Sensor size and Hyperfocal range (with the same lens)

[Chris Giles]

Hi all,

I've a question about Hyperfocal distance as I'm seeing all sorts of cross talk involving actual focus, perceived focus and so on.

If shooting a scene with a lens at say, 35mm focal length on a full frame sensor (could be medium format 6x7, 35mm FF or 6x4.5) at F8 would the area of focus be greater if using the same lens on a smaller sensor and focusing at the same distance at F8?

The reason I ask is that I'm using a wide angle zoom and am considering a smaller sensor body so I can get more reach with the same setting and flash power without resorting to a longer lens with a narrower DOF and searching for a correct understandable answer is appearing problematic.

 

[Dogs857]

I can't say in fully understand the maths but apparently yes sensor size makes a difference.

Punch some numbers into this and see what happens.

http://www.photopills.com/calculators/dof

 

[Godfrey]

Hi all,

I've a question about Hyperfocal distance as I'm seeing all sorts of cross talk involving actual focus, perceived focus and so on.

If shooting a scene with a lens at say, 35mm focal length on a full frame sensor (could be medium format 6x7, 35mm FF or 6x4.5) at F8 would the area of focus be greater if using the same lens on a smaller sensor and focusing at the same distance at F8?

The reason I ask is that I'm using a wide angle zoom and am considering a smaller sensor body so I can get more reach with the same setting and flash power without resorting to a longer lens with a narrower DOF and searching for a correct understandable answer is appearing problematic.

I used these two website tools to show you the difference between what you get with a Nikon DX vs Nikon FX sensor and a Field of View Calculator
DOFMaster Depth of Field Calculator

You can use them to plug in your format and lens choices to see what you get.

Nikon FX FoV and DoF:

Nikon DX FoV and DoF:

This pair of calculations holds focal length, focus distance, and aperture setting constant, letting field of view change. Notice that the 35mm lens on the smaller format has significantly less FoV, thus greater magnification in the captured image, which reduces DoF at the same lens opening. If you switched to a 24mm lens on the Nikon DX, be holding FoV constant, not focal length: Then you'd see similar FoV and magnification, and more DoF at the same aperture setting.

Make sense? Play with the calculators and see how to get what you want.

G

 

[gerald.d]

I'm not sure that's correct.

The two formats do not have a different magnification of the captured image.

The crop sensor is just that - a crop of what would be captured under identical settings with a larger sensor.

There is no difference in depth field of a captured image due to changing sensor size whilst maintaining focal length, aperture, and distance to subject.

Where there is a difference is in the depth of field of the print, assuming both images are printed to the same size and viewed from the same distance.

Kind regards,


Gerald.

 

[ErikKaffehr]

Hi,

My take is that if you used the same focal length/aperture combo the CoC for the same amount of defocus would be exactly the same, as it doesn't depend on the sensor.

But, with a smaller sensor you would need a higher magnification, so you would need a small CoC for perceived sharpness. So, a smaller sensor will give less DoF.

On the other hand, a smaller sensor needs a different focal length for the field of view, say you would use an 60 mm lens instead of 90 mm and that would give you a larger DoF.

Using a DoF calculator might be a good idea.







Best regards
Erik

Hi all,

I've a question about Hyperfocal distance as I'm seeing all sorts of cross talk involving actual focus, perceived focus and so on.

If shooting a scene with a lens at say, 35mm focal length on a full frame sensor (could be medium format 6x7, 35mm FF or 6x4.5) at F8 would the area of focus be greater if using the same lens on a smaller sensor and focusing at the same distance at F8?

The reason I ask is that I'm using a wide angle zoom and am considering a smaller sensor body so I can get more reach with the same setting and flash power without resorting to a longer lens with a narrower DOF and searching for a correct understandable answer is appearing problematic.

 

[Godfrey]

I'm not sure that's correct.

The two formats do not have a different magnification of the captured image.

The crop sensor is just that - a crop of what would be captured under identical settings with a larger sensor.

There is no difference in depth field of a captured image due to changing sensor size whilst maintaining focal length, aperture, and distance to subject.

Where there is a difference is in the depth of field of the print, assuming both images are printed to the same size and viewed from the same distance.

(bolded) That is the point: DoF calculations, the CoC etc, are always made with respect to a reference image size (usually a print, usually 8x10 inches (or metric equivalent) in size). Creating an image to that size from an APS-C size format implies more magnification of the image than from FF format.

G

 

[gerald.d]

(bolded) That is the point: DoF calculations, the CoC etc, are always made with respect to a reference image size (usually a print, usually 8x10 inches (or metric equivalent) in size). Creating an image to that size from an APS-C size format implies more magnification of the image than from FF format.

G

Hi Godfrey -

Does that still hold true in the digital age though?

Would it be true to say that a 24 megapixel APS-C image would require more "magnification" (all other things being equal) to print at the same size as a 6 megapixel FF DSLR image?


Kind regards,


Gerald.

 

[ErikKaffehr]

Hi,

In CoC terms yes. A high pixel camera just puts more pixels in that unsharp area. The CoC of confusion is the same, just more exactly described.

To complicate the issue, it is possible to deconvolve the CoC and having more pixels may be helpful in that.

Best regards
Erik

Hi Godfrey -

Does that still hold true in the digital age though?

Would it be true to say that a 24 megapixel APS-C image would require more "magnification" (all other things being equal) to print at the same size as a 6 megapixel FF DSLR image?


Kind regards,


Gerald.

 

[dchew]

Hi all,

I've a question about Hyperfocal distance as I'm seeing all sorts of cross talk involving actual focus, perceived focus and so on.

If shooting a scene with a lens at say, 35mm focal length on a full frame sensor (could be medium format 6x7, 35mm FF or 6x4.5) at F8 would the area of focus be greater if using the same lens on a smaller sensor and focusing at the same distance at F8?

The reason I ask is that I'm using a wide angle zoom and am considering a smaller sensor body so I can get more reach with the same setting and flash power without resorting to a longer lens with a narrower DOF and searching for a correct understandable answer is appearing problematic.

Chris, the two areas of bolded text imply a few things:

1. You will not change the camera position.
2. You would crop the final output file in the FF case to get the "reach" you need.

I want to "focus" on your first assumption about longer lenses having less DOF. If you do not change the camera-to-subject distance, the longer lens will have the same DOF as the wider lens after cropping on that FF camera, assuming the same f-stop and a reasonable magnification (not macro). If the final output angle of view is going to be the same, and your camera-to-subject position is going to be the same, then all this falls out in the wash (aka the crop). Just use the FF camera and the lens that provides the view you want. There are some minor effects of pixel pitch on COC, but I think that only applies to pixels, i.e. viewing at 100%; not a scaled output.

If you are considering different final output views in your two circumstances, then the one that gets magnified the least will have the most DOF. Again, all this assumes the camera position doesn't change (focusing at the same distance).

Dave

 

[ondebanks]

Hi Godfrey -

Does that still hold true in the digital age though?

Would it be true to say that a 24 megapixel APS-C image would require more "magnification" (all other things being equal) to print at the same size as a 6 megapixel FF DSLR image?


Kind regards,


Gerald.

Hi Gerald,

It doesn't matter what the pixel count is. Magnification in imaging is with respect to the physical image size projected onto the sensor, not how that sensor is subdivided into pixels. People just aren't as aware of this in the digital age because they don't see the captured image in its native size anymore; it's not like looking at various sized pieces of film on a lightbox.

Ray

 

[segedi]

Would the area of focus be greater?
In my experience, yes, as I could compare my 110mm f/2.8 lens on 6x7 to a 50mm f/1.4 lens on 35mm. When shot wide open, they offer about the same area of focus. The is highly distance based, as you know, with focus at nearer distances providing less of an area of focus.

The difference will be more noticeable if you are accustomed to a 50mm focal length on medium format and then use the same lens and aperture on 35mm (36x24mm) - f8 will give a greater area of focus on the smaller format. The effect would be less so on full frame 35mm compared to APS-C.

 

[ondebanks]

I used these two website tools to show you the difference between what you get with a Nikon DX vs Nikon FX sensor

This pair of calculations holds focal length, focus distance, and aperture setting constant, letting field of view change.

However it also lets the circle of confusion change, and that is the real reason why the calculated DOF changes.

If you enlarged both images by the same factor, clearly the DX print will be smaller than the FX print, but the DOF will be the same in both.

That calculator is not based on equal enlargements/magnifications, but on equal print sizes; and when you enlarge a smaller-sensor image by a larger factor to match the print sizes, you also enlarge the impression of blurriness in the areas just fore and aft of the plane of focus. The calculator adjusts for this by imposing a smaller limit to the circle of confusion in the original capture. You are now judging the same image projected by the same lens [in the areas where the sensors overlap] by a harsher criterion. But that is what the eye's tolerance for unsharpness demands from equal print sizes.

So, back to Chris' original question: "would the area of focus be greater if using the same lens on a smaller sensor and focusing at the same distance at F8?"
Answer: No. It will be either the same, or smaller. The same, if you enlarge them by the same factor, and end up with a smaller print from the smaller sensor. Smaller, if you enlarge them to the same print size.

Ray

 

[ondebanks]

Would the area of focus be greater?
In my experience, yes, as I could compare my 110mm f/2.8 lens on 6x7 to a 50mm f/1.4 lens on 35mm. When shot wide open, they offer about the same area of focus. The is highly distance based, as you know, with focus at nearer distances providing less of an area of focus.

That's not a good example for this thread, as you are talking about a different scenario - two different focal lengths.

The difference will be more noticeable if you are accustomed to a 50mm focal length on medium format and then use the same lens and aperture on 35mm (36x24mm) - f8 will give a greater area of focus on the smaller format. The effect would be less so on full frame 35mm compared to APS-C.

That is not physically possible. Please check my most recent post here.

Ray

 

[Godfrey]

Does that still hold true in the digital age though?

Would it be true to say that a 24 megapixel APS-C image would require more "magnification" (all other things being equal) to print at the same size as a 6 megapixel FF DSLR image?

You're still adjusting the size of the image from what the lens images on the sensor to where the ink is laid down on paper by the same amount. Don't think about the pixel count ... that affects resolution and dynamic range, not image geometry.

The evidence supports that it continues to work as the calculators suggest it ought to.

I tried it: I have both FF and Micro-FourThirds format cameras, 50mm and 25mm "normal" lenses respectively, and 35mm lenses for both.

I made the same photograph with my Leica M-P and my Olympus E-M1, using a 50mm lens at f/8 on the on Leica and 25mm lens at f/8 and f/4 on the Olympus. That's the "same FoV-same DoF comparison" ... printed full frame to 8x10s, the 50-f/8 and 25-f/4 images were identical, the 25-f/8 image had two stops more DoF.

I then made photos of the same target with 35-f/8 on the Leica and 35-f/8 on the Olympus: that's the "same focal length-same aperture comparison" ... printed full frame to 8x10s, the Leica image showed more FoV and more DoF, the Olympus image showed less FoV and less DoF.

I suggest you do some experimenting. It's the best way to get a feel for this kind of thing IMO, rather than trying to work it all out by numbers and debating the opinions of others who haven't done the experiments. You can replicate my experiment with a single camera, say a FF Nikon or Canon (whatever you have to hand):

1. Make a test exposure with a 50mm lens*at f/8, then a 25mm lens at f/8 and f/4 (or a 35mm if you want to compare to APS-C).
2. Print an 8x10 of the 50mm image.
3. Crop the 25mm image to the same number of pixels as either an APS-C or FourThirds format would have produced.
4. Print 8x10s of the 25mm images.
5. Make a test exposure with a 35mm lens at f/8.
6. Print an 8x10 of the 35mm image.
7. Crop the 35mm image to the same number of pixels as either an APS-C or FourThirds format would have produced.
8. Print an 8x10 of the cropped 35mm image.
9. Now compare them.

You don't have to do any calculation of COC or whatever to see the differences.
G

 

[gerald.d]

Hi Godfrey -

I appreciate you taking the time to go through this, and I think you're right - I'm going to have to test this out for myself.

The situation with regards different focal lengths, or moving the camera position to maintain the same FoV I totally get.

But I simply cannot get my head around why using the exact same lens from the exact same position on different sensor sizes would change the depth of field.

This article - https://photographylife.com/sensor-crop-factors-and-equivalence - claims that it doesn't.


Kind regards,


Gerald.

 

[Godfrey]

I appreciate you taking the time to go through this, and I think you're right - I'm going to have to test this out for myself.

The situation with regards different focal lengths, or moving the camera position to maintain the same FoV I totally get.

But I simply cannot get my head around why using the exact same lens from the exact same position on different sensor sizes would change the depth of field.

This article - https://photographylife.com/sensor-crop-factors-and-equivalence - claims that it doesn't.

That's why I don't read other people's theoretical articles. I test for myself.

If it didn't make any difference, then why all these insane debates about why FourThirds format doesn't give you the ability to make shallow depth of field, why FF lets you get beautiful bokeh, why MFD gives you such shallow and controllable DoF, etc etc?

The debates have been going on for a decade and a half. I did my first experiments on this when I bought my Canon 10D in 2003: I did it comparing the DoF using its APS-C 1.6x crop vs simulating the Sony F707's 5/8" sensor format. I've repeated the exercise with other APS-C, FourThirds and FF cameras over and over again since. The results are always the same: the calculators work.

Try it.

G

 

[ondebanks]

You're still adjusting the size of the image from what the lens images on the sensor to where the ink is laid down on paper by the same amount. Don't think about the pixel count ... that affects resolution and dynamic range, not image geometry.

I then made photos of the same target with 35-f/8 on the Leica and 35-f/8 on the Olympus: that's the "same focal length-same aperture comparison" ... printed full frame to 8x10s, the Leica image showed more FoV and more DoF, the Olympus image showed less FoV and less DoF.

Hi Godfrey,

Those are exactly the same points I made - glad to see we are in agreement. :salute:

I suggest you do some experimenting. It's the best way to get a feel for this kind of thing IMO, rather than trying to work it all out by numbers and debating the opinions of others who haven't done the experiments.

While as a physicist and observational astronomer, I will never dismiss the process of experimental verification, this truth also holds: well-established physics (like geometrical optics) does not yield to "debate" or "opinions". One does not argue with gravity, for example!

Cheers,
Ray

 

[Godfrey]

Hi Godfrey,

Those are exactly the same points I made - glad to see we are in agreement. :salute:

...

While as a physicist and observational astronomer, I will never dismiss the process of experimental verification, this truth also holds: well-established physics (like geometrical optics) does not yield to "debate" or "opinions". One does not argue with gravity, for example!

Cheers,
Ray

I agree, Ray. However, the vast majority of photographers are neither physicists, astronomers, nor even mathematicians (my particular affliction :scry, so there is an absurd amount of debate on subjects like geometrical optics for the simple reason that they aren't really very aware of the math and physics, nor know how to recognize real information vs hokery when the discussion is theoretical.

That's why I urge photographers to do easy experiments that demonstrate what happens clearly and simply. The level of experimentation required to understand photographic behaviors isn't rocket science.

enjoy!
G

 

[MGrayson]

well-established physics (like geometrical optics) does not yield to "debate" or "opinions".

Oh, agreed! The problem is that people don't always ask the question to which they actually want the answer. "Equivalence" debates rage between formats even though the equations are VERY easy to solve.

People use words like Depth of Field, but there are many reasonable definitions, and it's hard to convince everybody to at least start from the accepted definition in the literature, and then, if they want something else, figure out how their something else relates to that definition.

I enjoy coming up with non-technical explanations of these results, but they're no help if the question itself isn't defined.

Oh, and +1 on experimentation. In the end, we're all after looks. Sharp, pleasing bokeh, weird bokeh, whatever, and measurements really just aren't the same thing.

Best,

Matt

 

[GrahamWelland]

OMG - look into a pit and dive into it ...

Hyperfocal effectiveness is based on prints, not necessarily pixel peeping. If you pixel peep, use root(2) x pixel pitch for sharpness considerations. If you look at real photos printed on real paper, use the traditional calcs for viewing the prints and just relax. Heck even the DoF scale on your lens will work out just fine.

If you pixel peep, you'll just be disappointed and frustrated with your 1mm in focus region. :facesmack: