Do Male Or Females Get Cystic Fibrosis

Example color calibration of large-scale processing (wavelets)

What I show below is a simple way to process large-scale structures from the brainchild of Carlos Milovic (PTeam).

This sample processing large scale is made on an image of the Milky Way centered on the constellation of Cygnus. This is a single 30 minute exposure Fuji film Provia 400F chemistry through a standard lens 50mm F4 closed, taken in August 2004 under clear skies and dark (Ordesa).

This image is an excellent test which enhance large-scale structures. The picture is processed entirely in PixInsight Core. The adjustments include color calibration, correction of the backlight (ABE) by a division based on prior data linearization (Divide), histograms, curves and noise reduction for luminance (ACDNR).

In this photograph, to process large-scale structures a problem occurs. The stars, especially the brightest, for reflections in the upper layers of wavelets. This means that increasing the weight in layers where the stars are not present, they also exert some influence, creating strange halos.

The solution here is basically subtract the stars, to isolate them. It is not possible to simply disable the wavelets the lower layers (small scale) because the halos dark adversely affect the outcome. We must do the opposite: to duplicate the image (in this case the luminance) and her first disable all layers above 64 pixels. Then, it maximizes deringing parameter in the layers survivors and applied. This generates an image with the stars alone, without his halo negative.

then subtract the result (without climbing) and you get a picture without the small-scale components (smaller than 64 pixels).

This is usually enough to enhance large-scale structures by changing the weight of the upper layers of wavelets. However, if only enhance dark detail, bright structures dilute the effect. The solution is to invert the image with no stars, before and after processing the dark structures.

You must then add the result to the image of small-scale objects (mainly stars), this time Rescaling.

enhanced

Once large-scale structures, we combine the processed luminance original RGB image.

Here we have used the LRGBCombination process for several reasons. First, we import the luminance changing their relative weight with respect to the RGB data, on the other hand, we also avoid the loss of color saturation by light transfer function halftones on the channel C, while reduce the noise in the chrominance.

Finally I show the luminance processed with this technique.