F.Balázs: NGC 7000 area – There is something wrong with these colors

There was some debate about a picture published on the asztrofoto.hu website, whether the unfortunate colors are a result of processing or there is something wrong with the lens or the camera being used. The author (F.Balázs) published several similarly miscolored pictures and I thought I would investigate. The author was kind enough to publish one of his raws,  which I could use to verify some theories about what could have gone wrong and where. The picture to the right is the one published on the asztrofoto.hu forum. The metadata is Author: F.Balázs; Date: 2016-08-28 00:00; Optics: Nikon 300mm f4 fix; Mount: SkyWatcher HEQ-5 Pro with Syntrek; Camera: Nikon D700 (NOT modified); Exposures: 44*300s ISO 1600; Location: Hódmezővásárhely. It is obvious that something is very off.

The raw came in NEF format I could open with DSS and IrfanView. This is an 8bit jpeg cast as IrfanView output the raw file.

The NEF raw, cast as an 8bit JPEG by IrfanView

The raw picture’s telltale element is the orange hue, the light pollution coming from sodium lamps. Other than that, everything looks fine, I’ve seen plenty of this kind of images. The same picture revealed itself in DSS, so I checked this stage: there is nothing wrong with the camera or the lens. In addition, in RGB orange is #FFA500, full red and a good part of green – and this is a possible clue to what is wrong with the final picture.

Next, I followed three routes, each starting from the raw (or its jpeg cast), independent from each other.

1) in DSS (to open the raw) and adjust the histogram, and then in GIMP I tried the best I could do in a few minutes to get a fair picture.

stage1

adjusting the histogram – light polluted sky

playing with the curves

the final picture of route 1

2. Removing the light pollution’s orange hue by subtracting it from the picture, then recover with curves.

the sample to be removed

very little data is left

3) Setting a grey point with the pipette, thus adjusting RGB curves according to the light pollution with the knowledge of the region, where it should be black/gray/colorless enough for my experiment (the dusty areas).

removing the orange cast once

twice

My conclusion is that the camera and the lens are just fine, and the eery magenta hue is a result of faulty processing or a badly miscalibrated display.