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

Joined
Jan 22, 2019
Messages
3,393
Location
Jupiter, FL
Real Name
Andy
However, I'm a fan of simplicity.
Then test in a simpler way.
I feel like I've been extremely thorough with my testing.
I'm not trying to have "the last word," nor am I trying to antagonize you. Quite the opposite - you brought a legitimate and interesting question regarding DOF to the cafe in a very well-articulated way, and I want you to learn the answer to it. Many have offered constructive advice. You have nothing to lose and everything to gain by following a little more of it.

Best of luck - either in learning the answer, or being content in not doing so.
 
Joined
Jan 13, 2006
Messages
5,224
Location
Columbia, Maryland
Real Name
Walter Rowe
I've been pondering this question and I believe I have the answer. I won't try to explain with words - I think that has been done plenty. Instead, here are two diagrams for consideration.

Constants:
  1. A car, 16 feet in length is parked perpendicular to a treeline another 16 feet behind the car
  2. A FF camera, 135mm f/2 lens shot wide open is used
  3. A horizontal FOV of 15.15 degrees
  4. Focal plane is 80' from the camera
  5. DOF near limit is 74'
  6. DOF far limit is 87'
  7. All of the above are drawn to scale precisely
  8. The D850 camera body is represented at a larger scale for clarity
Driver side view, offset 35 degrees
View attachment 1689723

Passenger side view, offset 55 degrees
View attachment 1689722

For me, this completely solves the mystery. Other thoughts?
I think your assumptions about the setup are incorrect. Make the camera parallel with the trees and only change the orientation of the car 35º off axis, once rotated left and once rotated right. Also, if you look back at one of my illustrations you can see ground right next to the car on both sides where the left side is in focus and the right side is OOF, both clearly at the same distance from the camera. It has the appearance of a tilt-shift effect where the focal plane is not 100% parallel to the sensor due to something in the optical path not being aligned properly.

See this reply: https://www.nikoncafe.com/threads/i...on-weirdness-please-reply.329068/post-4175235
 
Joined
Aug 25, 2009
Messages
4,106
Location
Missouri
Then test in a simpler way.

I'm not trying to have "the last word," nor am I trying to antagonize you. Quite the opposite - you brought a legitimate and interesting question regarding DOF to the cafe in a very well-articulated way, and I want you to learn the answer to it. Many have offered constructive advice. You have nothing to lose and everything to gain by following a little more of it.

Best of luck - either in learning the answer, or being content in not doing so.

I don't know how else to say it other than I feel like the test should be relevant to what I'm actually using my equipment for, otherwise there is no point in learning the answer. You may disagree, and I understand that.

What's happening here is simple; people (like me) are not understanding what they are seeing, so they're offering advice that literally will get me no further down the road of understanding. Not sure how many times I've had to state that I have seen this effect on the D700 + 200/2, D800 + 200/2 (although the last test didn't show it at all), D800 + 135/2, D810 + 200/2, and D810 + 135/2. This literally cannot be equipment related, end of that story, seriously.

And as far as my testing methods, I'm sure they are off by a few inches, but you can clearly see by the overlays that they are super close, MUCH closer than what the dof would suggest.

I just don't see how it can be done better than this, so I've contacted a person that works in the industry. She is going to help me look into it further, or maybe find someone at Nikon that might be able to help me.

And as always, I appreciate your thoughtful responses.
 
Joined
Jan 13, 2006
Messages
5,224
Location
Columbia, Maryland
Real Name
Walter Rowe
I don't know how else to say it other than I feel like the test should be relevant to what I'm actually using my equipment for, otherwise there is no point in learning the answer. You may disagree, and I understand that.

What's happening here is simple; people (like me) are not understanding what they are seeing, so they're offering advice that literally will get me no further down the road of understanding. Not sure how many times I've had to state that I have seen this effect on the D700 + 200/2, D800 + 200/2 (although the last test didn't show it at all), D800 + 135/2, D810 + 200/2, and D810 + 135/2. This literally cannot be equipment related, end of that story, seriously.

And as far as my testing methods, I'm sure they are off by a few inches, but you can clearly see by the overlays that they are super close, MUCH closer than what the dof would suggest.

I just don't see how it can be done better than this, so I've contacted a person that works in the industry. She is going to help me look into it further, or maybe find someone at Nikon that might be able to help me.

And as always, I appreciate your thoughtful responses.
This can be tested very simply. Find a large, empty, flat parking lot with well painted lines. Position yourself on one end such that all the spaces are parallel to the camera and you are looking right down one line of spaces that face each other. Focus partway down the aisle of spaces with a similar aperture as you use for your car shots.

Given the example images you have shared, I anticipate there should be more lines in focus (more DOF) on the left side of the frame, and fewer lines in focus (less DOF) on the right. In addition, the point where both sides are in focus may not be in a plane parallel to camera, indicating as I've stated several times that there is a tilt-shift effect going on.

This image found on the internet illustrates the orientation of the test I suggest.

1632665178061.png
Subscribe to see EXIF info for this image (if available)


I maintain my argument that this is an equipment issue. Accurately functioning equipment would not exhibit this phenomenon. The car has NOTHING to do with this. In my opinion it only confuses the discussion and should be excluded from future sample images.
 
Last edited:
Joined
Aug 25, 2009
Messages
4,106
Location
Missouri
This can be tested very simply. Find a large, empty, flat parking lot with well painted lines. Position yourself on one end such that all the spaces are parallel to the camera and you are looking right down one line of spaces that face each other. Focus partway down the aisle of spaces with a similar aperture as you use for your car shots.

I maintain my argument that this is an equipment issue. Accurately functioning equipment would not exhibit this phenomenon. The car has NOTHING to do with this. In my opinion it only confuses the discussion and should be excluded from future sample images.

What you're saying is that if the phenomenon doesn't exist with just lines then the car...is...causing...it?

And the car DOES have something to contribute; it's an angled surface to show the exact progression of DOF difference between the right and the left.
 
Joined
Jan 13, 2006
Messages
5,224
Location
Columbia, Maryland
Real Name
Walter Rowe
What you're saying is that if the phenomenon doesn't exist with just lines then the car...is...causing...it?

And the car DOES have something to contribute; it's an angled surface to show the exact progression of DOF difference between the right and the left.
I'm saying it will exist. My annotations to his images in my earlier posts illustrate the phenomenon in areas unrelated to the car.

Take this image. The near side on the RIGHT, just below the red circle is in focus. That same area bottom left is NOT in focus. So the plane of focus is actually NOT parallel to the sensor. In addition, the DOF on the right is MUCH shallower than on the left. It can't be any more clear than this. You could remove the car entirely from this picture and this would still be the exhibited phenomenon.

1632667361598.png
Subscribe to see EXIF info for this image (if available)
 
Joined
Aug 25, 2009
Messages
4,106
Location
Missouri
Take this image. The near side on the RIGHT, just below the red circle is in focus. That same area bottom left is NOT in focus. So the plane of focus is actually NOT parallel to the sensor.

That's not actually a good picture to use because the car was parked on a gravel driveway that wasn't level and there were a lot of different elements.

That front shot is a requirement on my shoots, so I've done it hundreds of times. I use several markers to get it lined up as closely as possible, but always shoot an F4 as well as the F2 shot because inevitably one of the headlights will be oof (now that's shallow).

This is why I'm saying that lining up in a parking lot with just lines won't really work, as there is no possible way for me to get exactly 90° from the lines, and even if I could, at F2 it would take just millimeters to cause a dof discrepancy. Doing the front 3/4's looking down the sides of the car and the far ends seems to be a much better way to test IMO.

I keep going back to this image.

The tiny marker on the right is more OOF than the telephone pole on the left, yet the pole is further away. Everyone is saying it's an equipment issue, and if it is, it's with the D700 + 200/2, D800 + 200/2, D800 + 135/2, D810 + 200/2, and D810 + 135/2.

What I wouldn't give right now to have someone in my town with a 5DMKII and a 135L...
 

Attachments

  • VEP_5058.jpg
    VEP_5058.jpg
    379.5 KB · Views: 7
Joined
Jan 22, 2019
Messages
3,393
Location
Jupiter, FL
Real Name
Andy
I’d like to say thanks to the OP and everyone who tried to assist him. Within this thread I have learned something completely new (Brenizer method) and improved my existing understanding of how a shallow DOF is rendered with varying angles of FG/subject/BG.

I only wish that the OP has gained as much in the process. Best of luck to everyone who continues in that particular effort.
 
Joined
Aug 25, 2009
Messages
4,106
Location
Missouri
So for grins I went out tonight and tried my D800 + 85/1.4, and what do you know, it's still there at 85mm. I also did an overlay to show just how close the angles/distances were from right to left. Yes, this is a cropped image, and I did some straightening, but nothing else. Even the TERRIBLE CA is left (my apologies). And once again, the 2nd image (passenger's side) has been flipped for easier comparison. The last image is the overlay.
 

Attachments

  • VES_5369.jpg
    VES_5369.jpg
    218.7 KB · Views: 7
  • VES_5371.jpg
    VES_5371.jpg
    248.9 KB · Views: 10
  • VES_5371 overlay.jpg
    VES_5371 overlay.jpg
    210.4 KB · Views: 9
Joined
Jan 13, 2006
Messages
5,224
Location
Columbia, Maryland
Real Name
Walter Rowe
That's not actually a good picture to use because the car was parked on a gravel driveway that wasn't level and there were a lot of different elements.

That front shot is a requirement on my shoots, so I've done it hundreds of times. I use several markers to get it lined up as closely as possible, but always shoot an F4 as well as the F2 shot because inevitably one of the headlights will be oof (now that's shallow).

This is why I'm saying that lining up in a parking lot with just lines won't really work, as there is no possible way for me to get exactly 90° from the lines, and even if I could, at F2 it would take just millimeters to cause a dof discrepancy. Doing the front 3/4's looking down the sides of the car and the far ends seems to be a much better way to test IMO.

I keep going back to this image.

The tiny marker on the right is more OOF than the telephone pole on the left, yet the pole is further away. Everyone is saying it's an equipment issue, and if it is, it's with the D700 + 200/2, D800 + 200/2, D800 + 135/2, D810 + 200/2, and D810 + 135/2.

What I wouldn't give right now to have someone in my town with a 5DMKII and a 135L...
Perhaps a parking lot with far spaced lines isn’t the best setup. You could create the setup yourself. At that working distance and f/2, the DOF isn’t millimeters, but yes it could be several inches. Placing two rows of broom handles, equally spaced, one row on each side of the image, and insuring each pair is equidistance and parallel to the camera would give you the “calibration field” needed similar to the parking lot. Given what I saw in your sample image, the difference in DOF between left and right sides of the frame was so significant that I believe the parking lot might even provide what’s needed in terms of “proof”.

I still think the image I annotated was an excellent example. Not because of my “brilliant annotations” (LOL), but because it exhibits the issue quite clearly. You stated the problem and I think it is most evident and easily discernible in that image.

I wish you well in your pursuit and look forward to hearing what you discover. I have nothing else I can offer at this point.
 
Joined
Aug 25, 2009
Messages
4,106
Location
Missouri
Perhaps a parking lot with far spaced lines isn’t the best setup. You could create the setup yourself. At that working distance and f/2, the DOF isn’t millimeters, but yes it could be several inches. Placing two rows of broom handles, equally spaced, one row on each side of the image, and insuring each pair is equidistance and parallel to the camera would give you the “calibration field” needed similar to the parking lot. Given what I saw in your sample image, the difference in DOF between left and right sides of the frame was so significant that I believe the parking lot might even provide what’s needed in terms of “proof”.

I still think the image I annotated was an excellent example. Not because of my “brilliant annotations” (LOL), but because it exhibits the issue quite clearly. You stated the problem and I think it is most evident and easily discernible in that image.

I wish you well in your pursuit and look forward to hearing what you discover. I have nothing else I can offer at this point.

I did this shot last night while testing my 85, but just as I suspected, it doesn't help at all. If anything, the dof looks slightly more shallow on the left than the right, but just barely, and I'm sure that's because I was off in my relationship to the lines, or the lines themselves are off.
 

Attachments

  • VES_5373.jpg
    VES_5373.jpg
    203.6 KB · Views: 13
Joined
Jan 13, 2006
Messages
5,224
Location
Columbia, Maryland
Real Name
Walter Rowe
I did this shot last night while testing my 85, but just as I suspected, it doesn't help at all. If anything, the dof looks slightly more shallow on the left than the right, but just barely, and I'm sure that's because I was off in my relationship to the lines, or the lines themselves are off.
Taking into account the potential for uneven surface in the pavement I think this looks pretty much as I would expect and thus for me demonstrates the validity of the test method.
 
Joined
Aug 25, 2009
Messages
4,106
Location
Missouri
Taking into account the potential for uneven surface in the pavement I think this looks pretty much as I would expect and thus for me demonstrates the validity of the test method.
Sure, but does it help me in my quest for figuring out why the dof is off when I shoot a car right and left? See my point? It's not an equipment failure, it's an optics issue, and all of us photographers are just either scratching our heads, or making **** up to test that's meaningless (me included). I have a friend that shoots large format and he's calling it yaw, but I'm not buying that just yet.
 
Joined
Jan 13, 2006
Messages
5,224
Location
Columbia, Maryland
Real Name
Walter Rowe
Sure, but does it help me in my quest for figuring out why the dof is off when I shoot a car right and left? See my point? It's not an equipment failure, it's an optics issue, and all of us photographers are just either scratching our heads, or making **** up to test that's meaningless (me included). I have a friend that shoots large format and he's calling it yaw, but I'm not buying that just yet.
What it tells you is .. your 85mm lens does not exhibit the issue, and the body it is mounted on also does not exhibit the issue.

Is that body one of bodies used where images do exhibit the issue? If so, and that body is ruled out as the issue, then it MUST be one or both of the lenses (135/2 and 200/2) that are the issue. Next try the other body with the 85mm.

If you get acceptable results with the 85mm on both bodies you used for the 135/2 and 200/2 shots that did exhibit the issue, then those two lenses are even more suspect as being the source of the issue.

If camera A + 85mm does not exhibit the problem, then camera A and 85mm are good.
If camera B + 85mm does not exhibit the problem, then camera B is also good.

If camera A and camera B are used with the 135 and the 200, and cameras A & B are good, then the 135 AND 200 both are suspect.
 
Last edited:
Joined
Aug 25, 2009
Messages
4,106
Location
Missouri
What it tells you is .. your 85mm lens does not exhibit the issue, and the body it is mounted on also does not exhibit the issue.
No, it doesn't. Did you read my post above that?? The issue exists with the D800 + 85 as well! And I'm not even going to bother putting it on the D810, I'm sure it'll be there. And I've included the overlay to show just how close the angle was to perfect from side to side.
 

Attachments

  • VES_5369.jpg
    VES_5369.jpg
    218.7 KB · Views: 9
  • VES_5371.jpg
    VES_5371.jpg
    248.9 KB · Views: 11
  • VES_5371 overlay.jpg
    VES_5371 overlay.jpg
    210.4 KB · Views: 10
Joined
Jan 22, 2019
Messages
3,393
Location
Jupiter, FL
Real Name
Andy
I did this shot last night while testing my 85, but just as I suspected, it doesn't help at all. If anything, the dof looks slightly more shallow on the left than the right, but just barely, and I'm sure that's because I was off in my relationship to the lines, or the lines themselves are off.
Simple test = understandable results. At last.
 
Joined
Jan 13, 2006
Messages
5,224
Location
Columbia, Maryland
Real Name
Walter Rowe
No, it doesn't. Did you read my post above that?? The issue exists with the D800 + 85 as well! And I'm not even going to bother putting it on the D810, I'm sure it'll be there. And I've included the overlay to show just how close the angle was to perfect from side to side.
I think we will have to agree to disagree. The parking lot shot with the 85mm to me had a clear horizontal plane of focus that had fairly even DOF on both sides for a single frame as one would hope it to have. That is enough evidence to me that that camera + lens combination are free if any issues. If that same image were taken with the same lens and the other camera body and it looked the same, then that lens and both bodies are free of issues. That only leaves the lenses as suspect. This has to be measured from a single frame. I do not understand how the "overlay" illustration contributes to this.

The space between the green lines looks to me like your DOF region and it seems to run pretty well across the frame parallel to the camera. It cannot be any more clear to me that this camera + lens combination is clean.

VES_5373.jpeg
Subscribe to see EXIF info for this image (if available)


Personally, I would send both bodies and both lenses to Nikon and have them inspected and calibrated. You might be surprised what Nikon reports back about them.
 
Last edited:

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