5DIII Low ISO banding after boosting -> fixed

Started May 3, 2012 | Discussions thread
Peter 13
Veteran MemberPosts: 8,301
Re: fix this
In reply to John Sheehy, May 6, 2012

I agree. I meant - the clear advantage that you see there cannot be explained by some kind of cheating by Nikon.

John Sheehy wrote:

Peter 13 wrote:

And this is D7000 vs. 60D, IR images. No truncation can do this:

Truncation has its problems, but they are very minor compared to the differences in read noise quantity and quality amongst current cameras, for a single normal exposure.

Truncation at mean black in the RAW makes the data inferior for banding correction, but if you have less banding to begin with, that's moot. Truncation makes the data inferior for stacking, as you don't have "negative" noise to cancel out the positive noise. It is also slightly inferior for any downsampling or binning done to RAW data before black clipping, because, again, you are not binning "negative" noise with positive noise, resulting in a slightly lower SNR and non-linear means for signals very close to black. AFAIK, however, most converters resample the image after demosaicing and black-clipping, which is inferior, and negates any benefit from having negative noise in the RAW.

-- hide signature --


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