft
Total 46.8 ft

In front of subject 4.5 ft (10%)
Behind subject 42.3 ft (90%)


Changing to 50mm as the only change gives:

Depth of field
Near limit 9.33 ft
Far limit 10.8 ft
Total 1.45 ft

In front of subject 0.67 ft (46%)
Behind subject 0.78 ft (54%)

With the smaller sensors, you are just using different focal lengths to get the same field of view hence changing the depth of field....

QED ;)
 
Last edited:
woof woof said:
But moving a few inches (and that's all it was) forward or backward to get the framing pretty much spot on didn't change things that significantly not enough to account for the much greater LX2 DoF. That's what I'm trying to understand.
Click to expand...

Of course... that's because the LX2 has so much DOF that any slight movement just isn't noticeable, even close up, you have a LOT of leeway, whereas on larger formats, you don't.
 
Last edited:
These, and my own experiments convinced me that zoom length doesn't affect DoF..

http://www.film-and-video.com/dofmyth.htm

http://www.cybercollege.com/myths.htm

http://www.completedigitalphotograp...5cdContents/Chapter9/Focal Length and DOF.pdf

http://www.luminous-landscape.com/tutorials/dof2.shtml

They could be wrong... :) but my own shots seem to confirm what they say.

I've just shot some more shots with the LX2 and the only way that I can get the DoF to be as shallow as the 20D is to have it really close, so close that the framing and distance are obviously very different from that of the other 2 cameras like in the last image I posted.

I accept that logically the size of the sensor shouldn't matter and that it should be down to the camera to subject distance, the framing, but it doesn't appear to be that simple. Not in this little test.

Sitting in exactly the same place and taking the three pictures shows that the LX2 has the most DoF. Swaying backwards and forwards to get the framing as near as damn it doesn't alter that.

That's my problem.

Obviously I have the time to do this but if anyone else also has the time maybe they could try too.

At the moment I can't say what the issue is but it doesn't seem to be 100% or anything like 100% down to camera to subject distance. There seems to be something else, wether the cameras are reporting slightly wrong f values, I don't know, but there does seem to be something else going on.
 
Last edited:
woof woof. You're missing the point. In all those other ones you've quoted, they are changing the subject distance whilst keeping the sensor the same and zooming to get the same field of view. What you are doing with your experiments is keeping the distance (more or less) the same, changing the sensor and then using the zoom to get the same field of view. These are not interchangeable ;)

To get the same field of view as an APS-C sensor on a compact at the same distance, you will use a much shorter focal length which increases the depth of field at a given distance.

You are looking at this the wrong way around, and not only changing one thing at a time.
 
Here is an excellent example of how depth of field changes with focal length......

In the good old days Nikon used to use different colours for different f stops.... they then etched onto the lens barrel the indicated depth of field relative to focal length for any given aperture .....

http://www.photosynthesis.co.nz/nikon/ai80200n.jpg
 
"Now I am confused." You're not the only one.

One article I read said basically that it doesn't matter what the theory says what matters is that you understand enough to get the result that you want.
 
"Although focal length appears to affect depth of field (the area in focus along the lens axis), this appearance is actually based only on differences in camera-to-subject distance and target image size.

This has been confirmed by tests including the one illustrated in Popular Photography ("Depth of Field 101," July, 1994.)

For a specific lens-to-subject distance and comparable image size on the target, all lenses of comparable optical design will (regardless of focal length) have the same depth of field when used at the same f-stop.

Sure, it's true that a zoom lens used at 10mm appears to have a much greater depth of field than when the same lens used at 100mm. But the 10mm view simply is able to hide the existing lack of sharpness through reduced image size.

This explanation is more than just academic. It explains among other things why a subject which seems perfectly sharp can suddenly go completely out of focus when you zoom in.

By zooming in you end up magnifying the existing out of focus area until it becomes noticeable, and objectionable.

If you have correctly focused on and main subject to start with, you will generally see the background "appear" to go out of focus as you zoom in."

I'm in two minds about all this. There's a part of me that wants to just forget it and move on and there's a part of me that wants to fully understand.
 
woof woof said:
For a specific lens-to-subject distance and comparable image size on the target, all lenses of comparable optical design will (regardless of focal length) have the same depth of field when used at the same f-stop.
Click to expand...
And this is why it doesn't apply to your experiment ;) Because the sensors are different sizes in the tests, the target (i.e. the sensor) is of different sizes. What you are doing is keeping the lens-to-subject distance (near enough) constant but varying the size of the projected image as the size of the projected image needs to be different to fill the different sensor sizes.

This is why people say the sensor size matters. To get the same picture on a smaller sensor at the same subject distance, you need to zoom out (think of it being a smaller and smaller part of your APS-C sized sensor). Zooming out shortens the focal length which increases the field of view. When you get to the same amount of coverage as the compact has on your APS-C sensor, the depths of field will be the same. It's just you have a very small image on the APS-C sized sensor, but it would cover the same area as the full compact sensor.
 
"What you are doing is keeping the lens-to-subject distance (near enough) constant but varying the size of the projected image as the size of the projected image needs to be different to fill..."

But you see I'm not... I don't think so anyway.

I tried both shooting each camera from the one position and I also tried altering my position to make the framing and the size of the things in the frame look the same, or near as damn it the same, so that the final image looked pretty much the same. But what's phasing me is that even changing the camera to subject distance didn't make that much difference to the compact DoF unless the compact is veeeeery close, as in the last shot and in that shot the size of the things and the framing is really far out from the other shots.

I haven't convinced myself yet what's goin on and what I believe yet :thinking: :)

I'm happy enough when using just one camera but it's the difference between the compact and the others that I'm currently trying to understand as at the moment nothing I've tried has convinced me that it's the whole answer.

I might give this a go another day and pay even more attention to framing and image sizes but as I said, the only difference I've made today is by getting way too close and getting all sizes and relationships wrong.

It's been an intersting thing. :)
 
woof woof said:
"What you are doing is keeping the lens-to-subject distance (near enough) constant but varying the size of the projected image as the size of the projected image needs to be different to fill..."

But you see I'm not... I don't think so anyway.
Click to expand...
You are ;) You're forgetting that the sensors are different sizes, so you need different amounts of zoom to get the same field of view.

woof woof said:
I tried both shooting each camera from the one position and I also tried altering my position to make the framing and the size of the things in the frame look the same, or near as damn it the same, so that the final image looked pretty much the same.
Click to expand...
But what are the focal lengths in each of these cases. The thing you are missing is the sensor size varies wildly here. A compact camera at 12" will have the same sort of DoF as an APS-C at a much higher distance for the same zoom settings.

Basically, you are varying 2 things (sensor size and distance or zoom depending on your scenario) and trying to equate the results when you can't. All the comparisons you quote only ever change one thing at a time.
 
woof woof said:
But what's phasing me is that even changing the camera to subject distance didn't make that much difference to the compact DoF unless the compact is veeeeery close
Click to expand...

There's a very simple explanation for this.

Close to the hyperfocal distance, the crop sensor has much greater depth of field than the larger sensor.

Since the hyperfocal distance for a very small sensor is, in itself, quite close to the camera, that's the effect you're seeing.

http://www.bobatkins.com/photography/technical/digitaldof.html
 
To all, yes I know, but...

I've tried keeping the cameras at the same distance, I've tried getting close with the LX2, I've tried altering position to get the objects in the frame the same size so that the final images appear to be the same and I had thought that this would make a difference and get the LX2 DoF down but the only thing that demonstrates any real difference is getting way too close with the LX2 and making both the distance between the camera and the objects and the size of the objects in the frame extreme.

Granted I haven't got everything mm perfect but I'd have thought that I'd got it all accurate enough to show a difference and get the DoF to be somewhere near and yet what I have tried hasn't made any real difference until going to extremes.

Theory is one thing but what I wanted to do was try in practice and so far all I've done is prove what I already knew, that you have to get close with a compact to throw the background out, and we all know that. Day wasted? No, I've kept my mind active and I couldn't go out anyway.

Thanks to all contributors. I'll have a think and I might have a go another day but I don't know if comparing these cameras in practice (rather than theory) is even possible. If anyone else has two or three camera formats and wants to do a test, good luck. :o)
 
Last edited:
Musicman said:
There's a very simple explanation for this.

Close to the hyperfocal distance, the crop sensor has much greater depth of field than the larger sensor.

Since the hyperfocal distance for a very small sensor is, in itself, quite close to the camera, that's the effect you're seeing.

http://www.bobatkins.com/photography/technical/digitaldof.html
Click to expand...

BTW, this is why the makers of cheap 110 format film cameras in the 70s and 80s got away without including any focusing mechanism at all. Just fix the focal point the short, hyperfocal distance in front of the camera and everything is in focus.
 
It is better to ignore focal length and subject distance, and just consider magnification. If you use a longer lens, but move back so that the subject framing is the same, depth of field remains the same, because the magnification hasn't changed (other things will have though, like perspective and field of view, but not DoF).

When magnification is reduced (ie the image is smaller on the sensor) depth of field is increased, and vice versa. That's just how the physics works.

So, it follows that if you have a smaller sensor in the camera to start with, everything will be correspondingly reduced in size. Compact camera, tiny sensor, loads of DoF. Crop format and 4/3rds DSLRs, medium size sensor, medium amount of DoF. Full frame, bigger sensor, less DoF.

If you want to take even greater extremes, compare a camera phone to a large format film camera, though I don't know why you would want to. The physics is in the formulae that DofMaster uses, and unsurprisingly, they work.
 
chuckles said:
Here is an excellent example of how depth of field changes with focal length......

In the good old days Nikon used to use different colours for different f stops.... they then etched onto the lens barrel the indicated depth of field relative to focal length for any given aperture .....

http://www.photosynthesis.co.nz/nikon/ai80200n.jpg
Click to expand...

The heart of all D O F`s is the hyperfocal. Everything stems from this. Some of the nutters :) you get on these forums start talking about walking back etc. The basic formula for H is "FL^2/( a * C of C) /1000" Tried to send a few programs thru email but now you get " Cannot send, contains Executable
files".
 
classcams said:
The heart of all D O F`s is the hyperfocal. Everything stems from this. Some of the nutters :) you get on these forums start talking about walking back etc. The basic formula for H is "FL^2/( a * C of C) /1000" Tried to send a few programs thru email but now you get " Cannot send, contains Executable
files".
Click to expand...

Have to disagree slighty with this..... critical also is Circle of Confusion. It is almost arbitary but CoC is based upon what is acceptable and 'print viewing distance'. The further the viewing distance the more acceptable the 'confused' image.

Try zipping the file and then encrypt it with a password. Most mail programs now look at any attachment containing an 'executable' binary as being a potential, viral threat. Encrypting it usually stops scanners looking at the content.
 
chuckles said:
Have to disagree slighty with this..... critical also is Circle of Confusion. It is almost arbitary but CoC is based upon what is acceptable and 'print viewing distance'. The further the viewing distance the more acceptable the 'confused' image.

Try zipping the file and then encrypt it with a password. Most mail programs now look at any attachment containing an 'executable' binary as being a potential, viral threat. Encrypting it usually stops scanners looking at the content.
Click to expand...

Of course, all the variables are there. Once you have decided what you want ,A , FL and the distance you want to view the shot at,it`s all there.
 
I think that it's actually a good idea to ignore most of the theory on this and just go with what works :lol:

An update...

I thought that I knew all this or at least enough to get by before I did this little practical test so I had another go and this time as I'd started to confuse myself I started by confirmed again something that I thought that I already knew. I tried a 20mm and then a 50mm lens on the 20D and sure enough when I got the framing near enough (again not measuring anything to the nearest mm) the DoF was the same. That confirms what I've read, what I thought and how I tend to shoot and it makes sense to me to believe that focal length doesn't affect DoF, framing being the same.

So far so good with the 20D, I didn't bother with the GF1, I went straight to the LX2.

I again tried to frame the shot, trying to get the alarm clock the same size in the frame etc as I had with the 20D and again I couldn't get anywhere near the DoF of the 20D. The only way to match it is to get really close and increase the size of the clock in the frame way beyond the framing with the 20D.

This is the first time that I've tried a practical comparison between different sensor sizes and it has surprised me. The surprise being that matching the size of the foreground object within the frame with two different sized sensors (APS-C and Compact) and looking at the DoF shows that they're different. That might not surprise some, but I had thought that it would have been a much closer thing.

So ignoring all theory and formulas and going with what seems to work, for a given framing of a subject I have two ways of affecting DoF, aperture and sensor size. Like I said, maybe no surprise to some but for me it's been a worthwhile practical test.
 
woof woof said:
I think that it's actually a good idea to ignore most of the theory on this and just go with what works :lol:

An update...

I thought that I knew all this or at least enough to get by before I did this little practical test so I had another go and this time as I'd started to confuse myself I started by confirmed again something that I thought that I already knew. I tried a 20mm and then a 50mm lens on the 20D and sure enough when I got the framing near enough (again not measuring anything to the nearest mm) the DoF was the same. That confirms what I've read, what I thought and how I tend to shoot and it makes sense to me to believe that focal length doesn't affect DoF, framing being the same.

So far so good with the 20D, I didn't bother with the GF1, I went straight to the LX2.

I again tried to frame the shot, trying to get the alarm clock the same size in the frame etc as I had with the 20D and again I couldn't get anywhere near the DoF of the 20D. The only way to match it is to get really close and increase the size of the clock in the frame way beyond the framing with the 20D.

This is the first time that I've tried a practical comparison between different sensor sizes and it has surprised me. The surprise being that matching the size of the foreground object within the frame with two different sized sensors (APS-C and Compact) and looking at the DoF shows that they're different. That might not surprise some, but I had thought that it would have been a much closer thing.

So ignoring all theory and formulas and going with what seems to work, for a given framing of a subject I have two ways of affecting DoF, aperture and sensor size. Like I said, maybe no surprise to some but for me it's been a worthwhile practical test.
Click to expand...

Cropping is another way of effectively using a smaller format, and therefore increasing DoF. I've only ever used it for macro, when DoF is so shallow and you sometimes need all the help you can get. When you've run out of all other options, if you just move back so that the subject only fills half of the frame (by area, not linear) then you gain one stop of DoF.

It's a compromise though, like all these things, and while you certainly increase DoF, you lose out on a bit of sharpness generally through greater enlargement, increased noise etc.
 
Main thing here for me is that it's only really applicable to people who use a variety of cameras during a shoot and need this knowledge (about DoF Vs Sensor size) to form an idea of what result they'll get with a certain body/lens combo.

I use two bodies that are the same so A) they share lenses, and B) I know everything is the same across both bodies in terms of final image. Swapping between a 1x5 crop D2x and (for example) a 1.3x crop EOS 1D would be pretty mind boggling.

Fair play for the experiement though :)
 
"Main thing here for me is that it's only really applicable to people who use a variety of cameras during a shoot and need this knowledge (about DoF Vs Sensor size) to form an idea of what result they'll get with a certain body/lens combo."

Oh yeah, as I'm just an amateur this is of no real world use to me at all :lol: and in reality I can't change my sensor so the only real way to change DoF for a fixed composition is aperture.

It's just something that's been on my mind as back in the days when I had a Nikon 35mm and one lens I didn't have different formats / sensor / film sizes to play with and now I have three (well...possibly 4 if I count my phone...,) a compact LX2, a MFT GF1 and an APS-C 20D and as I said, this is the first time that I've tried to take the same shot with each.
 
woof woof said:
So ignoring all theory and formulas and going with what seems to work, for a given framing of a subject I have two ways of affecting DoF, aperture and sensor size. Like I said, maybe no surprise to some but for me it's been a worthwhile practical test.
Click to expand...
Not a surprise here ;). You don't need to ignore theory or formulas - they tell you exactly what you have found out - with the same framing, the larger the sensor size, the shallower the depth of field, all else being equal. All this comes from the fact you have to use a shorter focal length to get the same framing on smaller sensors. It's not actually the sensor size that gives you the DoF, but it is the combination of aperture and focal length used to get the shot.
 
Combination of FL and A only for DOF. Oh how true that is. Naff all to do with the sensor. Altogether now "I`m walking backwards for Christmas"
 
"Combination of FL and A only for DOF. Oh how true that is. Naff all to do with the sensor. Altogether now "I`m walking backwards for Christmas""

I'm not too sure if you're being funny or serious in all of that but actually after reading up I think that it is also a factor of sensor size and possibly of pixel size / density but I don't know how that would relate to different sensors of the same size. Maybe the densities aren't different enough yet. At least I've read some pretty convincing stuff that that's the case.

"It's not actually the sensor size that gives you the DoF, but it is the combination of aperture and focal length used to get the shot."

That's what I've not been able to convince myself of. I don't think focal length matters and some web sites seem to be saying that it doesn't. Some seem to be saying that the sensor size and even the size of the pixels and density matters.

The technical aspects of all this do interest me but there are so many web sites saying contradictory things that as I simply don't myself know it's impossible for me to know from reading stuff on the net. Lots of sites says the they're right and the other view is wrong... so I don't think that there's a definite answer yet or if there is it's hidden amongst all the rest and not acknowledged as the truth, the whole truth and nothing but the truth.

Looking at just one format, my 20D, focal length doesn't affect DoF if the framing is the same. I seem to believe that and I seem to be able to prove that to myself in my own little tests. If I change formats and use my LX2 and get the framing right so it looks like the 20D shot the DoF is different. That seems to be pretty conclusive to me so I'm now perfectly willing to believe that two things affect DoF besides the size and relationship of the things in the frame. The sensor (size, pixel size and density maybe if certain web sites are right) and aperture.

I said earlier that none of this makes much practical difference to me as I'm just an amateur but for some specialists I'm sure that this stuff matters and as compacts get better and better I suppose that the implications and possibilities will be there if you need them.
 
Last edited:
You must log in or register to reply here.
Back
Top