Reply To: Composite images

Forums Imaging Composite images Reply To: Composite images

#614775
Dawson
Participant

I agree that there should be very clear comments on what has been done to manipulate ANY image. For planetary this will usually be along the lines of “best 50% of 10,000 frames stacked, sharpened with wavelets and brightness, contrast and RGB adjusted”. The difficulty arises when dealing with images like this lunar occultation of Mars. I have seen several images of this occultation online where the processing comments should read along these lines:

1. Best 50% of 10,000 frames of Mars stacked (taken 60 minutes prior to lunar occultation). Derotated in Winjupos. Wavelets and RGB adjustments.
2. Best 50% of 10,000 frames of the lunar limb stacked (taken 60 minutes prior to the occultation of Mars). Wavelets and RGB adjustments.
3. Capture of lunar occultation taken during the event to show apparent spatial interaction between the Moon and Mars and to collect timing points.
4. Photoshop editing of two separate images (Mars and lunar limb) to merge into spatially correct apparent lunar occultation of Mars.

Using a bit of lunar limb and a full-face Mars from Richard’s original data (https://britastro.org/observations/observation.php?id=20221211_181703_df51689fca6965b0) as shown below, it is possible to construct the attached animated gif, which is totally made up, but just a rough representation of what was observed in time-lapse. But this face of Mars wasn’t the one that got occulted by the Moon, it is just the same full face of Mars slipping behind a layer in Photoshop. The issue then is neither the actual lunar limb, nor the actual face of Mars are correct; if the Mars image was taken 60 minutes before the Martian features will be in the wrong place, and the illumination of the lunar limb will be subtly wrong too.

But this all boils down to what is the purpose of the image? If it is for scientific scrutiny then this made-up composite has little value. If it is for maximal visual impact of a lay audience then this method is likely the one to follow.

James