VERY technical question about depth of field (you've been warned)

[Alex_B]

and because he's just a cool cat from what I know of him

Thanks, but stray dog describes me better I would guess

 

[Sark]

But "appear in focus" always relates to print size and viewing distance.

I agree. I guess the discussion depends on how many variable you bring into the equation. My opinion is based on the image at capture. After this there are various factors that determine image sharpness. These other variables are trickier to quantify. A full frame sensor can produce near flawless detail up to, say A4. But so can an APS sized sensor. Of course it is also true that once print size effects detail, it will effect it to a greater extent with images from the smaller sensor. And I guess this is what you are saying.

Now I think I'll go look for that Nikon v's Canon thread.

Sark

 

[jamesino]

“If the subject occupies the same fraction of the viewfinder (constant magnification) for both a wide angle and a telephoto lens, the total depth of field is virtually* constant with focal length”.

But in order for a shorter lens to have the same subject magnification as a longer lens, the shorter lens has to be closer to the subject. Therefore, the focal distance is decreased and since a decreased focal distance decreases the DOF, it's not really a far comparison between a shorter and longer lens.

 

[Helen B]

Depth of field is such a slippery thing.

Here are some simplified relationships (they are not totally accurate in all situations):

Depth of field is

1 - proportional to the square of the subject distance

2 - inversely proportional to the square of the focal length

3 - proportional to the f-number

4 - proportional to the diameter of the acceptable circle of confusion

It's easiest to make comparisons by varying two things at once. For ease of calculation consider a 4/3 sensor and a full frame sensor, and hold 1 and 3 constant (ie f-number and subject distance). The 4/3 sensor has a 'crop factor' of 2, so to get the same size print the image needs to be enlarged twice as much as an image from the full frame sensor. This means that the circle of confusion must be half the diameter (to end up with the same size blur circle in the print).

Therefore the DoF would be halved by switching from a full frame sensor to a 4/3 sensor. To keep the same angle of view from the same camera position, the focal length of the lens needs to be halved for the 4/3 sensor. As the DoF is inversely proportional to the focal length, halving the focal length quadruples the DoF. Therefore the DoF is halved, then quadrupled - which means that the overall effect is to double it by switching to the smaller sensor and keeping the same camera position and field of view.

(It might be worth noting that the diffraction-limited depth of field is unchanged by sensor size - the larger the sensor the more you can stop down for the same amount of blur caused by diffraction if you hold the print size constant.)

How does that sound?

Best,
Helen


PS I'm on holiday

 

[Sark]

Alex...It seems we posted at the same time yesterday, so I did not read all your replies.

Originally Posted by Sark
Photographers don’t, or rarely at least, change focal length to retain the same FOV.

This is exactly what you do. This is why with your medium format camera you have a totally different set of lenses than with your 35mm small format.

So in order to get the same composition on two different formats, you will need two different focal lengths respectively so you can keep your FOV identical.

True in that context, but that’s not what I was referring to. A photographer using, say a Nikon D3, doesn’t change the focal length with the intention of then changing the focal distance to compensate for the change in FOV. In a real life photography situation the point of using a different focal length is to change the FOV.
In the context of a change in focal length compensating for sensor size, or format size, I agree that it is relevant. Maybe that’s where we are at crossed purposes. I’m not basing my statements on more than one variable. The original post stated...

It says that a digital camera with a 1.5 or 1.6 sensor will have less of a depth of field at any given aperture and focal length than the film camera (i.e. 50mm at f/1.8). I told someone else here this, and they disagreed with me, and said the same lens should have the same depth of field no matter what the camera is, cause they have the same lens, so there shouldn't be any possible way for it to have a different depth of field.

If by same lens, and any given focal length, it meant equivalent lens/focal length, then yes, sensor size effects DOF because sensor size is affecting a change in the focal length. As I read it, the original post did not infer this, but assumed sensor size alone could affect DOF.
DOF on the image circle doesn’t change with a change in sensor size. Not unless you change the focal length to compensate for the smaller sensor. And this was not suggested.

Of course print size, among other factors, has a perceived effect, but that, as I have already stated, is a difficult one to accurately quantify. That is why I have tried to avoid the relevance of print size etc, in this discussion. However, I now accept it is an error on my part to separate optical DOF and image resolution determined by other factors. Reading more technical info on DOF, it is clear that these two factors are not generally discussed separately.

Personally, I would still argue that for the sake of clarity, DOF as a result of focal length and aperture setting, and image resolution from other factors, are separate issues. In a practical sense at least.

Consider the following situation. Two identical cameras with the same sensor size are focused on the same subject. Camera one has a high quality 50mm optic set to f 2. Camera two has an average quality 50mm optic set to f 16. Camera one would produce the better image resolution, but I believe most photographers viewing both resulting images displayed at the same size would consider camera two’s image as having the greatest depth of field. This may not be scientifically correct, but in my experience, most photographers relate DOF to the effect of the aperture setting.
Even at identical aperture settings, focal length effects DOF, but not necessarily optical resolution. Therefore the perceived image resolution at the important field of focus can remain high, whilst the background can appear more, or less, out of focus, due solely to the focal length affecting DOF.
This, at least, is how I have always percieved it.

Sark

 

[Helen B]

DOF on the image circle doesn’t change with a change in sensor size. Not unless you change the focal length to compensate for the smaller sensor.

DoF also changes if you wish to end up with the same print size from different sensor sizes, if everything else is held constant. DoF is not solely a property of the lens, it is also a property of the degree of magnification of the image.

Best,
Helen

 

[Dao]

This is a very interesting subject.

From what I read, you mean the DoF is related to the circle of confusion. The smaller the Circle of confusion, the sharper the image (in focus). Since the smaller sensor increase the size of the circle of confusion (at the finial print). So that is why it affect DoF (assuming other are constant)

Is that right?

 

[Senor Hound]

This is a very interesting subject.

From what I read, you mean the DoF is related to the circle of confusion. The smaller the Circle of confusion, the sharper the image (in focus). Since the smaller sensor increase the size of the circle of confusion (at the finial print). So that is why it affect DoF (assuming other are constant)

Is that right?

Perhaps we should call this thread, the thread of confusion...

 

[Ben-71]

The original question was about DOF, as influenced by the lens' focal length and the sensor
or film format.​

I see that DOF is regarded, by some here, as the perceived DOF in the
end product – the enlargement. ​

When the DOF button in my camera is pressed, the aperture closes to the preset value and
the DOF is visible in the viewfinder.
I kept a few 1K Watt lights in my studio just for this purpose – to check DOF on a View camera
screen or a viewfinder (it takes lots of light to check DOF at small apertures in 1:1+ Macro shots).​

So, it was the DOF-chicken that came before the egg-largement... and -- "appear in focus"
certainly does not "always relates to print size and viewing distance".
DOF is very much there, before the image is enlarged. ​

By changing focal length, we change the angle of view, so we change the DOF.
Changing the format of a sensor (or film) is another way of changing the angle of view, hence a
change of DOF.​

What if, instead of comparing two sensors of different sizes, we compare two sensors of
the same size, but with different pixel number? – Would the 'lesser' sensor give less DOF?
No. Only the resolution of the final image would change.​

When the day will come and a yet smaller, more advanced sensor, with
10 times more pixels, will replace the larger sensor that we now have,
would it change the DOF?
It would, because its smaller size will change the angle of view, while its larger pixel number
will just allow larger & sharper enlargements.
What we see in the viewfinder (through the same lens), when we press
the DOF button on the camera, will look exactly the same.​

If magnification enters the DOF definition, then what about the lens' resolution?
– Would a super-sharp lens give more DOF than a bad one of the same focal length & aperture?
It wouldn't.
(Now I see that Sark already pointed to this)​

So, the DOF is there, before other factors such as enlargement size,
viewing distance, contrast, and quite a few others, which influence the
end product, enter the game.​

Enlargement introduces yet more variables.
Even the choice of material (type of paper, canvas, or screen resolution, etc',), influences the
perceived DOF.​

True, by enlarging, part of the basic dots grows beyond 0.02mm or so, but this is just one of
the variables that change the resolution in general – perceived DOF included – in the final
enlargement.
Such factors are additional to, and not part of, the DOF, because enlarging does not change
the proportion between sharp and un-sharp according to distance.
It makes everything equally less sharp. ​

What if a sensor (or, say, a 6x4.5 back on a 4"x5" view camera) crops so that just the sharp part
out of the DOF is in the final image? Did it increase DOF?
Wasn't there DOF to begin with, but we introduced another element, which is not relevant to DOF ? ​

The perceived DOF in the final image is created by many variables that have nothing to do with
the DOF that was created by the lens & the recorded area of the image circle.​

It seems to me that this debate is mainly about semantics.
There's DOF, and there's perceived DOF in the final enlargement which is influenced by both
the DOF and by other variables which are not part of the DOF.​

Everyone who disagrees with this is kindly requested to return my 2 cents ​

 

[Alex_B]

OK, i have to comment on this: There is definitely nothing like absolute DOF. DOF is always related to what we perceive with our eyes. DOF is about perceiving with your eyes! If your eyes had infinite resolution, then DOF would always be perceived as being zero. But lucky as we are, our eyes have only a finite resolution and we do not realise how things in front of the plane the lens focussed on and behind that plane are in fact out of focus. But we just cannot resolve it, so it all seems in focus. This is called DOF then.

If you assume no enlargenment, then you would just look at the unenlarged film, and DOF will be larger than in the enlargement. Actually, on unenlarged 35mm film almost every shot is both in focus and has infinite DOF if you look at it with your bare eyes.

If you look through your viewfinder with your DOF preview button, the percieved image you see is much larger than when you look at the original film or sensor size.

 

[Alex_B]

the DOF that was created by the lens & the recorded area of the image circle.​

This is something which in principle is not and cannot be defined in a sensible way.

If you assume no enlargement, then you have an enlargement factor of 1 which is just a special case of enlargement. This is the case for contact prints. But unless you are into stamps and other miniatures, this makes no sense for small format film.

 

[Dao]

Quoted from Wikipedia

"The image format size also will affect the depth of field. The larger the format size, the longer a lens will need to be to capture the same framing as a smaller format. In motion pictures, for example, a frame with a 12 degree horizontal field of view will require a 50 mm lens on 16 mm film, a 100 mm lens on 35 mm film, and a 250 mm lens on 65 mm film. Conversely, using the same focal length lens with each of these formats will yield a progressively wider image as the film format gets larger: a 50 mm lens has a horizontal field of view of 12 degrees on 16 mm film, 23.6 degrees on 35 mm film, and 55.6 degrees on 65 mm film. What this all means is that because the larger formats require longer lenses than the smaller ones, they will accordingly have a smaller depth of field. Therefore, compensations in exposure, framing, or subject distance need to be made in order to make one format look like it was filmed in another format."

 

[Alex_B]

Dao, true, but this is the other effect, the lens related DOF effect.

There is also a (smaller) effect related purely to the format, if you keep the lens at the same length. And this is the one which causes dispute here

 

[Sark]

Just to clarify this whole discussion.

I believe you can differentiate between image resolution and depth of field. Whilst I accept they are connected, I believe you can also discuss them separately. I also believe that many photographers will think like this.

Science agrees there is a perceived level of sharpness that is based on the CoC. At the capture point, controlling the desired sharp subject elements is generally achieved by focusing. Controlling those elements you desire to be out of focus is determined by the f-stop. The f-stop will also affect the depth of perceived sharpness of those elements near the focus point. But it is also possible for the f-stop to have negligible effect on the important in-focus elements, whilst affecting the desired out of focus elements significantly. A single flower in a field of flowers for example. The f-stop can therefore affect DOF without affecting the perceived image resolution at the important point of focus. In this example, enlargement will affect the in-focus element to a far greater extent than the out of focus elements (a very blurred object is a very blurred object even at a higher magnification). Now I would consider this loss of sharpness at the important point of focus (due to print size) as an image resolution issue, seperate to depth of field.

Sark

 

[Ben-71]

Alex_B
<< OK, i have to comment on this: There is definitely nothing like absolute DOF. DOF is always related to what we perceive with our eyes. DOF is about perceiving with your eyes! If your eyes had infinite resolution, then DOF would always be perceived as being zero. But lucky as we are, our eyes have only a finite resolution and we do not realise how things in front of the plane the lens focused on and behind that plane are in fact out of focus. But we just cannot resolve it, so it all seems in focus. This is called DOF then.>>​

Agreed. Except for what is, exactly, the definition of 'DOF'... ​

<< If you assume no enlargenment, then you would just look at the unenlarged film, and DOF will be larger than in the enlargement. Actually, on unenlarged 35mm film almost every shot is both in focus and has infinite DOF if you look at it with your bare eyes. >>​

Not the DOF will be larger on film than in the enlargement, but, typically
the grain will be smaller. But, again, I do think that this is a question of
semantics: Do we include more variables, such as contrast, type of paper,
etc', or we do not.​

BTW, I once had no choice but use 35mm for 1~1.5m B&W prints.
I used a special super-high resolution Kodak film (I think it was the TP1524
or something, a line film that could also be developed [in Rodinal] as full-tone).
The only grain you could see, was the paper's grain.
This, I think, strengthens my conviction that enlargements do not change the DOF,
they can only introduce OTHER variables, which may change the perceived DOF
in the end result.​

<< If you look through your viewfinder with your DOF preview button, the percieved image you see is much larger than when you look at the original film or sensor size. >>​

Why much larger? Nikon's specification for the D300 :
Viewfinder Frame Coverage - Approx. 100% (vertical and horizontal)
Viewfinder Magnification -Approx. 0.94x with 50mm lens at infinity​

And, on the Mamiya 6x7 and the Sinar 4"x5" that I used, the matt glasses
were 1:1 with the film format.​

<< This is something which in principle is not and cannot be defined in a sensible way.>>​

The definition of DOF is, of course, an arbitrary one, so how about this simple, arbitrary definition &#8211;
"A dot up to Xmm on the recording media (film or sensor) is sharp. A dot larger
than Xmm is not sharp" ?
(I didn't invent this "wheel") ​

<< If you assume no enlargement, then you have an enlargement factor of 1 which is just a special case of enlargement. This is the case for contact prints.
But unless you are into stamps and other miniatures, this makes no sense for small format film.>>​

I don't think so. It is a special case of 1:1 enlargement if a 1:1 enlargement
is made. Even a contact print can introduce more variables such as contrast,
or the type of paper used - variables which can change the perceived DOF.​

I was talking about no enlargement, as every enlargement introduces more
variables, that aren't part of what makes the DOF.​

It stil seems to me that this debate is mainly about semantics.​

There's DOF, and there's perceived DOF in the final enlargement.
The perceived DOF in the final enlargement is influenced by both the DOF
and by other variables which are not part of the DOF.​

It's 4 cents by now&#8230; ​