Is DOF like a rectangle or a cone...our picture seems weird

W

walnutcrunch

Well-known member
So I had a previous thread about some distortions in the picture coming out of our 500. The centre of the frame would be in focus, and the frame would lose focus as you moved out to the edges.

I always believed that there was a focal plane, kind of like a rectangular box that moved back and forth as you changed focus. Just like you see in After Effects 3d camera.

I thought that as you racked focus, a focus plane, perpendicular to your lens would move back and forth and focus or lack of focus would be consistent across the frame.

Our lens/camera does not work like that as you rack the focus the center of the frame comes into focus first, and then the sides of the frame come in to focus.

Instead of a rectangular box, it seems like some kind of cone of focus extends from the lens.

Is this normal. The engineers looked at a our camera, saw what they thought was a problem. Panasonic took our camera, said they couldn't see a problem, reloaded lens data, sent it back and said it is fine.

The camera picture looks slightly different, but has the same effect on a high end canon lens as our cac lens.

Is this normal? We're using a test sheet filled with letters, completely flat and square to the camera. As we bring it into focus, a circle in the center is in focus first, then the edges come into focus, then everything is in focus.

Is that how a lens/camera normally performs?
 
it's flat. if you are focused on a flat brick wall and the camera is perfectly level and perpendicular to the wall, the entire frame from edge to edge should be in focus. if it's sharp in the middle but not on the sides, etc. then you have issues.
 
and its probably the lens. Try renting the Canon 13x wide Angle for a day. Its an amazing lens and has great edge to edge. Ive seen many ENG lenses that have a super sharp center then fall off and get soft as you move away from the center of the image.
 
maybe the individual elements of your lens are not aligned with each other properly. anything get dropped or exposed to extreme temps?
 
You can calculate via the pythagorean theorem the distance from the lens to the target at a non-central point p and compare that distance to the central distance c. Then the question is whether the critical focus point is a sphere or a plane for that particular lens. Unless it is a copy lens, it's more likely that it's a sphere. Then, is the distance delta within the theoretical focal distance for the defined circle of confusion you've determined is acceptable for the lens at that focal length? In any case, what you have described goes well beyond what you might expect in focus falloff shooting a flat target.

In the case of the particular issue you mention, in the previous threads on this I thought that it was determined that setting the camera to its default settings resolved the situation, which makes it sound like there was some image manipulation. Could it somehow be related to the CAC implementation?
 
The image settings seems to have been a false result. All in all it is pretty confusing. We've had the local Panny approved engineer look at it, and Pansonic Canada has looked at it. The local guy could see the problem, Panasonic could not. Information flow has not been stellar.

J Michael, your serious math looks like it makes sense, but man, I have no idea how to do that. Or even figure out the significance of the numbers if I could.

If we focus on a flat object (like we are) we can get the whole image pretty sharp corner to corner, it's when we rack the focus slightly that the center stays in focus and the edges go out. Eventually the full frame will go out of focus.

The problem is, that with focus assist, and focus in red, there is absolutely no way to tell when you have center focus or full frame focus. It's quite perplexing.

And considering we are in conditions that don't often let us take our monitor we rely heavily on focus assist.

I attached a file of a screengrab, and here's the weird part, this is with a loaner 500, and a high end canon lens. Same effect as on our camera, same effect as the guy had with the 3000.

I can't find the variable that makes any of this make sense, other than "that's what all cameras do you lunk". But people are telling me all cameras don't do this. So why is it happening?
 

Attachments

  • camera test oct 30.jpg
    camera test oct 30.jpg
    202 KB · Views: 0
I used to shoot 4X5 copy negs years ago, and we had a special lens for flat art. Very sharp, edge to edge, even fully opened. I also have a Nikkor 55mm macro lens for close-ups and flat art. All of my Nikkor zooms are a bit softer on the edge fully opened. Alot better stopped down a couple.

I have never seen a video camera zoom lens sharp edge to edge on flat art with the iris fully opened. A couple of stops down, it's better. At f16 is gets softer.

What f stop are you using?

Go to a dealer,etc. and try different lenses. Try wide open and then at f5.6.

JohnnyD
 
Last edited:
Focal length is a factor too. So, what f-stop, focal length, what is distance from lens to target, and what's the size of the target? If you are sharp corner-to-corner you should expect the corners to fall off first as the focus is pulled, since the corners are at the greatest distance relative to the center of the target.
 
Last edited:
You must log in or register to reply here.
Back
Top