In-depth, depth of field question. Weirdness. PLEASE reply.

Joined
Aug 25, 2009
Messages
4,071
Location
Missouri
OK my peoples, I have some dof weirdness that I need explained.

First, I shoot cars, and normally in doing so I shoot a front 3/4 from the driver's side, and the same front 3/4 from the passenger's side.

I've been doing this for YEARS, and have noticed that when shooting from the passenger's side, even if it's at the same angle (sorry that the below pix are not), the dof is always DEEPER.

So look at these two pictures.

I have circled the areas I want you to pay close attention to.

First, look at the driver's side pic and notice how out of focus the rear of the car is. Now look at the passenger's side pic and notice how in focus the rear is.

Now, it's easy to see that the rear of the car in the passenger's side pic is CLOSER to the camera than the driver's side pic because of the angle, so it should be sharper.

Easy stuff, right?

Now for the hard bits.

Look at the big circles in the background.

Look at the far left, and the far right of each picture, which should be literally the exact same degree out of focus if the dof was constant...but they're not.

Can someone tell me why?

Beyond that, and you're not going to believe this one at all, but if you download the passenger's side image and look at the driver's headlight, it's more out of focus than the rear of the car, and it was absolutely closer to me.

Can someone explain this weirdness to me??
 

Attachments

  • VEP_5055.jpg
    VEP_5055.jpg
    406.5 KB · Views: 43
  • VEP_5058.jpg
    VEP_5058.jpg
    379.5 KB · Views: 39
Joined
Jan 22, 2019
Messages
2,801
Location
Jupiter, FL
Real Name
Andy
It appears that the lower left corner of both frames is a bit more OOF than the lower right. If the pavement surface is flat and level, I would not expect to see any difference in how OOF the left foreground is compared to the right.

The EXIF data indicates that these were made at 135mm and f/2, a lens and aperture which I presume you selected to render the BG somewhat OOF. However the working distance between the camera and car is fairly large - I'm guessing about 100 feet. And the trees/road forming your BG appear to be about that same distance behind the car.

So I think there might be two things going on here:

1. Too much camera to subject distance (and not enough subject to BG distance)
2. A slightly de-centered lens.
 
Joined
Aug 25, 2009
Messages
4,071
Location
Missouri
It appears that the lower left corner of both frames is a bit more OOF than the lower right. If the pavement surface is flat and level, I would not expect to see any difference in how OOF the left foreground is compared to the right.

The EXIF data indicates that these were made at 135mm and f/2, a lens and aperture which I presume you selected to render the BG somewhat OOF. However the working distance between the camera and car is fairly large - I'm guessing about 100 feet. And the trees/road forming your BG appear to be about that same distance behind the car.

So I think there might be two things going on here:

1. Too much camera to subject distance (and not enough subject to BG distance)
2. A slightly de-centered lens.

Yes, the lower left is more oof than the lower right, just the opposite at the far end. The ground was pretty level, it was a parking lot.

Yes, 200/2, and yes, to maximize the oof nature of the background. Far right was a bit further distance, but not nearly enough to warrant the massive difference in DOF.

Even if there was too much camera to subject distance, what does that have to do with the oof nature of the background? They should render the same, no?

De-centered lens. Interesting, never heard of that. Same thing happens with my 200/2. Maybe a de-centered mount on my camera?
 
Joined
Jan 13, 2006
Messages
5,113
Location
Columbia, Maryland
Real Name
Walter Rowe
Are you certain the left was just as far away as the right? The road in the background looks to be further away on the right making me think those trees also might be further away, but maybe its just misinterpretation of what I'm seeing.
 
Joined
Jan 22, 2019
Messages
2,801
Location
Jupiter, FL
Real Name
Andy
I just browsed through your 500px. Firstly, a very nice portfolio! At a glance, I don't see the phenomenon shown with your two Cadillac photos in the gallery images.

Even if there was too much camera to subject distance, what does that have to do with the oof nature of the background? They should render the same, no?
The road in the background slants away so that the right side elements are farther from camera. This would make them more OOF, as they would fall farther beyond the plane of focus.
De-centered lens. Interesting, never heard of that. Same thing happens with my 200/2. Maybe a de-centered mount on my camera?
De-centering can occur to a lens, but typically only if it has been bumped hard or taken a big fall. A misaligned lens mount on your camera could cause all your lenses to point slightly off axis, which might be why you've seen this with more than one lens. Do you have a backup body you can try with the same lens used for the Caddy?
 
Joined
Jan 22, 2019
Messages
2,801
Location
Jupiter, FL
Real Name
Andy
I'm starting to think that you have an overall optical misalignment issue on the camera side. This could mean that either the lens mount (on the camera) and/or the sensor itself is skewed so that the optical axis of your lenses is not precisely perpendicular to the sensor. This would create the effect seen in your Caddy photos. I would advise trying a couple of your lenses on a different body.
 
Joined
Sep 13, 2007
Messages
20,966
Location
Northern VA suburb of Washington, DC
I'm assuming you didn't crop either image. If I'm right about that, the distance between the subject and camera has changed. You even acknowledged that the distance between the two was changed. It's also evident as the height of the car is smaller in the passenger-side view than in the driver-side view. That would explain why the rear of the car in that view is more in focus.

If you did crop one or both images, your comparison isn't valid; you should expect differences barring coincidences that would render the two images the same.

I didn't take the time to evaluate why the more distant objects appear differently.

In the end, the factors of depth of field are always the same and are unalterable laws of physics: distance between subject and camera, focal length and aperture. If you change one or more factor, the resulting photo will also change.
 
Joined
Aug 25, 2009
Messages
4,071
Location
Missouri
Guys.

Look at the passenger side pic.

Look at the little orange and white marker on the far right side.

Now look on the far left side and see the RR sign.

That RR sign was FURTHER away than the little yellow and white marker, by quite a margin.

This makes no DOF sense.
 
Joined
Aug 25, 2009
Messages
4,071
Location
Missouri
I just browsed through your 500px. Firstly, a very nice portfolio! At a glance, I don't see the phenomenon shown with your two Cadillac photos in the gallery images.


The road in the background slants away so that the right side elements are farther from camera. This would make them more OOF, as they would fall farther beyond the plane of focus.

De-centering can occur to a lens, but typically only if it has been bumped hard or taken a big fall. A misaligned lens mount on your camera could cause all your lenses to point slightly off axis, which might be why you've seen this with more than one lens. Do you have a backup body you can try with the same lens used for the Caddy?

Thanks :)

I hide the phenomenon.

I first thought it was the lens mount, but it's not, I have seen this on my D800 + 200/2, and D810 + 135/2.

I think this weekend I'm going to do some testing to see exactly what's going on.
 
Joined
Aug 25, 2009
Messages
4,071
Location
Missouri
I'm assuming you didn't crop either image. If I'm right about that, the distance between the subject and camera has changed. You even acknowledged that the distance between the two was changed. It's also evident as the height of the car is smaller in the passenger-side view than in the driver-side view. That would explain why the rear of the car in that view is more in focus.

If you did crop one or both images, your comparison isn't valid; you should expect differences barring coincidences that would render the two images the same.

I didn't take the time to evaluate why the more distant objects appear differently.

In the end, the factors of depth of field are always the same and are unalterable laws of physics: distance between subject and camera, focal length and aperture. If you change one or more factor, the resulting photo will also change.

No crops.

The distance in front of the subject will not matter whatsoever, it would be the distance behind. And the far right is indeed further away than the far left. However, this is a MINIMAL distance difference, and not nearly enough to warrant such an obvious oof difference. And I've seen it over and over.
 
Joined
Sep 13, 2007
Messages
20,966
Location
Northern VA suburb of Washington, DC
Look at the passenger side pic.

Look at the little orange and white marker on the far right side.

Now look on the far left side and see the RR sign.

That RR sign was FURTHER away than the little yellow and white marker, by quite a margin.

It's clear to me that you're either wrong about the relative distance between those two objects, your lens is defective, or you used a special kind of lens (tilt-shift lens or a LensBaby) that introduced that characteristic. I realize the last possibility is remote at best, but it's still worth mentioning.
 
Joined
Sep 13, 2007
Messages
20,966
Location
Northern VA suburb of Washington, DC
The distance in front of the subject will not matter whatsoever

That distance may not produce discernible differences, especially when viewing the photo at such a relatively small size as displayed here. However, to say that it doesn't matter whatsoever is to ignore the laws of physics. That's a losing game.
 
Joined
Jan 22, 2019
Messages
2,801
Location
Jupiter, FL
Real Name
Andy
I'm reluctant to suggest this because I know how tedious testing lenses can be. However, if this is something that has long been present in your work, it might be useful to perform some simple tests. To really make the effect of DOF obvious, I suggest you set up your test in a very well controlled environment with fairly short working distances (camera to subject and subject to background much less than you would use for your automotive subjects). Make sure you have items that will appear in all four corners of the frame that are distinct enough from their surroundings. The front pair and rear pair of items should fall on planed that are fairly precisely parallel to the camera sensor. I think this is the only way to completely rule out equipment calibration issues. Then the debate can begin regarding angles!
 
Joined
Aug 25, 2009
Messages
4,071
Location
Missouri
It's clear to me that you're either wrong about the relative distance between those two objects, your lens is defective, or you used a special kind of lens (tilt-shift lens or a LensBaby) that introduced that characteristic. I realize the last possibility is remote at best, but it's still worth mentioning.

I'm not wrong about the distance, and even if I were, this is a recurring phenomenon.

The lens is not defective, it has happened on multiple lenses, and multiple bodies.

I did not use a TS lens or a lensbaby, for god's sake.

Forgive me, but you're not helping. I need someone with a better understanding of dof.
 
Joined
Aug 25, 2009
Messages
4,071
Location
Missouri
I'm reluctant to suggest this because I know how tedious testing lenses can be. However, if this is something that has long been present in your work, it might be useful to perform some simple tests. To really make the effect of DOF obvious, I suggest you set up your test in a very well controlled environment with fairly short working distances (camera to subject and subject to background much less than you would use for your automotive subjects). Make sure you have items that will appear in all four corners of the frame that are distinct enough from their surroundings. The front pair and rear pair of items should fall on planed that are fairly precisely parallel to the camera sensor. I think this is the only way to completely rule out equipment calibration issues. Then the debate can begin regarding angles!

This is how it's going to have to be done, I was just hoping that someone had seen the phenomenon before.

The other thing that's weird is that I get mirror slap blur even with my 135/2 at slow speeds. I expect that with the 200/2, but the 135/2 was a surprise for sure.
 

Latest threads

Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Nikon Cafe is a fan site and not associated with Nikon Corporation.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Forum GIFs powered by GIPHY: https://giphy.com/
Copyright © Amin Forums, LLC
Top Bottom