Depth of field issues

[David Mullen ASC]

I got a private question about whether a wide-angle shot cropped to a narrow view has the same depth of field as switching to a longer lens to get the same view as the cropped shot.

 

Of course, the way students are often taught is that wide-angle lenses don't actually have more depth of field because if you cropped them to a telephoto view, you'd see the same depth of field as a telephoto shot. But of course, if you think about it a little deeper, the cropped wide-angle shot should still have more depth of field, just as when you switch from a 35mm camera to a 16mm camera and switch to a lens with half the focal length to match the field of view -- essentially you are cropping a shorter focal length lens image by putting it on a 16mm camera.

 

I quickly shot this test in my kitchen just now. Forgive the fact that the shot sizes don't match exactly, it was a bit of an eyeball cropping job but I made each final photo 600 pixels wide.

 

The first is a 48mm shot at f/8:

 

This is a 24mm shot cropped in half to match the view of the 48mm, also shot at f/8. You see there is more depth of field:

 

Now to match depth of field better, still cropping a 24mm shot by half, I opened up from f/8 to f/4:

[John Brawley]

Of course, the way students are often taught is that wide-angle lenses don't actually have more depth of field because if you cropped them to a telephoto view, you'd see the same depth of field as a telephoto shot. But of course, if you think about it a little deeper, the cropped wide-angle shot should still have more depth of field, just as when you switch from a 35mm camera to a 16mm camera and switch to a lens with half the focal length to match the field of view -- essentially you are cropping a shorter focal length lens image by putting it on a 16mm camera.

 

 

Hi David...

 

Doesn't it depend on the subject distance ? I think in the above example (and to paraphrase), you mentioned that students are taught that the depth of field doesn't change when the image size is the same. But what you have done and illustrated really well is that the you can get a difference in depth *IF* you crop in on the actual image (post lens if you like )

 

And so yes, the depth of field is now different. But the trade off is a fairly hefty image blow up right ?

 

jb

[David Mullen ASC]

Hi David...

 

Doesn't it depend on the subject distance ? I think in the above example (and to paraphrase), you mentioned that students are taught that the depth of field doesn't change when the image size is the same. But what you have done and illustrated really well is that the you can get a difference in depth *IF* you crop in on the actual image (post lens if you like )

 

And so yes, the depth of field is now different. But the trade off is a fairly hefty image blow up right ?

 

jb

 

I wasn't advocating cropping as a way of eliminating the need for longer focal length lenses.

[John Brawley]

I wasn't advocating cropping as a way of eliminating the need for longer focal length lenses.

 

Sure sure. Was just clarifying that's what you had done.

 

jb

[Peter Martin]

David,

 

what you did is to enlarge the 24mm-shot only. This proofs that a 24mm lens has more depth compared to a longer lens.

 

You have to come with the 24mm closer to the glass in order to make the qouted rule working. Try another test photographing the glass in the same scale (48 and 24mm) in relation to the sensor size .

 

The effect should be: identical depth of field but more background visible due to the differnt angle of view of both lenses.

 

Peter Martin

 

Vantage Film - Hawk Lens Design

www.vantagefilm.com

[David Mullen ASC]

You're missing the point. I was asked by someone if you cropped a wide-angle shot to match a telephoto shot, whether the FOV and DOF would be the same. You can match FOV by cropping... but you'd get more DOF unless you adjusted the f-stop to compensate.

 

The question wasn't about matching FOV by moving in closer, etc. Yes, I realize that moving in closer causes the depth of field to drop.

 

If you look on a DOF chart and compare a 20mm focused at 3' to a 40mm focused at 6', you'd see that the depth of field was similar at the same f-stop (though not exactly the same, close enough.)

[Peter Martin]

David

 

I do not know any other rule than same DOF with same scale. Shorter lenses even cropped will give more DOF.

 

Peter Martin

[David Mullen ASC]

Now that we've got your attention, Peter... ;)

 

What exactly is going on with the Hawk anamorphic lenses in terms of offering a 1.33X or 1.34X squeeze to fit 2.39 onto a 1.78 sensor / film format? I hear it's now some sort of option or conversion?

[Peter Martin]

We launched recently Hawk V-Lite Anamorphics. They are prepared to work as regular (0.5x) or 0.77x anamorphics. The lenses can be ordered with two different cylindrical groups which have the same mechanical dimensions. The rental house can swap them in order to deliver anamorphics as requested by the customer. The groups are positioned in front of the spherical lens as in all other Hawks. One can compare that to a camera where the 4-perf movement is replaced by a 3-perf movement.

 

We call the new 0.77x lenses 3PD (3-perf & Digital). They can be used on 3-perf film cameras, S16mm or 16:9 sensors.

 

Peter Martin

[David Mullen ASC]

We launched recently Hawk V-Lite Anamorphics. They are prepared to work as regular (0.5x) or 0.77x anamorphics. The lenses can be ordered with two different cylindrical groups which have the same mechanical dimensions. The rental house can swap them in order to deliver anamorphics as requested by the customer. The groups are positioned in front of the spherical lens as in all other Hawks. One can compare that to a camera where the 4-perf movement is replaced by a 3-perf movement.

 

We call the new 0.77x lenses 3PD (3-perf & Digital). They can be used on 3-perf film cameras, S16mm or 16:9 sensors.

 

Peter Martin

 

Why do you use the term 0.5X to describe the 2X squeeze of standard anamorphic? Seems like a non-industry way of describing the lenses... Traditionally anamorphics have been described as 2X (CinemaScope), 1.5X (Technirama), 1.25X (Ultra Panavision), etc.

 

I'm assuming that .77X is the same thing as saying a 1.33X-ish squeeze??? I'm not sure of the math... When I divide 1.78 by .77, I get 2.31... When I multiply, I get 1.37...

 

What would be the "normal" way of describing a .77X squeeze, if .5X = 2X?

[John Sprung]

What would be the "normal" way of describing a .77X squeeze, if .5X = 2X?

It seems to be inverses, so 1X/(0.77) = 1.3X. I also find it strange to be using the numbers this way. To me, numbers bigger than one mean tall and skinny on the film or chip, while numbers less than one would be short and fat.

 

 

 

 

-- J.S.

[Peter Martin]

When calculating anamorphic lenses we work with such numbers. For marketing reasons we would not use them. Just name them 2x or 1.3x.

 

Peter Martin

[Chris Keth]

David,

 

what you did is to enlarge the 24mm-shot only. This proofs that a 24mm lens has more depth compared to a longer lens.

 

You have to come with the 24mm closer to the glass in order to make the qouted rule working. Try another test photographing the glass in the same scale (48 and 24mm) in relation to the sensor size .

 

The effect should be: identical depth of field but more background visible due to the differnt angle of view of both lenses.

 

Peter Martin

 

Vantage Film - Hawk Lens Design

www.vantagefilm.com

 

You are correct. Both would show the same depth of field. The 24mm one would show more background. The trick would be that, because of a smaller enlargement factor for the background, the 24mm one would appear to have more depth of field.

[Danny Lachman]

Is there any practical purpose to this question?

I can see why you'd be curious in terms of canceling out differences by changing the relative perspective, but is there some advantage to doing this? All I can see is that you loose resolution trying to match with cropping - what might this provide otherwise? I'm actually curious.

[Chris Keth]

Is there any practical purpose to this question?

I can see why you'd be curious in terms of canceling out differences by changing the relative perspective, but is there some advantage to doing this? All I can see is that you loose resolution trying to match with cropping - what might this provide otherwise? I'm actually curious.

 

You bet it serves a purpose! Every couple of weeks or so we get people asking about using 35mm format lenses on S16mm and whether a 25mm lens is now a 50mm lens when you put it on an SR3, et cetera.

 

This is a nice visual demonstration (thanks, David. We should start linking new members here first. :lol: ) on why smaller formats tend to exhibit greater depth of field than formats with a larger film or sensor area. In reality it's a simple question with a simple answer but, for some reason, it keeps coming up over and over (and over).

[David Mullen ASC]

Is there any practical purpose to this question?

 

It's just a question that I get now & then. It's not so much about the practical uses of cropping, but about whether a telephoto image is exactly the same as a wide-angle one that is cropped.

[Sam Wells]

This is a nice visual demonstration (thanks, David. We should start linking new members here first. :lol: ) on why smaller formats tend to exhibit greater depth of field than formats with a larger film or sensor area. In reality it's a simple question with a simple answer but, for some reason, it keeps coming up over and over (and over).

 

Hear Hear !

 

I've refered more than one person to Ansel Adams' "The Camera" not knowing what else to do, but this thread really could be the best introduction to the issues..... (not that Adams isn't well worth the read)

 

-Sam

[Bill DiPietra]

What do you guys mean by "cropping?" Are you referring to the "framing" of the shot??

[Hal Smith]

Great thread. Very clear thinking on the subject by all concerned (David first of all - as usual).

[David Mullen ASC]

What do you guys mean by "cropping?" Are you referring to the "framing" of the shot??

 

No, cropping as in cropping in this case, though when talking about switching to a smaller film format, then you could say it is framing.

 

Look, it's simple -- we often tell students that wide-angle lenses often seem to have more depth of field because the background is smaller and farther away-looking, but if you cropped and enlarged the wie-angle shot to create a telephoto image, you'd see just how truly out of focus the background is (which is a different phenomenon than depth of field, which describes just the range around the point of focus that is acceptably sharp).

 

But of course, you actually do get more depth of field by cropping to a narrower view compared to a true telephoto-lensed shot. This is why using a smaller film format gives you generally more depth of field than a larger one -- you essentially are using a smaller area of the lens image and thus need to use shorter focal lengths to compensate to maintain the same field of view, but getting an increase in depth of field unless you compensate by changing the f-stop.

 

This would all make sense if you saw the private question I got that caused me to take the photos to explain the answer to him. I only posted them to show you guys. Here was the original question, though it was not about depth of field, only perspective:

 

Would the hvx at 55mm pointed at an object very near to the lens create the same image (not taking DOF into account) as if you would shoot the same object with the same distance to the lens with the lens at 4.2mm and crop the image so you get the same framing as you got when you shot the object with the 55mm lens? In other words, could you photograph a landscape (also with objects in the foreground) with a wide angle lens and later crop the picture in photoshop and it would look exactly like you used a telephoto lens (except for the DOF and resolution loss)?

This is one of the questions I could never actually resolve. I never was able to come to a conclusion when I played with my HVX.

 

You can crop a wide-angle shot to match the view of a telephoto shot. There may be some minor lens characteristic issues that come into play, but in theory, it could end up the same image in terms of view and perspective.