The biggest problem of the DP2M (truth and facts) : BANDING

Started Aug 16, 2012 | Discussions thread
balchinian
Regular MemberPosts: 293
Like?
Re: Solution path is clear
In reply to balchinian, Aug 17, 2012

balchinian wrote:

PrebenR wrote:

I used Roland's x3f tool to extract the three "layers/colours"* separately.
) Have not quite understood how the RAW data is.

Here you can see the three different colours when pushing the levels to provoke the banding patterns.

As I say I'm not sure yet how to interpret the colours/layers, so I cannot conclude anything, but seems to be more banding in the blue than the red above.

-- hide signature --

AF takes away the Zen in photography

PrebenR wrote:

Cropped another place in the photo and banding corresponded. Actually at pixel accuracy.

We'll see...
--
AF takes away the Zen in photography

jbr wrote:

http://forums.dpreview.com/forums/read.asp?forum=1027&message=41420931

PrebenR wrote:

Robert F. Tobler wrote:

After some more experiments with the debanding option of Topaz DeNoise 5 I found a serious problem with it: if there is image content that is pixel-perfectly aligned (either

horizontally or vertically) it is interpreted as banding, and thus serious artefacts are
introduced. So its back to Nik Dfine for debanding...

If the pattern is fixed, one should find the signature so that one can use it to correct for the banding. So the best way would be to find out how to best obtain the signature. It does look like the Blue colour contains the most banding, so maybe that is the best place to start.

--
AF takes away the Zen in photography

Brilliant! It's not noise! Noise reduction unnecessary! This can be fixed in firmware with no loss of detail!

Problem solved:

http://forums.dpreview.com/forums/read.asp?forum=1027&message=42281999

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow