How come Gradient ND filters actually work?

hahaha... I guess I didn't keep my picture simple enough, huh? lol!
 
It's easy to see where harz is coming from - there's a lot that is correct in the original question, and there is an understanding of the fundamental properties of light and of optics. If you scaled everything to the size of an eye, and put the ND grad right in front of the eye in the same way that a filter is put right in front of a camera lens, the ND grad would have very little gradient effect for exactly the reasons that harz states. The big difference between a camera lens and an eye is that the eye's pupil (which is the eye's entrance pupil - that's where the name came from) is right next to the eye's front element. For most, but not all, camera lenses the entrance pupil is usually a little way back from the front element, and the entrance pupil is where the lens sees the world from.

Using a simple, single-element lens to illustrate why an ND grad has an effect is a little misleading, because the entrance pupil of a single element lens is usually the lens itself, or very close to the lens. There is, therefore, very little distance between the entrance pupil of a single element lens and an ND grad right in front of the lens, and it is the distance between the entrance pupil and the ND grad that allows the ND grad to have an effect. Understanding what the entrance pupil is and what it does is the key to a few of the properties of lenses, such as aperture, evenness of illumination of the image, perspective, rotation point for stitched panoramas, lens hood design and, as here, the use of ND grads.
 
OP:

1. Get a pair of sunglasses
2. Put sunglasses over half your eye
3. Understand how ND grad filters work

Got it yet??
 
I'm not sure you're getting what the OP was asking, or at least where the paradox might lie. You need to scale your experiment:

1. Get a very tiny pair of sunglasses such that one lens is roughly the same size as your pupil, then take a lens out (you don't want to poke your eye out with the tiny sunglasses frames)
2. Put sunglass lens right up against your eye - a millimetre or less from your cornea - so that it only covers half the pupil
3. Observe the effect

Harz had 'got it' by post #9 I suspect.

I think that the OP asked an intelligent question with a sense of genuine inquiry and some understanding of optics, and deserves an appropriate quality of answer.
 
Last edited:
I'm not sure you're getting what the OP was asking, or at least where the paradox might lie. You need to scale your experiment:

1. Get a very tiny pair of sunglasses such that one lens is roughly the same size as your pupil, then take a lens out (you don't want to poke your eye out with the tiny sunglasses frames)
2. Put sunglass lens right up against your eye - a millimetre or less from your cornea - so that it only covers half the pupil
3. Observe the effect

Harz had 'got it' by post #9 I suspect.

I think that the OP asked an intelligent question with a sense of genuine inquiry and some understanding of optics, and deserves an appropriate quality of answer.

That was a serious answer, believe it or not, and have you actually tried doing it because it works.
 
Hold your hand up to the sun on a bright day. A grad-ND filter works the same way.

Sent from my Galaxy Nexus using Tapatalk
 
I'm not sure you're getting what the OP was asking, or at least where the paradox might lie. You need to scale your experiment:

1. Get a very tiny pair of sunglasses such that one lens is roughly the same size as your pupil, then take a lens out (you don't want to poke your eye out with the tiny sunglasses frames)
2. Put sunglass lens right up against your eye - a millimetre or less from your cornea - so that it only covers half the pupil
3. Observe the effect

Harz had 'got it' by post #9 I suspect.

I think that the OP asked an intelligent question with a sense of genuine inquiry and some understanding of optics, and deserves an appropriate quality of answer.

That was a serious answer, believe it or not, and have you actually tried doing it because it works.

I don't doubt that it was a serious answer, and of course it works. I just didn't think that it really answered the OP's question (which seems to have been answered by post #9 anyway) so that's why I suggested a different version of your experiment, in scale.

Does nobody understand what the OP was getting at? Have you guys read and understood the question? Why are you all treating the OP like an idiot, which is clearly an incorrect assumption? Is there something wrong with the original answer?
 
I don't think light get mix up at any point if it got mix up at any point before it hit the sensor there would be weird photo. How would the light un-mix after being mix up?

Even if it meet at one small point to be inverted by the lens I don't thing it actually mix. By mix I mean every single ray of light mix to one single ray point.
 
I'm not sure you're getting what the OP was asking, or at least where the paradox might lie. You need to scale your experiment:

Exactly.

Harz had 'got it' by post #9 I suspect.

Sadly, I still don't get WHY graduated filters work. I've been googling it and can't find out, and I don't need a more verbose re-iteration of what apparently happens - I suspect I am going to contact the Nikkor engineers over this.

Does nobody understand what the OP was getting at? Have you guys read and understood the question? Why are you all treating the OP like an idiot, which is clearly an incorrect assumption? Is there something wrong with the original answer?

I think treating my like an idiot started with some sarcasm somewhere early on. Regardless in terms of getting the picture, #2 had it right ....
 
Sadly, I still don't get WHY graduated filters work. I've been googling it and can't find out, and I don't need a more verbose re-iteration of what apparently happens - I suspect I am going to contact the Nikkor engineers over this.

That's disappointing. I hoped that I had explained why they work, why the light from a point that will later form the image of that point is not distrubuted evenly across the grad when the grad is not at the entrance pupil. What part of my explanation didn't you get? Do you understand the concept of the entrance pupil, and did you do a practical experiment to see how the apparent location of the entrance pupil moves across the lens' front element as you move your eye away from the lens axis?

I suspect that the key is to understand the concept of the entrance pupil. Once you've got that clear, everything else is obvious. If you are going to talk to the guys at Nikon then it would be a good idea to have done your homework first, so they don't have to explain the basics to you - so you don't come across as an idiot to them as well. Start your homework by understanding the entrance pupil. (By the way, I find it easier to talk to the guys at Zeiss, and they will respect you more if you show them that you already know your stuff.)

If I have time I'll draw a diagram for you.
 
This post has started giving me an inquisitive mind. Yesterday when I got into my car and put it into reverse to back out of the garage I started trying to visualize everything that is going on in the transmission. I know that putting the shift selector into R does exactly what I want the transmission to do but I can't grasp how it does what it does. Would someone please explain and draw a diagram.

Don't forget the old saying "curiosity killed the cat". I don't have a full understanding of electricity but I know when I flip a wall switch the ceiling light comes on. I'm just failing to see why trying to understand the intricate details of how this works is so intriguing.

Jerry
 
I do understand the "pupil", and have even taken an old lens apart once, and I have looked through my lenses off-camera! No need to draw diagrams either.

Some of this light is stopped by the "iris" / aperture blades - that is understood: in particular this affects light hitting the outer region of the lens, but more importantly the iris affects the top and bottom half in the same way, by design (or at least it attempts to, which is why the number of aperture blades are important).

For simplicity lets assume a static iris at some arbitrary setting, or a smaller lens with the same effective light gathering potential. If you still want to fixate on the "pupil", please assume a wide open aperture for the rest of this discussion.

All the explanations, including yours, ignore the crux of my question. Why doesn't the light from any point in either half the scene get affected simultaneously by BOTH halves of the filter.

All the light from one point (A) in the image that hits the lens anywhere is focused back to a single spot on the image (Assuming of course that point A is within the area that is in focus well enough so that the CoC is small enough to be negligible, known as the in-focus area and based on the effective Depth of Field.

Light reflecting from this point A spreads out, and some of it falls on the lens within the area that is not blocked by the iris. Half of this light is affected by the top half of the graduated ND filter, and the other half by the bottom, and just to be pedantic, we assume a graduated filter with a sharp dividing line, or else adjust the "halve" statement to read something like "equal portions of the light bla bla bla"

The point is that all of this light is bent by the lens to focus on a single point in the resulting image, irrespective of whether it passed through the top or bottom half of the filter.

Conversely the filter is so far out of focus as to have no effect.

But clearly this is wrong.

I have also observed that very out of focus objects (fingers, fence wire right against the lens, camera straps) are still visible. And I guess this is what the graduated filter is doing - it is being vaguely in focus, and therefore functional!
 
All the explanations, including yours, ignore the crux of my question. Why doesn't the light from any point in either half the scene get affected simultaneously by BOTH halves of the filter.

Because the light that forms the image of a point doesn't pass through both halves of the filter, unless the filter is at the iris (when it is at the entrance pupil), as I've already explained. The filter is not at the entrance pupil in almost all cases. It's really simple. Why don't you understand that?

Don't forget the old saying "curiosity killed the cat". I don't have a full understanding of electricity but I know when I flip a wall switch the ceiling light comes on. I'm just failing to see why trying to understand the intricate details of how this works is so intriguing.

Jerry

I guess that by your own reasoning you are happy to remain in a state of not knowing why this is so intriguing for the OP.
 
Last edited:

Most reactions

Back
Top