Magenta edge tinging at high ISO and a possible solution...

Started Aug 17, 2009 | Discussions thread
GordonBGood Veteran Member • Posts: 6,312
PROGRESS: version now supports PEF files...

dlacoutu wrote:

Maybe you could allow some settings by command line or INI file, like blur window, hot pixel threshold, and so on...

This would had the added benefit of 1) easing fine-tuning the application, and 2) allow some camera-specific tuning (if a camera needs a smaller blur radius, for example)...

Dominique, I have avoided adding settings in order to keep the user interface simple and because I am too lazy to want to have to document how to use command line arguments or a settings/INI type of file. I like the current "just go do it" format, and so far see no need to change have settings.

You have suggested in the past being able to set the ISO at which the program starts to do conversions, to which I counter that once we know the ISO at which conversions are no longer necessary, I will just code that limit into the program (likely below ISO 200), but it really does no harm to apply the correction to lower ISO images as the correction will be very near to zero anyway.

Now you suggest being able to adjust the blur window, but I have just played with smaller blur windows and find that there is no advantage since the smallest blur window I can use without adding to the Horizontal Pattern Noise doesn't do anything about getting rid of your "amp boom" horizontal ridge unique to probably only a few cameras.

I also played with algorithms that do straight averaging without respect to hot pixels and my current technique that eliminates hot pixels by just determining a population limit and find no real difference in results. Thus, I have reverted to my original algorithm, which doesn't really have a hot pixel threshold is it doesn't even see them (they are just above the population level that I track).

Any other options you can think of?

To all: interested, from the Read Me file:

"Further proposed changes are to add support for lossless compressed DNG files that have been converted from the uncompressed camera DNG files by the Adobe DNG converter or by Adobe Lightroom by default. The program will not work with DNG files converted from PEF's by the current version of the Adobe DNG Converter application due to this program stripping out the necessary black masked-to-light photosites at the right and bottom borders of the sensor in landscape orientation which are used by the correction algorithm.

This program may also eventually include a correction for wide banding vertical pattern noise as encountered for some K20D cameras and if users report slow processing times for huge numbers of conversions, addition of support to mulltiple processors to reduce batch processing time (currently less than a second per image file).

Version (3/9/2009) has the following changes:

1) Corrected a very slight offset bug in the calculation of correction values for the top (landscape orientation) although the size of the error would never be noticed.

2) Added support for PEF format raw files.

3) Determined that this method can not correct for "amp bloom" coloured ridges that may appear across the image (landscape orientation) when these "blooms" have no corresponding influence on the black masked border photosites on the right hand side."

And the link to the program files, as follows:

Compressed DNG file support will likely follow in a couple of days, as the bulk of the decompression/compression work has already been done for the PEF files. Note that I can not help any of you if you converted your PEF files to DNG format and threw away the original files, as the current version of the Adobe DNG Converter (including the embedded one in Lightroom) strips the masked to light black border needed to do this correction.

Regards, GordonBGood

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