Myth busted: larger sensors have thinner DOF.

[David Rasberry]

For a given focal distance and aperture, DOF is an optical property of lens focal length regardless of sensor size.
Shorter focal lengths needed to maintain equal FOV on a smaller format have greater DOF at the same distance and aperture. One can compensate with aperture over some range of differences to match DOF on lenses of different focal lengths, but obviously there is a practical limit as to how far one can go with this.
The difference in DOF vs FOV characteristics between S35 and 2/3"/16mm formats is reasonably bridgeable with appropriate optics and techniques.
It is questionable whether the longer lenses required for larger than S35 formats have adequate DOF for practical motion picture work, especially in low light conditions. The ultra shallow DOF look is a fad in my opinion, more effective as a still composition technique than as an effective narrative complement in movies. Indeed there seems to be something of a movement among still photographers towards ultra deep DOF with the new generation optics and digital cameras, a very non-film, pure digital aesthetic. I find it refreshing.

 

[morgan_moore]

I think isolation of the subject helps the viewer concentrate on what they are 'supposed' to be looking at and therefore helps the viewer understant the image quickly and simply - no fad - part of the toolkit of communicating through images

serious out of focus is a fad or a look

I would suggest serious set environments where every item in the frame is chosen or controlled have less requirement for shallow DOF than less controlled environments where one may want to blur distracting objects

S

 

[David Rasberry]

I think isolation of the subject helps the viewer concentrate on what they are 'supposed' to be looking at and therefore helps the viewer understant the image quickly and simply - no fad - part of the toolkit of communicating through images

serious out of focus is a fad or a look

I would suggest serious set environments where every item in the frame is chosen or controlled have less requirement for shallow DOF than less controlled environments where one may want to blur distracting objects

S

No question creative use of DOF is a valuable tool. There are also other ways to control focus of attention in a shot. The format ranges between 16mm size and S35 size for motion pictures provide a practical range of DOF control under a wide range of shooting conditions. But when you reach a situation where you can't keep all of an actor's face in focus in a closeup, as has happened to me shooting documentary stills of stage productions with long lenses on FF35 under low lights, it can be a frustrating problem as much as a useful tool.

 

[Stefan Christou]

This may have been said already. But where are you going to find a f0.9 lens for 2/3" cameras?

Also the real benefit that medium format backs have is that for any given field of view you are on a longer and hence more flattering and natural focal length.

FWIW Shallow depth of field is viable artistically and also useful for storytelling and avoiding judder in tracking shots (grandma, eggs, suck).

I do find it enlightening to understand aperture by it's actual size rather than relative

 

[David Rasberry]

There are still some c-mount Angenieux 25mm f.95 primes for 16mm floating around out there in the used market that will work just fine. It has been very popular with micro 4/3 format shooters even though it doesn't fully cover the sensor, so prices remain high. They still command prices in the used market of $800 and up. Schneider also makes a couple of C-mount hi-resolution industrial lenses,17mm and 25mm, at f.95, about $1500 each.
Angenieux also made some 16mm-44mm T1.3 zooms too. I recently acquired a Rank Taylor Hobson 8-24mm f1.6 16mm zoom that I am looking forward to trying out on Scarlet.

I'm sure medium format can be used to good effect for some types of projects. But for the kind of live performance theatrical and music event shoots I enjoy doing, the DOF characteristics would make it useless. 2/3" is really just about ideal for those kind of low to moderate light level conditions. Much easier to focus, more flexible and forgiving in fast moving situations. Judder is not so much of a problem at 30fps. :devil:

I do appreciate the perspective qualities of longer focal length lenses on medium format cameras. My personal favorites are in the 85-100mm range and they do seem to match the natural perspective of my eyes best anyway.

 

[Shane Betts]

Seems to me this thread is a bit misleading. There is no myth, there is only physics.

The basic fact here is this. If you're framed up on a shot, with the doorframe on the left of frame and the pot plant on the right, you will have chosen a/A focal length and b/A camera to subject distance to give you that frame. Change bodies to a camera with a smaller or larger sensor size and you will have to do one of two things to return to the same frame (door on left, fern on right), either change the camera to subject distance (affects depth of field) or change focal length (affects depth of field). I've had arguments before with people who should know better where they countered with, "But what if you made it a close-up?" or "But if I turned on a 10k". All nonsense designed to defend a position.

You can use all the math you like to counter the argument but, in order to disprove this basic rule of physics, you have to introduce "but if" statements which, no matter how convoluted your formulae, is a very unscientific approach to any subject and tantamount to slight of hand. All other stats must remain consistent for any conclusion to hold up.

Cooke (one assumes they know what they're doing) have, on their website, tables for depth of field for their lenses and hyperfocal distance calculations. Pixel size and ISO ratings are not a part of the formulae.

 

[David Rasberry]

Seems to me this thread is a bit misleading. There is no myth, there is only physics.

The basic fact here is this. If you're framed up on a shot, with the doorframe on the left of frame and the pot plant on the right, you will have chosen a/A focal length and b/A camera to subject distance to give you that frame. Change bodies to a camera with a smaller or larger sensor size and you will have to do one of two things to return to the same frame (door on left, fern on right), either change the camera to subject distance (affects depth of field) or change focal length (affects depth of field). I've had arguments before with people who should know better where they countered with, "But what if you made it a close-up?" or "But if I turned on a 10k". All nonsense designed to defend a position.

You can use all the math you like to counter the argument but, in order to disprove this basic rule of physics, you have to introduce "but if" statements which, no matter how convoluted your formulae, is a very unscientific approach to any subject and tantamount to slight of hand. All other stats must remain consistent for any conclusion to hold up.

Cooke (one assumes they know what they're doing) have, on their website, tables for depth of field for their lenses and hyperfocal distance calculations. Pixel size and ISO ratings are not a part of the formulae.

Well said, but aperture too affects depth of field, so some compensation with aperture change is possible. If you are seeking an ultra shallow depth of field look, nothing will substitute for a larger sensor/longer lens. The opposite is also true. If you want extended depth of field under some conditions a smaller sensor/shorter lens may be preferable.

I was looking at award winning work by new photographers in Digital Photo Pro magazine. Out of the eight images only one, a full face closeup, used shallow depth of field as element of the composition. The rest were fully in focus throughout the depth of the image. Focus of attention in the compositions was controlled by geometry, mass, color, light, shadow, and emotional drama.

 

[Daniel Browning]

I did a test with some stills cameras

Thanks, those are interesting. I'll try to look into them further when I have time.

Also the real benefit that medium format backs have is that for any given field of view you are on a longer and hence more flattering and natural focal length.

14mm on 2/3" has the exact same perspective as 76mm on medium format; therefore, it is just as flattering and natural. See the earlier comments in this thread by Paris Remillard and Eddie.

Seems to me this thread is a bit misleading.

I kindly disagree. The fact is that larger sensors only have thinner DOF if you *want* them to. It's never a requirement nor any kind of natural consequence. Most photographers don't realize that it's possible to keep DOF the same. They think smaller formats have a "deep DOF" advantage, but that's just a myth, even in low light.

The root cause of all these misconceptions is that people have only ever considered f-number. It takes an entire paradigm shift to begin to understand the importance and implications of iris diameter.

[I'm only using the phrase "iris diameter" because the true meaning of the word aperture has been lost to so many people that I can't be understood when I use it correctly.]

There is no myth, there is only physics.

Your position is not bolstered just by throwing out the word "physics".

I've had arguments before with people who should know better where they countered with, "But what if you made it a close-up?" or "But if I turned on a 10k". All nonsense designed to defend a position.

Well, if you see any nonsense in my posts, I'm sure you'll point it out in specific.

You can use all the math you like to counter the argument but, in order to disprove this basic rule of physics,

I don't think you meant that the "larger sensors have thinner DOF" myth is a basic rule of physics. At least, I didn't read about it in any of my physics textbooks.

Cooke (one assumes they know what they're doing) have, on their website, tables for depth of field for their lenses and hyperfocal distance calculations. Pixel size and ISO ratings are not a part of the formulae.

First, I brought up pixel size and ISO ratings to make it simple to explain how larger formats can use narrower f-numbers and still have the same total amount of light and the same noise. I was not referring to their effect on DOF when all other variables are fixed.

Second, there would be nothing incorrect about discussing such effects because they are indeed real. Cooke's information, just like most discussions of DOF (including this thread), only relates to the most important factors, such as the lens, and it makes the discussion much simpler if all the other factors are assumed to be equal. But that does not mean they don't exist. There are a host of other factors between the lens and the viewer's brain that affect the viewer's DOF (acceptably sharp CoC), including filters, OLPF, sensor (including pixel size and ISO ratings), raw conversion, post processing, mastering, display size/resolution, viewer distance/acuity, etc.

Change bodies to a camera with a smaller or larger sensor size and you will have to do one of two things to return to the same frame (door on left, fern on right), either change the camera to subject distance (affects depth of field) or change focal length (affects depth of field).

That statement is based on the assumption that f-number is the same, but such an assumption is arbitrary and unnecessary. Nothing dictates that the same f-number must be used.

In some cases it is not even possible to keep f-number the same. That is precisely the reason why FF35 cameras are capable of much thinner DOF than any MFDB, despite the fact that MFDB have much larger sensors. That is because medium format lenses have narrower maximum iris diameters for the same angle of view.

If you keep maximum f-number the same, it results in:

* Less DOF
* Less diffraction
* Less noise in low light
* Lenses are generally bigger, heavier, and cost more.

If you keep maximum iris diameter the same:

* Same DOF
* Same diffraction
* Same noise in low light
* Lenses are generally close in size, weight, and cost.

Saying "smaller sensors have the advantage of deeper DOF than large sensors" is like saying "commuter sedans have the advantage of slower speeds than sports cars". That's not an advantage, because sports cars can drive slow too.

Large sensors can do anything the small sensors can do, and more.

 

[David Rasberry]

Large sensors can do anything the small sensors can do, and more.

Within limits. One may have to crop a larger sensor to get the necessary depth of field vs FOV with a shorter lens under some circumstances, or suffer higher noise from under exposure with the longer lens stopped down.
There have been instances where I have composed shots with two actors on a stage 15 to 20 feet apart front to back in relatively low light where I wanted both actors faces in focus. Much easier to do with a 2/3" video camera than with a 35mm still camera for the same general composition.

 

[Daniel Browning]

Within limits.

The only limits are the ones I outlined in the OP.

One may have to crop a larger sensor to get the necessary depth of field vs FOV with a shorter lens under some circumstances, or suffer higher noise from under exposure with the longer lens stopped down.

The only time that is the case is when conditions are unequal. When "all else is equal", the myth is busted. By "all else", I mean the same technology, lens transmissivity, raw pre-conditioning, raw conversion, post-processing, display size, and bellows factor.

There have been instances where I have composed shots with two actors on a stage 15 to 20 feet apart front to back in relatively low light where I wanted both actors faces in focus. Much easier to do with a 2/3" video camera than with a 35mm still camera for the same general composition.

The 2/3" video camera had a 3-chip design; that explains why you found it easier. [For the same sensor size and DOF, 3-chip gathers more light than Bayer, but that's not the topic of this thread.] A Bayer 2/3" video camera might have prompted you to reach a different conclusion.

 

[Shane Betts]

Well said, but aperture too affects depth of field, so some compensation with aperture change is possible. If you are seeking an ultra shallow depth of field look, nothing will substitute for a larger sensor/longer lens. The opposite is also true. If you want extended depth of field under some conditions a smaller sensor/shorter lens may be preferable.

I was looking at award winning work by new photographers in Digital Photo Pro magazine. Out of the eight images only one, a full face closeup, used shallow depth of field as element of the composition. The rest were fully in focus throughout the depth of the image. Focus of attention in the compositions was controlled by geometry, mass, color, light, shadow, and emotional drama.

Absolutely right. Depth of field is not a one size fits all thing. Sometimes you want the world in focus, sometimes nothing but the eyes and changing stops will deliver great change. The limiter is when the sensor size you're using (assuming you don't want to change from 35 to mini DV in the one scene) is restrictive and just won't allow you to drop unwanted elements out of focus the way you want to, or when you have to be 100' away on the long end just to get some definition and separation.