I was reading Edmund Optics application notes on relative illumination, which talks about what it is, its affects, its drawbacks, positive aspects and so on. However, out of all the types of vignetting, the one that stumped me was Optical vignetting. I especially cant seem to visualize how it ties in with the stopping up or down the aperture, and consequently the effect on the Entrance pupil and bokeh? (I also believe it relates to coma aberration or are they completely different)

I found the page, What is vignetting? at photographylife.com, which, to my understanding, explained that the reason vignetting decreases when stopping down the aperture, is due to the fact that the relative difference in the amount of rays (illumination) coming from the off axis points compared to the more centered points decreases. From the page,

Optical vignetting naturally occurs in all lenses. Depending on the optical design and construction of the lens, it can be quite strong on some lenses, while being barely noticeable on others. Still, vignetting occurs on most modern lenses, especially on prime / fixed lenses with very large apertures. There are two causes for this. First, at the widest apertures, the light than enters the lens is partially blocked by the lens barrel, as indicated by the below diagram:

Optical vignetting diagram, from photographylife
Diagram of optical vignetting, by Nasim Mansurov, from "What is vignetting?" at Photographylife.com.

Due to the length of the lens barrel and its relative size of the front and back frames, peripheral light rays that travel at extreme angles are partially blocked. As a result, the light that reaches the image plane at such angles naturally falls off (decreases in brightness) towards the extreme corners of the frame.

Note that such vignetting is mostly evident at large apertures, since it is the physical lens barrel that mostly blocks the peripheral light from the front and back of the lens barrel. Once stopped down, the smaller size of the aperture in the center is visible even from the corners, allowing the light to pass through. That’s why most fast aperture prime lenses have plenty of vignetting at the widest aperture and dramatically improve as the aperture is stopped down.

Pay a close attention to the entrance pupil in the above example. As you can see, it is circular in the center, but takes a different shape that some call “cat’s eye” in the corners. If you have a fast aperture prime lens, you might have seen this effect on the lens bokeh – the bokeh shapes stay circular in the center, but gradually change in shape towards the corners, just like you see in the below crops ...

I think this is right, but I cannot seem to grasp it visually, because what is also confusing me in the diagram, is why the rays from the off axis source don't bend? Also when I view the diagram I am imagining the aperture blades closing down reducing the ratio between light from center axis vs off axis, but eventually stopping down also affects the off axis light rays and seeing as its already smaller, in relative terms more than the on axis rays?

Edmund Optics, in Relative Illumination, Roll-Off, and Vignetting, also mentioned that it is harder to focus the light (objects) onto the periphery of the image sensor, why is this the case?

Could someone please help explain and answer some of the questions I have above?

  • 1
    \$\begingroup\$ You might find this article archived at The Wayback Machine helpful. \$\endgroup\$
    – Michael C
    Dec 10, 2023 at 7:16
  • \$\begingroup\$ @MichaelC Funnily enough I just came across that paper yesterday and was talking about it in the comments on Steven Kersting's answer, but it seems I forgot to link the article again. Thank you for suggesting and linking it here. Its a great and intuitive resource. \$\endgroup\$
    – vannira
    Dec 10, 2023 at 19:30
  • 1
    \$\begingroup\$ @vannira I'm struggling to understand your questions. This post is unfocused; it seems your question re: what was said at Edmund Optics is unrelated specifically to vignetting as the rest of the question pertains. I think I've suggested it before, but I suggest again: to clarify your question to help us answer it, try to write your title into a single, concise complete sentence. A title of "Optical Vignetting" is a subject, not a question. If you can make it into a complete question sentence, maybe we can help you better. \$\endgroup\$
    – scottbb
    Dec 10, 2023 at 22:52
  • \$\begingroup\$ @scottbb yeah I think it’s perhaps a bad habit of mine, I just have a tendency to fit all the questions into one subject, rather than multiple questions, although the latter may be better for a “question and answer” forum. \$\endgroup\$
    – vannira
    Dec 14, 2023 at 11:29
  • \$\begingroup\$ "I just have a tendency to fit all the questions into one subject, rather than multiple questions". Stop doing that, please. It reduces the quality of your questions here, and makes answers less focused, more difficult to read, and leads to lots of comments requesting or adding clarifications for each post. Thank you. \$\endgroup\$
    – scottbb
    Dec 14, 2023 at 20:28

2 Answers 2


The vignette:

If you were an insect walking about on film or image sensor and looking up at the camera lens during the exposure, you would see the aperture as an illuminated circle if on center (axis). Walking towards the corner of the sensitive surface, you would see not a circle but an ellipse.

The square millimeter aera of the working aperture is a key ingredient to image brightness. Image brightness falls off towards the edges. This phenomenon is called cosign error. Image brilliance fall off is mainly due to the fact that a circle is more compact and then its elliptical counterpart.

Mathematically we compute the image brilliance fall-of using the formula cos^4 angle.

0 angle (axis) = 1

10° off axis = 0.96

20° off axis = 0.78

30° off axis =0.56 (nearly a 50% fall-off)

Think about shining a flashlight head-on, the light creates a circular spot. Aim the flashlight at some angle other than “normal” and the spot of light becomes an eclipse making the spot of light weaker.

Let me add – in the analog negative / positive system, positive prints of negatives are produced by projection printing. The enlarger lens also displays a vignette. The vignette in the camera lens is counterbalanced by the vignette of the enlarger lens when printing negatives.

When making prints from slides, unfortunately the resulting print displays a 2X vignette because positive to positive projection printing exhibits an additive vignette.

  • \$\begingroup\$ Thank you for the great response, the flashlight analogy was very intuitive. Now regarding you latter point, does that mean that the printing mechanism also involves a form of aperture or why would it also produce a vignette of the same fashion? \$\endgroup\$
    – vannira
    Dec 14, 2023 at 11:27
  • \$\begingroup\$ The vignette plagues every lens. We are talking camera lens, telescope lens, microscope lens, projector lens, still camera, movie camera -- every lens! \$\endgroup\$ Dec 14, 2023 at 15:00
  • \$\begingroup\$ I guess because I am not well versed with film printing, I wasn’t aware that they also employed a lens. Thank you for the clarification. \$\endgroup\$
    – vannira
    Dec 15, 2023 at 17:31

I don't fully understand all of the questions, but I'll answer what I can.

Rays from off center do bend; the issue with the drawing is that it is including only selected rays. I.e. all points on the objective lens receives light from all points within the scene; and bends those rays to focus in the correct location. I added another ray to show this.

enter image description here

The entrance pupil is the size of the aperture as magnified by the objective element. As such, it cannot be larger than the element itself, but the objective element can be as large, or even larger than the lens barrel. When the size of the entrance pupil is equal to the size of the lens barrel you get vignetting because the images at the periphery include the edge of the lens barrel (hood/etc) within their field of view.

I.e. all points on an objective element includes light from all parts of the scene required to make a complete image as scene from their perspective. I.e. larger apertures stacks an increasing number of virtual images. And apertures that include the edge of the objective element are more likely to include the lens barrel within their field of view.

  • \$\begingroup\$ Thank you kindly for your response, although I don't quite understand the part "larger aperture stacks an increasing number of virtual images". I also just found this paper, which has a brilliant photo showing how stopping down the lens improves optical vignetting. It elegantly showed what I already thought was the reason behind it, but wasn't quite sure about. \$\endgroup\$
    – vannira
    Dec 9, 2023 at 21:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.