Shot with the Zeiss 35 2.8 on my A7iii. Any input is much appreciated!

Uncropped Photos Also, here is some other shots I did, with some other settings too.

enter image description here

enter image description here

  • 2
    Do you have any filters on your lens? – xiota Apr 18 '19 at 2:46
  • Not sure about the lines...but iso2500 in midday? – OnBreak. Apr 18 '19 at 2:52
  • 1
    Are the lines in all three images aligned (if you rotate all images into the same orientation)? Does it occur in other images – mattdm Apr 18 '19 at 4:23
  • 1
    Is there a lens hood on your lens? – xiota Apr 18 '19 at 5:43
  • 1
    @Hueco yeah, I know it's high, but it seems like the higher iso-settings are revealing the curves more. – Endless Videography Apr 18 '19 at 17:17

I do not have the same lens you do, but I have a different 35/2.8 lens that appears to exhibit similar behavior when light is shining into the lens. If you can get the arcs to come and go depending on the position of a lamp shining into the lens, then the problem is likely flare.

Here is a sample photo taken with a lamp shining toward the camera from the right. Moving the lamp shows different portions of the arc on the left.

without hood

Using a lens hood may reduce the problem. Here is a photo taken with lens hood, at the approximately the same position as the above photo.

with hood

Similar appearances are known to have been caused by filters.

filter reflection

Some combinations of lenses, teleconverters, and sensors are known to produce flare-like artifacts, such as hot spots and blue dots. This is a possible scenario if using the same lens with different equipment produces different artifacts.


This is a guess rather than a definitive answer. However there it points to some experiments you could do which would indicate whether it is correct or not.

First of all all of your examples show the lines on a clear blue sky. Secondly I think that, if you extended the curves to circles, the centres of those circles would perhaps appear to be at the opposite point to where the Sun is in the sky. Finally the patterns don't appear to be aperture-dependent, and in particular they don't appear only when the lens is wide-open or nearly wide-open.

So my guess is that what you are seeing is some kind of artifact due to polarization of the sky. Light from the sky comes from Rayleigh scattering of sunlight, and it is significantly polarized. The degree of polarization depends on the Sun, but it is greatest at 90 degrees to it, dropping to zero both where the Sun is and opposite it. This means that if you measure the polarisation of the sky you will see arcs of circles both surrounding the Sun and surrounding the point opposite it.

So I hypothesize that something in the camera's optical system – either the lens or the sensor – is sensitive to polarization, and what you are seeing is some artifact due to that.

As I said above: this is a hypothesis not a definitive answer (but it was way too long for a comment). There are two obvious experiments you could do to see if this is true:

  • check whether the centres of the arcs really are opposite the position of the Sun (this means they will be below the horizon of course);
  • do some experiments with a polarizing filter to see what that does to the artifacts.

Finally it's also possible that the lens is seeing something else in the sky: these patterns kind of look like rainbows (which also centre on the point opposite the Sun): the reason I'm suggesting polarization is that this is something you would not see with your eyes. If they are some brightness/colour change, then you should be able to see it with your eyes (although it may be faint), and other lenses should see it. Such faint rainbow-like artifacts might occur because of aerosols in the atmosphere I think (aerosols in this context means a suspension of small particles in the atmosphere: fog & clouds are aerosols, but there are many other kinds, including ones as a result of human activity).

One variant of the above suggestion is that the optical system of the camera might be sensitive somewhat into the infrared or ultraviolet, and what you are seeing is indeed a rainbow, but one which has only IR or UV components. I would have to do the maths about this but I suspect that UV-only rainbows might be a possible result of aerosols with very small particle sizes. I'm not sure IR-only ones make sense in the same way. You could test for this by getting a UV filter and seeing if the artifacts go away.

Note that all these suggestions fail if the arcs are not centred on the point opposite the Sun!


Digital image files can be quite large. Such large files consume oodles of storage space. Regularly, software contained in the camera endeavors to lessen the file’s size This technique is known as file compression. JPEG is commonly in use, residing inside the camera (Joint Photo Experts Group).

This software works by casting out repetitive sequences of data, substituting a code that restores the data when needed. Blue sky is an example of a mundane area where adjacent pixels share like values. This is an example of an artifact due to the failure of the algorithm. Photographers often prefer cameras with more sophistication. These are cameras capable of storing files as RAW i.e. no compression utilized.

  • 3
    Those don't look like compression artifacts to me. – xiota Apr 18 '19 at 2:45
  • 1
    I'm not sure what image viewer is being used (so it might be showing JPEG previews), but we can see in the footer that these are ARW (Sony RAW) images. I also agree that this is not typical of JPEG artifacts. – mattdm Apr 18 '19 at 4:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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