[Dartmoorman]

I had this problem with my skies the other day and thought it was my new 18-135 lens on my new K5 mk2 producing those lines. I shoot raw and jpegs are set to four star 16mp, no filters. Then someone told me it was not the lens but a resolution problem. I'm not sure I fully understood the explanation so can anyone help me out and advise what settings are causing this. It does not happen all the time, and I shoot in P mode.

Attached Images

View Picture EXIF

File Date

 

[fgaudet]

Are you talking about the gradient lines in the sky?

If yes, try changing your color space (I suspect you're shooting in SRGB) try AdobeRGB. Those lines usually happen when the image is converted between color spaces. Some website converters do that too, Facebook was very good at creating those gradients when I was posting pictures.

Edit:

Or do the opposite, put your camera in sRGB if it's in AdobeRGB and see if that helps.

[jatrax]

Are you seeing this on prints or on images posted to the web?

[Miguel]

Were you using a Polarizing filter?

M

[Dartmoorman]

Thanks folks,
The lines show up in the Raw file (19+MB) straight from the camera and are still there when I convert to Jpeg and print out like that. No Filter of any sort used.I will swap the colour space to see if that helps.

[Steve.Ledger]

Which RAW processor are you using? Banding should not show in RAW.

► https://www.pentaxforums.com/forums/120-general-technical-troubleshooting/221...nding-sky.html

► https://www.flickr.com/groups/nikond7100/discuss/72157639178604255/

[Na Horuk]

Yep, its very odd that raw files would display this. I would try using a different raw editing software. Raw Therapee, Gimp, FastStone are some free ones you can try, and many others allow a free trial version.
Keep in mind that if you are using certain raw viewers, they don't really show the raw data, they only show the embedded preview jpeg.

Edit: Oh, and its not so much "resolution" (the detail measured in a given area) as it is a "depth" problem. Each pixel contains a certain amount of data, and when the gradient is too fine for this data to convey the differences slowly, banding like that may occur. And jpegs have a much smaller bit depth than raw files (+jpegs have lossy compression, emphasizing colour problems). With raw files, each pixel should have more than enough data for the sky to be smooth, without transitions or banding. With jpegs, especially if the compression is high (=low quality), this is common.

[Dartmoorman]

Thank you, I'm using Faststone RAW converter, It is not something that happens all the time which made me think it was something I am doing to the camera. I will remove Faststone and revert to Pentax supplied software to see what happens. Attached is a Faststone conversion without the problem.

[Imageman]

That just isn't the way raw files work.


The individual pixels in a raw are separate and unique with their own chroma and luminance values unrelated to adjacent pixels. (a pixel is not a sensor photosite)


This kind of banding which is a grouping together of related pixels in discrete and similar band groups is just not going to happen.


It suggests there is a lossy compression going on such as Jpeg which you are then viewing.


I suggest you make some test images, the workflow is:-


Raw - output to TIFF - output to High resolution Jpeg - output to Low resolution Jpeg.


Then examine the files, I believe that you wont see any banding in the TIFF but banding will be evident in the Jpegs, and if you see banding in the original Raw but not the TIFF created from it, that shows that there wasn't really any banding in the Raw at all, but a viewing effect only.

[clackers]

Dartmoorman, I bet the RAWs aren't posterized, but you haven't developed them.

That is, when looking at the RAW files in software you're only looking at their JPEG thumbnails.

[Lowell Goudge]

Irrespective or raw or JPEG the issue is that most likely there was an exposure error and while correcting exposure error the expansion of the mid tone range in the sky leaves insufficient data between each color. Banding is the result of over expansion of a tonal range

[clackers]

Irrespective or raw or JPEG the issue is that most likely there was an exposure error and while correcting exposure error the expansion of the mid tone range in the sky leaves insufficient data between each color.

The OP's claim is the RAW files looked like that straight from the camera - no corrections.

[Dartmoorman]

The Raw files did shown this right out of the camera.The Raw files were about 23Mb in size. I imported them from them from the sd card with Faststone which I like using. It must be something I am doing (have done), So am going to reset the camera and reinstall Faststone to see if that cures the problem. Thanks to all for your help.

[Lowell Goudge]

The OP's claim is the RAW files looked like that straight from the camera - no corrections.

Even with JPEG it is very very unlikely that you would have noticeable banding on the sky, because normally there are about 40 steps within a stop in the middle 5-6 stops of the image

To get the type of banding here you need to cut that down to about 6 steps per stop, which means you are working well out of the middle +/-3 stops of dynamic range.

I suspect there has been some form of either SCurve correction or mid tone range expansion done. Maybe an auto exposure application upon generating thumbnails

[normhead]

As Lowell says, banding is caused by trying to put more information into an image than is there. With a 14 bit engine, that is very unlikely to happen. It's unlikely to happen even with a 12 bit K-01 or K-30. It was fairly common with 8 bit cameras. I've had it happen when manipulating images saved on the camera in jpeg before. It's definitely not the lens.

I've also seen images of skies that looked like banding....