Depth of Field & Sensor Sizes

[John Sheehy]

I was watching a video course on the foundations of Photography, and course material says that a smaller sensor camera is better suited to landscapes because it has a broader depth of field vs. a larger (i.e. Full Frame) sensor. Is this true?

Not exactly. What is true is that if you do need significant DOF, then a small sensor is not an impediment to SNR, but this is only true above base ISO, when you need a specific shutter speed. On a tripod with a static scene, you can give each sensor a standard base-ISO exposure, and the larger the sensor, the more total light is detected. If your subject is still and the camera is still, however, you can take as many images as you want and add them together, for minimal noise and maximum sharpening potential.

It seems to me that full frame cameras are used quite a lot in landscape photography as well. Does it become difficult to get a broader depth of field (that is, no blurriness in the photo) with a full frame camera under certain conditions, or does this assertion have no basis in reality?

Not really. Many FF lenses stop down pretty far, providing as much DOF or more than some smaller-sensor formats that have limited ability to stop the lenses down. You have to look at the available maximum f-numbers and the focal lengths and do the math, to see the potential DOF. If two systems have the same field of view, then the smaller the entrance pupil, the more DOF there is, and the larger the pupil, the less DOF there is. You get the pupil by dividing the true focal length (not the equivalent one) by the f-number. So if you use 24mm on FF, and 12mm on m43, the m43 will only give more DOF if the f-number is greater than half that used on the FF. IF the m43 lens stops at f/16, and the FF at f/32, then they have the same maximum DOF. If the FF stops at f/22, then the m43 will have more maximum DOF. No matter how many ways you try to look at things, it is the entrance pupil that is actually at the core of all non-time-related photographic characteristics.

What is absolutely true is that for the smaller sensors, there are no lenses that can give as shallow a DOF as larger sensors can, with available lenses. Lenses "faster" than f/1.4 are rare, faster than f/0.9 virtually non-existent.

Smaller sensors are an obstacle, when you want the shallowest DOF at wide and medium angles of view, but that does not make them superior at deep DOF; that is limited only by how small lens manufacturers are willing to go, with entrance pupils.

 

[MarshallG]

Nailed it!

 

[lehill]

I've wanted to try this for a while and finally got around to it today. It's very simple. I put the very same 50mm lens on two cameras (FF & APS-C) and took two photos at the same distance (~ 80 cm), and the same f-number (f/1.4). The results are predictable but perhaps enlightening to someone fresh to photography.

I used my old Lenscal focus calibrator with scale markings adjusted for a 45° angle, ie, the distance between each major hashmark measures 1.4 cm. So the "2" numerals on the scale are 2 cm in front or behind the target.

How did you view the images to decide what was in-focus and what was out-of-focus?

I just eyeballed it and asked myself "What's not fuzzy?".

No matter what size I view the image the "3" numerals are always fuzzy. The "2" numerals are sorta borderline so I made the call and said the DOF is about ±2 cm.

It's not going to bother me if someone wants to determine DOF differently.

 

[MarshallG]

I've wanted to try this for a while and finally got around to it today. It's very simple. I put the very same 50mm lens on two cameras (FF & APS-C) and took two photos at the same distance (~ 80 cm), and the same f-number (f/1.4). The results are predictable but perhaps enlightening to someone fresh to photography.

I used my old Lenscal focus calibrator with scale markings adjusted for a 45° angle, ie, the distance between each major hashmark measures 1.4 cm. So the "2" numerals on the scale are 2 cm in front or behind the target.

Full-frame camera (A7RV, 61 mp):


I'll say the DOF is about ± 2 cm. People with better eyes might say ± 1 cm or even less. Hover your mouse pointer over this photo for EXIF info.

APS-C camera (A6500, 24 mp):


Looking at the scale the DOF measures the same to me. Hover your mouse pointer over the photo for EXIF info.

So when someone says "APS-C cameras have more depth-of-field." what do they really mean? If we measure it with a scale, the DOF seems to be the same.

Thank you for the test. Hovewer concept is not correct in my opinion.

I use or used FF, apsc and m4/3 cameras. My favourite lens is 35mm FF eq. prime. I don't really care what number is written on the lens, I want my favourite field of view - 35mm on FF, 24mm on apsc or 17mm on m4/3.

So, you should do the test with 33 or 35mm apsc lens to match the field of view of 50mm FF lens. So yes, apsc cameras in general offer more depth of field.

No, APS-C cameras do not provide more depth of field than full frame cameras. The depth of field from the shorter focal length you use on the APS-C camera can easily be offset with a full frame camera by using a smaller aperture, without losing quality.

Remember that if you’re doing a critical comparison between the two, depth of field is about blur.

 

[chrisfisheye]

Depth of field only has three factors: the lens diameter, the distance to the subject, and a subjective factor for how tolerant your eyes are to blur.

Can somebody explain me this ?

The post has many likes, so maybe one of these persons can explain. But this looks very surprising to me...

 

[Tom Axford]

Depth of field only has three factors: the lens diameter, the distance to the subject, and a subjective factor for how tolerant your eyes are to blur.

Can somebody explain me this ?

The post has many likes, so maybe one of these persons can explain. But this looks very surprising to me...

This post may help to explain it. Or this post goes into a little more detail about the physics.

 

[chrisfisheye]

Depth of field only has three factors: the lens diameter, the distance to the subject, and a subjective factor for how tolerant your eyes are to blur.

Can somebody explain me this ?

The post has many likes, so maybe one of these persons can explain. But this looks very surprising to me...

This post may help to explain it. Or this post goes into a little more detail about the physics.

Magnification depends on focal length.

Why don't you just show me the equation with the 3 factors that Mark refers to ?

 

[Tom Axford]

Depth of field only has three factors: the lens diameter, the distance to the subject, and a subjective factor for how tolerant your eyes are to blur.

Can somebody explain me this ?

The post has many likes, so maybe one of these persons can explain. But this looks very surprising to me...

This post may help to explain it. Or this post goes into a little more detail about the physics.

Magnification depends on focal length.

Why don't you just show me the equation with the 3 factors that Mark refers to ?

I think Mark has omitted a variable such as the angle of view.

 

[chrisfisheye]

Depth of field only has three factors: the lens diameter, the distance to the subject, and a subjective factor for how tolerant your eyes are to blur.

Can somebody explain me this ?

The post has many likes, so maybe one of these persons can explain. But this looks very surprising to me...

This post may help to explain it. Or this post goes into a little more detail about the physics.

Magnification depends on focal length.

Why don't you just show me the equation with the 3 factors that Mark refers to ?

I think Mark has omitted a variable such as the angle of view.

Thanks Tom for your answer.

 

[Bill Ferris]

I was watching a video course on the foundations of Photography, and course material says that a smaller sensor camera is better suited to landscapes because it has a broader depth of field vs. a larger (i.e. Full Frame) sensor. Is this true?

It's an odd opinion founded upon misunderstandings about cameras of different formats.

Cameras of different formats cam be used to make the same photo. Equivalence is a tool photographers use to determine which settings allow different-format cameras to make the same photo. By definition, a larger format camera will work with the same shutter speed and a larger f-number to make a photo equivalent to an image made with a smaller format camera.

I suppose the author(s) of the support material interpret a smaller f-number delivering the same depth of field as some kind of advantage. What they're overlooking is that a full-frame camera can easily match the depth of field a smaller format camera works with. Full-frame lenses typically go to at least f/22. An f-stop of between f/8 and f/16 will produce more than enough depth of field at the focal lengths commonly used in landscape work.

The support material also overlooks the potential advantage a full-frame camera brings to the table of capturing more total light at the same exposure. Thos translates directly to increased dynamic range and less noise.

This isn't to say that fantastic landscape photos can't be made using APS-C or micro fourthirds systems. But if any system has a great potential advantage in terms of image quality, it's full-frame...or medium format, if you're serious about the genre.

It seems to me that full frame cameras are used quite a lot in landscape photography as well. Does it become difficult to get a broader depth of field (that is, no blurriness in the photo) with a full frame camera under certain conditions, or does this assertion have no basis in reality?

The author(s) of the support material would probably counter what I've written above by saying something like, "If the full-frame photographer wants to be at f/16 for extreme depth of field, I can use my APS-C camera at f/16 for even more depth of field and the same exposure."

What they're overlooking, is that smaller format cameras experience the same detail-robbing effects of diffraction at faster apertures. If a full-frame camera is on the verge of significant distraction at f/16, an APS-C camera capturing the same angle of view, will be on that same precipice at f/11 and experience noticeable diffraction effects at f/16.

Also, at the same f-stop and shutter speed in the same light, all formats work with the same exposure...but not the same total light energy. The larger the format, the more total light energy is captured during a shutter actuation.

 

[lehill]

I've wanted to try this for a while and finally got around to it today. It's very simple. I put the very same 50mm lens on two cameras (FF & APS-C) and took two photos at the same distance (~ 80 cm), and the same f-number (f/1.4). The results are predictable but perhaps enlightening to someone fresh to photography.

I used my old Lenscal focus calibrator with scale markings adjusted for a 45° angle, ie, the distance between each major hashmark measures 1.4 cm. So the "2" numerals on the scale are 2 cm in front or behind the target.

Full-frame camera (A7RV, 61 mp):


I'll say the DOF is about ± 2 cm. People with better eyes might say ± 1 cm or even less. Hover your mouse pointer over this photo for EXIF info.

APS-C camera (A6500, 24 mp):


Looking at the scale the DOF measures the same to me. Hover your mouse pointer over the photo for EXIF info.

So when someone says "APS-C cameras have more depth-of-field." what do they really mean? If we measure it with a scale, the DOF seems to be the same.

Thank you for the test. Hovewer concept is not correct in my opinion.

I use or used FF, apsc and m4/3 cameras. My favourite lens is 35mm FF eq. prime. I don't really care what number is written on the lens, I want my favourite field of view - 35mm on FF, 24mm on apsc or 17mm on m4/3.

So, you should do the test with 33 or 35mm apsc lens to match the field of view of 50mm FF lens. So yes, apsc cameras in general offer more depth of field.

I'd do the very same thing and even use something like DOF Master to check if the resulting DOF suits my needs. But I don't believe that was the OP's question.

Taking two photos while holding as many variables constant as I could seemed a pretty simple way to show sensor size alone does not affect depth of field.

I probably should have quoted the OP's post and highlighted relevant parts.

--
Lance H

 

[Michael Fryd]

I was watching a video course on the foundations of Photography, and course material says that a smaller sensor camera is better suited to landscapes because it has a broader depth of field vs. a larger (i.e. Full Frame) sensor. Is this true?

It seems to me that full frame cameras are used quite a lot in landscape photography as well. Does it become difficult to get a broader depth of field (that is, no blurriness in the photo) with a full frame camera under certain conditions, or does this assertion have no basis in reality?

The simple answer is that at the same angle of view, same shutter speed, and same aperture diameter, you get the same result, independent of sensor size.

By "same result", I mean the same framing, same depth of field, same diffraction, same motion blur, and same overall image noise.

.

For instance, consider someone shooting a landscape with a 50mm lens at f/8, 1/400 and ISO 400. That's a 46° angle of view, and a 6.25mm aperture diameter (50mm /8 = 6.25mm).

Perhaps the battery dies and he wants to switch to a 2X crop body, and continue shooting. With the 2X crop body, he could use a 50mm lens. However, that would give a different field of view and therefore different results. He may very well choose to shoot with a 25mm lens to get that same 46° angle of view.

He could continue to shoot at f/8, but 25mm/8 is a different aperture diameter, and that's going to give a different depth of field, and different diffraction. He may very well choose to keep the same 6.25mm aperture diameter, and that's f/4 on a 25mm lens.

He could keep targeting an ISO 400 exposure, but that's noisier on a 2X crop body. He may target an ISO 100 and will get an image with the same lightness and overall image noise.

The bottom line is that crop bodies don't yield more depth of field. However, you will get more depth of field if you shoot them using the same f/stop as a full frame. But just as you wouldn't use the same focal length, why would you use the same f/stop.

The advantage of a full frame is that you generally have the option of using wider aperture diameters. Those larger diameters yield shallower depth of field and a corresponding decrease in noise. If you don't need the shallow depth of field, you may not need a full frame camera.

Full frame cameras may also offer wider lenses than what you can get with some crop bodies.

.

The bottom line is that any shot you can get with a crop body can almost always be matched with a full frame. However, full frames can allow you to get some shallow depth of field shots that you can't get with a crop body.

One generally doesn't need shallow depth of field for landscapes, and therefore one can save some money by buying a crop body, rather than a full frame. You won't get more depth of field, but you might save some money.

 

[Dick Dastardly]

An M43 camera is better for landscapes because i can shoot at f1.7 all day long and still get sufficient DoF, with FF you need F11 and a tripod to do the same, but the overall IQ is somewhat better.

The best compromise is APSC, especially when considering price.

 

[Michael Fryd]

An M43 camera is better for landscapes because i can shoot at f1.7 all day long and still get sufficient DoF, with FF you need F11 and a tripod to do the same, but the overall IQ is somewhat better.

The best compromise is APSC, especially when considering price.

Not really.

An M43 camera has a 2X crop factor. At the same angle of view a full frame at f/3.5 will yield the same depth of field as an M43 at f/1.7.

Consider an M43 camera with a 14mm lens (75° angle of view) at f/1.7. If you focus the lens at 28', your depth of field will range from 13.5' to infinity.

With a full frame you need a 28mm lens to get that same 75° angle of view. If you focus that lens at 28' you get the same depth of field (13.5' to infinity).

If you kept the same shutter speed, you would need to set the ISO on the full frame two stops higher. As full frames have a reputation for two stops better noise, you end up with the same overall image quality as the crop body.

At the same shutter speed you get the same amount of motion blur. If you didn't need a tripod with the M43, then you won't need one with the full frame (You might need a tripod if you have more hand shake holding a heavier full frame camera).

 

[Martin_99]

An M43 camera is better for landscapes because i can shoot at f1.7 all day long and still get sufficient DoF, with FF you need F11 and a tripod to do the same, but the overall IQ is somewhat better.

Equivalent of m4/3 f1.7 is FF f3.4 (crop factor 2). Your landscape photos are probably kind of flat if have sufficient DoF at this setting.

Tripod is a great tool for landscape, but not necessary, often stabilisation do it's job well, for all formats.

The best compromise is APSC, especially when considering price.

 

[MarshallG]

An M43 camera is better for landscapes because i can shoot at f1.7 all day long and still get sufficient DoF, with FF you need F11 and a tripod to do the same, but the overall IQ is somewhat better.

The best compromise is APSC, especially when considering price.

Not really.

An M43 camera has a 2X crop factor. At the same angle of view a full frame at f/3.5 will yield the same depth of field as an M43 at f/1.7.

Consider an M43 camera with a 14mm lens (75° angle of view) at f/1.7. If you focus the lens at 28', your depth of field will range from 13.5' to infinity.

With a full frame you need a 28mm lens to get that same 75° angle of view. If you focus that lens at 28' you get the same depth of field (13.5' to infinity).

If you kept the same shutter speed, you would need to set the ISO on the full frame two stops higher. As full frames have a reputation for two stops better noise, you end up with the same overall image quality as the crop body.

At the same shutter speed you get the same amount of motion blur. If you didn't need a tripod with the M43, then you won't need one with the full frame (You might need a tripod if you have more hand shake holding a heavier full frame camera).

Of course, this isn’t what most landscape photographers would do. We would use the full frame at ISO 100, with the shutter as long as it needs to be for the aperture we want… and the result is the full frame shot is better than the APS-C or 4/3.

So in the worst case, full frame is equal, but overall it lets you take better photos.

 

[Michael Fryd]

An M43 camera is better for landscapes because i can shoot at f1.7 all day long and still get sufficient DoF, with FF you need F11 and a tripod to do the same, but the overall IQ is somewhat better.

The best compromise is APSC, especially when considering price.

Not really.

An M43 camera has a 2X crop factor. At the same angle of view a full frame at f/3.5 will yield the same depth of field as an M43 at f/1.7.

Consider an M43 camera with a 14mm lens (75° angle of view) at f/1.7. If you focus the lens at 28', your depth of field will range from 13.5' to infinity.

With a full frame you need a 28mm lens to get that same 75° angle of view. If you focus that lens at 28' you get the same depth of field (13.5' to infinity).

If you kept the same shutter speed, you would need to set the ISO on the full frame two stops higher. As full frames have a reputation for two stops better noise, you end up with the same overall image quality as the crop body.

At the same shutter speed you get the same amount of motion blur. If you didn't need a tripod with the M43, then you won't need one with the full frame (You might need a tripod if you have more hand shake holding a heavier full frame camera).

Of course, this isn’t what most landscape photographers would do. We would use the full frame at ISO 100, with the shutter as long as it needs to be for the aperture we want… and the result is the full frame shot is better than the APS-C or 4/3.

So in the worst case, full frame is equal, but overall it lets you take better photos.

That's certainly an option. It's one of the advantages of the full frame over a crop body. In situations where you are not light limited (i.e. you can use an arbitrarily long shutter speed) then the larger sensor can generally capture more total light. As a general rule, the more total light you capture, the better the image quality. Total light is the exposure (light per unit area) multiplied by the sensor size.

However, in situations where you are not using a tripod, motion blur applies the same shutter speed limits to full frame as a crop body.

There is also the question of "good enough". At some point the quality is good enough for the dynamic range to be sufficient to capture the scene, and produce a noise level that is not visible to the unaided human eye. Once you hit that point, there may be little advantages to additional increases in image quality. If your images from a crop body are visually noise free, then you may not get a visible difference by moving to a full frame.

 

[J A C S]

That's certainly an option. It's one of the advantages of the full frame over a crop body. In situations where you are not light limited (i.e. you can use an arbitrarily long shutter speed) then the larger sensor can generally capture more total light. As a general rule, the more total light you capture, the better the image quality. Total light is the exposure (light per unit area) multiplied by the sensor size.

There is also the ability to get more light with a larger aperture without scarifying the needed DOF. With smaller sensors, the DOF is often deeper than needed.

 

[Michael Fryd]

That's certainly an option. It's one of the advantages of the full frame over a crop body. In situations where you are not light limited (i.e. you can use an arbitrarily long shutter speed) then the larger sensor can generally capture more total light. As a general rule, the more total light you capture, the better the image quality. Total light is the exposure (light per unit area) multiplied by the sensor size.

There is also the ability to get more light with a larger aperture without scarifying the needed DOF. With smaller sensors, the DOF is often deeper than needed.

Yes. Full frame bodies tend to offer the option of shallower depth of field. If you are shooting your crop body with the lens wide open, and you want shallower depth of field, you may benefit from moving to a full frame.

At the same shutter speed, you're not going to get more total light without moving to shallower depth of field. That's independent of sensor size.

If you are stopped down a couple stops on your crop body, then you can open up the crop body lens instead of moving to a full frame. At the same angle of view, it's the aperture diameter, not the f/stop, that we care about. Full frame usually offers the option of a larger maximum aperture diameter.

 

[mujana]

An M43 camera is better for landscapes because i can shoot at f1.7 all day long and still get sufficient DoF, with FF you need F11 and a tripod to do the same, but the overall IQ is somewhat better.

Equivalent of m4/3 f1.7 is FF f3.4 (crop factor 2). Your landscape photos are probably kind of flat if have sufficient DoF at this setting.

Tripod is a great tool for landscape, but not necessary, often stabilisation do it's job well, for all formats.

(also) correct. So much so, that I’m still in doubt wether to take my tripod on my next holiday, or not.
Well, probably going to take it anyway. Not sure yet.

The best compromise is APSC, especially when considering price.