Sensor size affect DOF? Methinks not.

SkipD wrote in post #17671375
Folks, it seems that most of the posts here are missing the boat about depth of field calculations.

Actually looks like it's mainly the OP who is missing the boat No-one has agreed with him, as far as I can see

I am still struggling to understand how anyone can look at two photos from vastly different size sensors, taken at the same aperture, distance and focal length, that display identical DOF, can continue to insist that what they are staring at is not possible. Those photos are a demonstration that sensor size does not affect DOF. I sincerely regret opening this can of worms. Apologies all around.

AJSJones wrote in post #17671386
Actually looks like it's mainly the OP who is missing the boat No-one has agreed with him, as far as I can see

Nobody agreed with Copernicus, Galileo, or Columbus, either. Near as I can tell, the earth is still round, and the planets still orbit the sun.

SkipD wrote in post #17671375
The "circle of confusion" is what folks need to understand to begin to understand the DOF calculations.

Indeed. It does all come down to the size of your CoC. Errr. Um.. I mean...

Yes; standard DOF charts assume a particular print size viewed at a particular distance by a person with a particular quality of eyesight. There are just so many variables that it does make comparisons difficult.

The way I see it; if you have the same...

• Framing/field of view
  
• Subject distance
  
• Print size
  
• Viewing distance (for the print)
  
• One viewer (i.e. the same quality of eyesight)
  

GeoKras1989 wrote in post #17671392
I am still struggling to understand how anyone can look at two photos from vastly different size sensors, taken at the same aperture, distance and focal length, that display identical DOF, can continue to insist that what they are staring at is not possible. Those photos are a demonstration that sensor size does not affect DOF. I sincerely regret opening this can of worms. Apologies all around.

They're not the same aperture though (as AJSJones explained). What's happened is that you've managed to produce images with different settings (or crops) that match on different formats; that's fine, but it's different from your assertion.

EDIT: If your posted focal lengths for the G15 are from its stated ~6-30mm range then you've also not matched the focal lengths; as 30mm on the G15 is equivalent to 140mm on full frame.

GeoKras1989 wrote in post #17671392
I am still struggling to understand how anyone can look at two photos from vastly different size sensors, taken at the same aperture, distance and focal length, that display identical DOF, can continue to insist that what they are staring at is not possible. Those photos are a demonstration that sensor size does not affect DOF. I sincerely regret opening this can of worms. Apologies all around.

Your example is the same as setting up the camera, keeping the FL and aperture the same but "cutting out a small piece of the negative" from the 35 mm negative that is the same size as the 110 negative and printing them up the same size - of course they will be the same D'oh! No-one is quibbling with that experiment's outcome - only its relevance! That simply tells us that when you use a crop sized piece of film and a given FL and aperture you get the same results each time We could have cut it out of an 8x10 sheet of film - once you crop the capture area to the same sizes, it doesn't matter what the bigger size sheet was...


The whole point of getting a camera with a bigger sensor is to actually use it, not just the 13x17 bit in the middle When you do, you have to use a different lens - a different lens - a different lens - to get the same FoV to land on the bigger sensor. If you use the same lens, you won't get the same FoV. And we've explained about the "entrance pupil/physical aperture" being different from f stop.

By cropping the 6D you made the sensor size the same.

Nonnit wrote in post #17671427
By cropping the 6D you made the sensor size the same.

A very pithy nutshell

Many folks - photography newbies and experienced photographers alike - assume that the depth of field is analyzed at the film plane or sensor (the "in-camera" image formed inside the rear of the camera by the lens). If that were true, then the OP would be correct in stating that camera format does not affect the depth of field of images. However, the depth of field is definitely NOT analyzed at the film plane or sensor but on an enlargement that's made at a single standard size for images from all camera formats. The difference in magnification between the in-camera images and the standard print size for analysis is what makes a definite change in DOF for various format cameras using the same focal length, same aperture, and same camera-subject distance.

Exactly.

George, read this: http://photo.net/learn​/optics/dofdigital/. It's the best practical explanation of the impact of sensor size on depth of field I have read.

SkipD wrote in post #17671375
The "circle of confusion" is what folks need to understand to begin to understand the DOF calculations.

Which also coincides with crop factor: look up circle of confusion values of different cameras and you'll see they're the same proportion as the sensor size/crop factor. Circle of confusion and enlargement sizes were standardized with Kodak during the film era. With the digital age, and different smaller sized sensors, the physics are the same (and proportionally reduced from the smaller surface area). You later mention any print from the same camera having different DOF with different enlargements: that's only if you go large enough to start exceeding the CoC diameter limit (IE going larger in perceptual resolution then the sensor's native resolution). With any modern camera (which is high MP), I don't see how any would show a different DOF with a 5x7 vs 8x10 for example (printing from the same exposure). Tony Northrup has some good examples of achieving the exact same DOF with FF, crop, or M 4/3rds. Not sure what the OP is trying to prove, as it looks like examples are extreme in focus vs OOF background. All DOF calculators do have DOF ranges that correspond to sensor size (and why there's an argument for considering "format factor" instead of just "crop factor"). If it's a myth that cameras have different DOFs (and specifically crop factors)....well that would mean every photography source is wrong Just as there's a FL equivalent when comparing field of view with camera formats, there's aperture equivalents for exact DOF equivalents. That along with ISO, diffraction limit, or camera stability are some of the main factors in deciding what size format camera to use IMHO.

You're correct in thinking aperture doesn't change with sensor size. Never understood the people who state "equivalent" numbers when changing formats.

If that was the case F3.5 on my 7x7 would be like 0.75 on 35mm

Sensor size doesn't affect DOF any more than sensor size affects the focal length of your lens.

"But wait," you say, "I have this link to an article on the internet which says my crop sensor makes the lens focal length 1.6 times as long." Um, no. The focal length of your lens is the same, and your DOF is the same. People talk about the "effective focal length" of a lens on a crop sensor camera based on how much larger you would have to make a print for it to be the same size as a print from a "full frame" 24x36 mm sensor.

Here's an example. Say you want to make a 4x6" print. Starting from a "full frame" 24x36mm sensor, you would have to enlarge the image captured by the sensor by 4.23 times in each dimension. Starting from a 15x22.5 APS-C crop sensor, you would have to enlarge the image captured by the sensor by 6.77x to get the same 4x6 inch print. 6.77/4.23 = 1.6. Apologies for my freewheeling mixing of imperial and metric units. I'm an American.

So if you had the same lens with the same focal length on both cameras, detail in the image center would be 1.6x as large on the 4x6 inch print from the crop sensor, hence the 1.6x "crop factor."

The same is true with DOF. DOF is based on the "circle of confusion", which, for the purpose of this discussion, is the largest circle that looks effectively like a point on the print. At the image plane, the circle of confusion is the same for the same lens at the same focal ratio. The cone of light that produces a given circle of confusion doesn't know or care if it's striking a full frame sensor, a crop sensor, or a piece of 8x10 sheet film. The difference between sensor sizes only becomes relevant when you make a print (or a larger image). Then, in order to have the same size final image, the image from the crop sensor has to be enlarged by 60% more than the image from the full frame sensor. That enlarges the circles of confusion in the final image. As a result, points that appeared in focus in the image from the full frame sensor may start to look a little soft in the image from the crop sensor, so the DOF appears to be less. This is a result of the greater enlargement and is not an inherent property of the crop sensor.

So in truth, the answer depends on how you evaluate DOF. The OP is correct that sensor size doesn't affect DOF at the sensor when the image is captured. However, the DOF calculator web sites ask you whether you are using full frame or crop sensor because the same size maximum circle of confusion at the image sensor will result in less DOF at a 60% greater final enlargement.

-Mike
I don't pretend to know a lot about optics but there are plenty of people around here who do.

DoF does not exist until you view the image. (print or monitor)
DoF does not exist until you view the image.
DoF does not exist until you view the image.
DoF does not exist until you view the image.
DoF does not exist until you view the image.
DoF does not exist until you view the image.

If you are in front of your monitor and you keep zooming in, you are changing the DoF. Try it on pretty much any image that has any areas OOF. DoF is based on *the perception of sharpness* at a particular print size and viewing distance NOT the sharpness of the original image captured.

AJSJones wrote in post #17671518
DoF does not exist until you view the image. (print or monitor)
DoF does not exist until you view the image.
DoF does not exist until you view the image.
DoF does not exist until you view the image.
DoF does not exist until you view the image.
DoF does not exist until you view the image.

If you are in front of your monitor and you keep zooming in, you are changing the DoF. Try it on pretty much any image that has any areas OOF. DoF is based on *the perception of sharpness* at a particular print size and viewing distance NOT the sharpness of the original image captured.

VERY true....

AJSJones wrote in post #17671518
If you are in front of your monitor and you keep zooming in, you are changing the DoF. Try it on pretty much any image that has any areas OOF. DoF is based on *the perception of sharpness* at a particular print size and viewing distance NOT the sharpness of the original image captured.

That correlates to perceptual resolution (and is important for enlargements that exceed CoC...if your perceptual resolution doesn't exceed the original image's captured resolution, you'll see no difference in sharpness). Guess there is some difference with nomenclature/terms (where you say DOF, I say perceptual resolution)...but I think we are agreeing with major concepts. I think the only main point I'm adding to is that with digital images: perception of sharpness first starts with perceptual resolution (the arguments with DOF are for larger prints where the CoC is exceeding perceptual resolution: and becomes more apparent with close viewing distance/etc). Aperture is a function of the lens, and does seem to correspond to sensor size the way FL does... it's all equivalences, and there are pros and cons of different sensor sizes/may be no difference depending on application.