Has deep focus ever been the norm?

[cole t parzenn]

Outside of mainstream Western film, perhaps? Razor thin as a stylistic choice seems more recent but I haven't seen a lot of true deep focus in any era.

Edited February 16, 2016 by cole t parzenn

[David Mullen ASC]

Deep focus takes a certain extra amount of effort and a larger lighting package so it would have never become "the norm" but it was most popular with b&w cinematography starting after 1941 with "Citizen Kane" and going through to the decline of b&w by the early 1970's (one of the last major b&w movies of the 1970's like "Paper Moon" and "Young Frankenstein" are fairly deep focus, particularly "Paper Moon", which was all shot at f/16 on a 25mm lens.)

 

However, semi-deep focus was more or less an industry goal for a lot of the 1950's and 60's, not "Citizen Kane" levels of deep focus, but many DP's aimed for an f/5.6 at least, if practical. That was easier to do in b&w than in color, since the color stocks were much slower.

 

What you see in those decades is directors and cinematographers staging shots in depth, whether or not the lens and the stop could hold everything in sharp focus. You see this a lot in "Touch of Evil", which had a lot of location work and night work, so actually lighting to a deep stop wasn't possible, but many shots were done on an 18mm lens to get a pseudo-deep look. Same goes for John Frankenheimer's movies of the 1960's like "Seconds" or "Seven Days in May".

 

And of course, it was easier to get deep focus outdoors in daylight, as in all those westerns.

 

There is also Kurosawa's 1960's b&w movies shot on longer lenses in anamorphic, but often stopped down to f/16 or even f/22 for a deep focus look ("Yojimbo", "Red Beard", "High and Low", etc.)

 

 

So for many directors and DP's of those decades, deep focus was often aimed for, even if not achieved consistently.

[Carl Looper]

Here's an interesting article on the subject.

 

It takes issue with Orson Welles as the origin of deep focus. Or Renoir for that matter (as others have sought to argue). This is not to dismiss Welles - as Welles (amongst others) will contribute to the popularity of deep focus as a technique (very different from tracking influences and origins). In general, the article suggests that deep focus was something that many filmmakers were attempting, for quite a long time, long before Welles arrived.

 

Indeed one might even suggest those working on the manufacture of faster film stocks had deep focus as their guiding idea - and/or that this idea (as a choice) has been in the works from the very beginning (the first photograph is a 9 hour exposure).

 

http://fredrikonfilm.blogspot.com.au/2013/10/the-deep-focus-conundrum.html

Edited February 16, 2016 by Carl Looper

[Carl Looper]

A wider lens provides for the same deeper sense of focus as a smaller iris does. One can use either method. It is not a case of one being pseudo and the other not so.

 

For all f/stops (and all lens magnifications), there is only one and the same plane cutting through a scene where the image is at it's sharpest. Either side of such a plane, the signal becomes softer. One can minimise this softness (but not elliminate it), using either a wider lens or a smaller iris. Both provide for the same result. An iris just provides the means to go further than that which a change in lens magnification might achieve.

 

 

C

Edited February 16, 2016 by Carl Looper

[Carl Looper]

In the following images there is a difference in the amount of depth of field.

 

As we zoom in closer, the less depth of field there is. By this is meant that in the top picture we might say the depth of field includes most of the padlock. But in the bottom picture we'd say the depth of field is a lot smaller than this.

 

But each of these images are from the exact same source image. What differs between them here is only the magnification. As we increase the magnification we decrease the depth of field. Because in reality there is no real depth of field - there is only one distance at which the signal is actually in focus. Either side of that is some arbitrary definition of an acceptable amount of softness before we declare the signal too soft, ie. deem it out of focus. As we magnify the image we magnify that which we might have otherwise included within acceptable focus - and it it goes out of focus (out of the range of what we accept as in focus).

 

By decreasing the magnification (using a wider lens) some of that which we found was out of focus in the telephoto shot, goes back into focus, ie. back into that depth range we accept as the depth of field.

 

 

Edited February 17, 2016 by Carl Looper

[cole t parzenn]

But cropping isn't the same as using a different focal length; the dof is "baked in" and you're not capturing the subject with more film/sensor area for greater detail on the plane of focus.

[David Mullen ASC]

Actually cropping the image is similar to cropping by using a lens with a narrower field of view, the only thing that makes them different is that you'd use a different circle of confusion figure to calculate depth of field when you crop and enlarge the image in post.

 

Filming closer with wider angle lenses doesn't actually increase depth of field since you end up focusing closer but since the background recedes in size faster compared to a longer focal length it is harder to tell that it is not in sharp focus.

[cole t parzenn]

I think that that's more or less what I meant, that subject and format size need to be considered?

[Carl Looper]

Actually cropping the image is similar to cropping by using a lens with a narrower field of view, the only thing that makes them different is that you'd use a different circle of confusion figure to calculate depth of field when you crop and enlarge the image in post.

 

Filming closer with wider angle lenses doesn't actually increase depth of field since you end up focusing closer but since the background recedes in size faster compared to a longer focal length it is harder to tell that it is not in sharp focus.

 

 

If the size of the background shrinks with respect to a subject that is kept the same size in relation to the frame (the vertigo effect) the shrinking size of the background is exactly the same as decreasing the magnification on the background while otherwise maintaining the same magnification on the subject. In other words, in terms of depth of field, the background literally becomes more "in focus" (as per previous discussion)

 

Basically, if it looks sharper, it actually is sharper.

 

Depth of field is defined in terms of how spread out (into a circle), a point in the scene would be at the film plane. The larger the circle the softer we say it is. The smaller the circle the sharper we say it is. As we decrease the magnification on an otherwise soft point (ie. "zoom out"), we are literally decreasing the size of the circle it is otherwise occupying at the film plane. We are literally making it sharper at the film plane.

 

So we are literally increasing the depth of field.

 

A distant background may not enter that zone one considers acceptable focus, but insofar as it gets sharper, it means any background closer to the subject also gets sharper, and somewhere in space (between distant background and in focus subject), will be a location, that will move across that boundary between that deemed out of focus and that which we otherwise accept as in focus.

 

C

Edited February 17, 2016 by Carl Looper

[Carl Looper]

One doesn't have to take my word for it.

 

Careful consideration of any depth of field table will confirm what I'm otherwise saying.

 

C

[David Mullen ASC]

No, not really. Depth of field is about the range in each direction from the point of focus that would be considered acceptably sharp -- it's not technically about how the far background looks.

 

I don't have my charts here but if you shoot an object with a 50mm lens focused at 6 feet and switch to a 25mm lens focused at 3 feet (having moved the camera closer by half in order to maintain the same subject size) you should find that the DOF range would be similar, that you haven't gained more depth of field. However, the background would have become smaller looking on the 25mm lens and thus the 25mm shot at 3' might feel deeper in focus than the 50mm shot at 6'.

[Carl Looper]

No, not really. Depth of field is about the range in each direction from the point of focus that would be considered acceptably sharp -- it's not technically about how the far background looks.

 

I don't have my charts here but if you shoot an object with a 50mm lens focused at 6 feet and switch to a 25mm lens focused at 3 feet (having moved the camera closer by half in order to maintain the same subject size) you should find that the DOF range would be similar, that you haven't gained more depth of field. However, the background would have become smaller looking on the 25mm lens and thus the 25mm shot at 3' might feel deeper in focus than the 50mm shot at 6'.

 

I'd have a closer look at some depth of field charts - nobody uses them anymore - but they will confirm what I'm otherwise saying.

 

Depth of field is certainly the range in which something is in acceptable focus (no argument with that), but the way in which "acceptable focus" is defined (the point spread size, or circle of confusion) means the range of acceptable focus varies as a function of lens focal length, and one will find that the range (of acceptable focus) does indeed increase as one decreases the focal length. The greater the change in focal length, the greater will be the change in range (of acceptable focus).

 

Making the background smaller is the same as making it's circles of confusion (or point spread functions) smaller. The background literally becomes sharper, and so too must anything closer to the focus plane. The depth of field has no choice but to get larger.

 

C

Edited February 17, 2016 by Carl Looper

[David Mullen ASC]

I found my old AC Manual chart.

 

50mm at f/2.8 focused at 6' has a DOF from 5' 9" to 6' 4" -- a 7" range

 

25mm af f/2.8 focused at 3' has a DOF of 2' 9" to 3' 4" -- a 7" range.

 

So there is no increase in DOF by switching to a wider-angle lens if you maintain the same subject size by moving closer, it's just that the background looks smaller in size on a 25mm instead of a 50mm, which makes it harder to tell that it is soft so the shot "feels" deeper in focus. But the actual depth of field hasn't changed really.

 

Sure, if you didn't cut the distance in half, the 25mm shot will have more depth of field than a 50mm shot, but then it's a wider shot of the subject unless you cropped the 25mm in half, either in post or by putting the lens on a camera with half the size film format or sensor in it (ignoring that you'd technically have to pick a new CoC figure if you enlarged that cropped shot back up to the viewing size of the uncropped version.)

[Carl Looper]

Yes, sometimes the change in depth of field is too small to see.

 

But here are some values from a DOF table - with distance measurements in meters, where I've sought settings in which the effect would be more visible. Note the entries for 50mm at 5 meters and 25mm at 2.5 meters. These are not unreasonable settings one might have in a shot.

 

f/8

Dist Near Far Range
50mm 5 3.84 7.15 3.31
30mm 3 2.00 6.02 4.02
25mm 2.5 1.56 6.28 4.72
20mm 2 1.14 8.08 6.94

10mm 1 0.40 ∞
5mm 0.5 0.12 ∞

 

The table is for a Canon 7D, using this online calculator:

 

http://www.dofmaster.com/doftable.html

 

The circle of confusion is defined as 0.019 mm in all cases (same camera sensor size).

Edited February 17, 2016 by Carl Looper