In processing images in PS from raw 5d2 files (also seen with 1ds & 1ds2) clear blue skies will quickly show banding. When raw files are first converted to tifs they appear fine but with any manipulation of the sky the banding starts to become evident. This happens with files converted in both DPP and ACR. I work in either Prophoto RGB or Wide Gamut RGB color spaces 16 bit. Does anyone else have this problem? Can anyone provide information to lessen or completly avoid this happening?
Thanks for any help or info, its much appreciated.

Click to view attachment

Peter

Architectural Photography by Peter Montanti, www.mountainphotographics.com

Peter,

I'm assuming the file attached here is a screen grab. So what you are seeing may not be the file or an issue with the 5DII (and other cameras) but is an issue of the screen profiling or quality of the screen. Have you printed the file to see whether or not the banding shows on output even with heavy curves applied?

I would bet money, if you view these files on a higher quality monitor or better profile, the banding will not be present.

I am on an Apple 23" aluminum LCD. I've used an Eye-One II for years with the Eye-One Match software. I tried ColorEyes to calibrate and the quality of the profile is even better. But I still have not seen banding like that from these cameras.

The last time I saw sky banding like that was when I owned a Fuji S1 in 2001. The newest cameras have very clean gradations from raw files. I have owned a 1Ds, 1D, 5D, 5DII all have not shown that banding.

The most usual cause of banding given the settings you mention would be that the shots are under-exposed, causing the darker blue tones to sit on lower portions of the histogram where the number of tonal levels is exponentially lower exist in the upper portion, such that adding contrast to the sky would separate them further, causing loss of smooth tonal transition. This is a logical inference from what you say here, not a certainty, as I haven't seen the image histogram and where the blue tones sit in it. But have a look at that as a possibility. If that seems as if it could be the cause, make some test shots with plenty of blue sky, shooting further to the right (more exposure) without clipping highlights wherein you need to preserve detail, and try processing the sky in those shots to see whether or not banding still happens.

1. Do you see the banding when viewing the image in 100% on monitor?

2. If not, i.e. the banding appears only in reduced size: instead of reducing the zooming in the view, make a "proper" downsampling (onloy for test). DO you see the banding now?

I am asking these, because the size reduction in viewing is crude (but fast) compared to downsizing, i.e. you may see banding, which is present only in the actual monitor view.

I had assumed the scenario here is banding in a print, but if we are only talking about monitor view, then yes, depending on the quality of the display and the display profile it is possible to have display banding without corresponding banding in the print. The surest way to check for this is to make a print of an image that looks banded in the display and compare.

Another factor that comes to mind is whether anything you have set-up in the display or in your image processing programs has been changed recently - in the sense that given the equipment you are describing, presumably there was a period of time in which this problem did not happen?

It's the monitor. Meaning, the quality of the monitor or how the monitor is set or profiled.

I don't even get banding like that from my G9 which can make lovely prints.