12
\$\begingroup\$

Some photos of the sky taken on our Nikon D5100 have a diamond-like grid pattern. I'd like to know what may be the cause. The photo was taken in RAW and the pattern appears regardless whether it was processed with Apple Photos, Nikon's own software, or RawTherapee:

photo 1

Here it is (cropped) with contrast increased:

enter image description here

Lens: AF-S DX VR Zoom-Nikkor 55-200mm ƒ4-5.6G IF-ED

ISO 250, 82 mm, 0.3 ev, ƒ6.3, 1/400 s

Here's the original NEF file

Another picture, taken on different day:

enter image description here

AF-S DX VR Zoom-Nikkor 55-200mm ƒ4-5.6G IF-ED

ISO 160, 247 mm, -1 ev, ƒ5.3, 1/1000 s

original NEF file

\$\endgroup\$
14
  • 1
    \$\begingroup\$ Was this taken from inside a vehicle, through the windshield or a closed side window? \$\endgroup\$
    – scottbb
    Dec 6, 2021 at 18:47
  • \$\begingroup\$ No, it was taken outside, no filters on lens, nothing between the lens and the air. \$\endgroup\$ Dec 6, 2021 at 18:53
  • 2
    \$\begingroup\$ Using Nikon's software, if you disable lens correction, does the pattern go away? \$\endgroup\$
    – scottbb
    Dec 6, 2021 at 20:38
  • \$\begingroup\$ @scottbb Unfortunately not, the pattern is still there. Here's the original NEF if someone wants to play with it: dropbox.com/s/0qraa5qhbd37a59/DSC_0211.NEF?dl=0 \$\endgroup\$ Dec 7, 2021 at 15:16
  • 1
    \$\begingroup\$ @BobMacaroniMcStevens it's a crop. I didn't try printing it. \$\endgroup\$ Dec 8, 2021 at 22:06

3 Answers 3

1
\$\begingroup\$

You can do an experiment to isolate the problem and possibly rule out something happening in the lens electronics.

First, if you have another lens, try it! I assume you would have done this already were it an option, but I mention it just in case.

Second, take an exposure with the lens off and point the naked sensor at something gray like the sky, so it stands out the same. If the noise is still there, it's probably that decade old sensor failing on you.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks for suggestions. The other lens (AF-S DX VR Zoom-Nikkor 18-55mm ƒ3.5-5.6G) also produces the grid, so it may be indeed a problem with the sensor. The camera was used frequently for first few years, and then was left unused for a few years. \$\endgroup\$ Dec 11, 2021 at 22:03
1
\$\begingroup\$

One thing ahead: A RAW files usually is not as "raw" sensor data as many think. There might be some processing even on that data.

What I believe, we see here, is the processing of the RAW within the camera becoming visible. To see the artifacts, you have to push the image beyond usual processing - and probably beyond the dynamic range of the sensor.

I also found out that the problem becomes less pronounced if you apply the profile corrections for the lens you used.

I am attaching the images from Adobe RAW here. Sorry for using the German version here. But I think you can interpolate the settings from the description.

Neutral Image Image Neutral

Image with Black pushed -100 Image Pushed Black -100

Image with pushed Blacks and Lens Profile Image Pushed Black -100 with Profile

So what you do, is to change some of the image by -3 f-stops. This might be enough to introduce some sort of artifacts. Although I never saw diamond shaped banding, this might just be what happens.

If you look on the histogram you see that at -100 Black, a tiny sliver of the histogram becomes almost 50% of the histogram. The tonality here is stretched pretty much to the breaking point.

One Solution

If you take images that need such a wide dynamic range, you might be better using bracketing - i.e. taking several images with different shutter speeds and use a hdr software or even just photoshop to combine them into an hdr image. This way you conserve the areas where your camera is best while also having more information to work with. Your camera can do automated bracketing, so unless you have a lot of movement in your image, this would be the way to go.

\$\endgroup\$
1
  • \$\begingroup\$ @DmitryChestnykh I added bracketing as a potential solution to your problem. \$\endgroup\$ Jan 23, 2023 at 8:28
0
\$\begingroup\$

I have personally observed this type of pattern on clear day blue sky with the naked eye. I was wearing polarized sunglasses, I took them off and the hexagon pattern was still visible without the glasses.

I observed this in 2021 mid-spring in Montreal, Canada.

Here is a thread on Reddit I discovered.

\$\endgroup\$
2
  • 4
    \$\begingroup\$ You seem to have forgotten the link, please edit it in. \$\endgroup\$
    – FreeMan
    Jan 19, 2023 at 18:34
  • \$\begingroup\$ Interesting theory, however I haven't seen this pattern with my eyes (or via the lens when shooting), and it didn't appear in iPhone photos of the same subject. \$\endgroup\$ Jan 20, 2023 at 19:14

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.