Bad Advice from the Pros (Page 2): You are correct. The 35mm lens shot has deeper DOF. Cropping in post does change DOF but not to an equal degree such as to offset the increased DOF...

Ysarex Loc: St. Louis

BebuLamar wrote:

DOF are different because the image taken with the 35mm then crop becomes a smaller format. Different format has different DOF.

You are correct. The 35mm lens shot has deeper DOF. Cropping in post does change DOF but not to an equal degree such as to offset the increased DOF from the shorter lens when the photos are taken.

Easy to solve -- just do it. Here's the proof. I set up the camera on a tripod. I mounted a 58mm lens and took the photo at f/4. Without moving the camera I switched lenses to a 35mm and took the photo at f/4. I was careful in both cases to focus each lens to the same place in the scene. I then cropped the 35mm image to match the 58mm and enlarged then both to the same degree. You don't even have to download them to see the 35mm lens image has more DOF.

(Download)

selmslie Loc: Fernandina Beach, FL, USA

Ysarex wrote:

... I set up the camera on a tripod. I mounted a 58mm lens and took the photo at f/4. Without moving the camera I switched lenses to a 35mm and took the photo at f/4. I was careful in both cases to focus each lens to the same place in the scene. I then cropped the 35mm image to match the 58mm and enlarged then both to the same degree. You don't even have to download them to see the 35mm lens image has more DOF.

You have a point. I will need to repeat this experiment later today to confirm it.

Nevertheless, it's clear that changing the focal length and cropping accordingly has no effect on perspective.

The problem I can see is that the focus distances are very small. With greater distances the focal length does not measurably change the subject distance.

CHG_CANON Loc: the Windy City

BebuLamar wrote:

He said he crop the 35mm shot to give the FOV of the 50mm in effect changes the the format to approximately APS and thus the DOF is 0.29m which is a less than that of the uncropped image but still more than that of the 50mm shot.

Actual pictures with an appropriate subject are needed to demonstrate. I have more important things this afternoon like washing my hair than to set-up the specific demo. I did try to find similar images in LR for the two lenses on the same body at the same aperture, but didn't find two that I might call both apples.

The same distance and cropping the 35mm to the same-ish image will give a near identical ish image when resized to the same pixel resolution. One really has to have an eye for the details of the background, if any, to show differences. So to generate those differences, you need a subject within the frame and a common aperture.

It seems much easier to artificially change the question and then argue why the answers are wrong. That's the kind of inattention to detail that doesn't belong in the Advice from Pros section.

Ysarex Loc: St. Louis

CHG_CANON wrote:

Actual pictures with an appropriate subject are needed to demonstrate.

Done: https://www.uglyhedgehog.com/t-702446-2.html#12355112

CHG_CANON wrote:

I have more important things this afternoon like w... (show quote)

CHG_CANON Loc: the Windy City

Ysarex wrote:

Done: https://www.uglyhedgehog.com/t-702446-2.html#12355112

My thoughts exactly, thanks! I took the liberty of pulling out the detailed comparison.

Nitpick aside: It would be better the EXIF matched the images. I recognize the lenses are adapted to the mirrorless body, some might not.

(Download)

PHRubin Loc: Nashville TN USA

selmslie wrote:

That's only true if you also make a correspondingly smaller image from the result.

But if the final images are the same size, it doesn't matter whether you cropped in the camera (used a Dx crop on an Fx sensor) or if you did a 2/3 stop in post processing.

For more on this see https://www.uglyhedgehog.com/t-698801-1.html#12280939

But they are made with different focal length lenses which enter into the equation.

Ysarex Loc: St. Louis

selmslie wrote:

You have a point. I will need to repeat this experiment later today to confirm it.

Nevertheless, it's clear that changing the focal length and cropping accordingly has no effect on perspective.

But it does affect DOF, so bad advise from you too.

selmslie wrote:

The problem I can see is that the focus distances are very small. With greater distances the focal length does not measurably change the subject distance.

I've encountered this question decades ago and numerous times. Basically the question is can you shoot with a shorter lens and then crop to gain an increase in DOF? The answer is yes. Although cropping does alter DOF it won't alter DOF enough to completely offset the increase gained from the shorter lens before cropping. I've repeated this test under many circumstances both with film and digital cameras -- same conclusion every time.

tramsey Loc: Texas

selmslie Loc: Fernandina Beach, FL, USA

selmslie wrote:

You have a point. I will need to repeat this experiment later today to confirm it. ....

I can show what I am talking about with this example. To make the math easier I used a 50mm and a 75mm lens and a 1.5 crop factor.

Note that the cropped 50mm version is on a 1.5 crop factor sensor so that it produces an image with the same perspective as a 75mm lens on a full frame sensor.

But the cropped image needs to be enlarged 1.5x as much as the full frame image to get the same relative print dimension.

The consequence is that the hyperfocal distance is different by only 1.1 meters and the DoF at 10 meters is different by only 0.02 meters.

This is where the calculator from Cambridge in Colour comes in handy. It accounts for the different enlargement needed to get a Dx crop image to print at the same size as a full frame image.

On the other hand, if you make a print with a max dimension of 10 inches, the cropped 50mm image will have a greater DoF.

BebuLamar

selmslie wrote:

I can show what I am talking about with this examp... (show quote)

Why you make 1 print 10 inch and the other 15 inch?

Ysarex Loc: St. Louis

selmslie wrote:

I can show what I am talking about with this example. To make the math easier I used a 50mm and a 75mm lens and a 1.5 crop factor.

Note that the cropped 50mm version is on a 1.5 crop factor sensor so that it produces an image with the same perspective as a 75mm lens on a full frame sensor.

The perspective is the same because the camera position (distance to subject) is the same.

selmslie wrote:

But the cropped image needs to be enlarged 1.5x as much as the full frame image to get the same relative print dimension.

No it does not. From the same camera position with the same subject distance, a 75mm lens on a FF camera and a 50mm lens on an APS-C camera will produce the same framing and so enlarge equally to the same print dimension. You're taking the same photo with both cameras.

selmslie Loc: Fernandina Beach, FL, USA

Ysarex wrote:

No it does not. From the same camera position with the same subject distance, a 75mm lens on a FF camera and a 50mm lens on an APS-C camera will produce the same framing and so enlarge equally to the same print dimension. You're taking the same photo with both cameras.

You are right and here is why:

The DoF is developed by first calculating the hyperfocal distance (HD). At subject distances significantly grater than the focal length (50 and 75 mm are only about 2 and 3 inches), the second equation is all we need. What it shows is that a crop factor of 1.5 does not offset the 1.5x change in focal length because f is squared.

Once we have HD we can calculate the near and far limits of the DoF Depth of Field, Part II: The Math. It's easy to see why the DoF becomes infinite once you focus beyond HD.

But if we also increase the print size by 50% the change in the CoC and the change in magnification together offset the change in focal length. That's where I went astray. I didn't need to increase the print size.

Aside from the DoF difference, all of the perspective issues remain identical.

BushDog Loc: San Antonio, TX

Thank you everyone for your responses. 👍🏽

wdross Loc: Castle Rock, Colorado

Ysarex wrote:

I've encountered this question decades ago and numerous times. Basically the question is can you shoot with a shorter lens and then crop to gain an increase in DOF? The answer is yes. Although cropping does alter DOF it won't alter DOF enough to completely offset the increase gained from the shorter lens before cropping. I've repeated this test under many circumstances both with film and digital cameras -- same conclusion every time.

Thank you for the information. I stand corrected. I usually shoot for "filling the canvas" and leave as little for cropping as possible. It has been many, many years since I cropped anything close to a 50mm view out of a 35mm shot. I usually only allow enough margins to allow for 8x10 through 20x24 prints.

selmslie Loc: Fernandina Beach, FL, USA

At the risk of beating a dead horse, there is one other factor that is often overlooked - diffraction limits.

If you use an aperture just below the diffraction limit for full frame (about f/11) you might need to adjust the aperture for a 1.5x crop to about f/8.

In this example, both images are close to the diffraction limit. The DoF is not quite the same because a one stop change in aperture does not change the results by a factor of 1.5, but it's close.

The numbers would be closer if we were looking at a 2x crop factor and a two stop change in aperture. But sensors with a 2x crop factor also have a 4:3 aspect ratio rather than 3:2 so it's difficult to replicate this with a DoF calculator.