Using AstroCrop to crop and register Bridge camera images taken on a static tripod.

The camera used is a Panasonic Lumix DMC-FZ72, 60x optical zoom bridge camera mounted on a sturdy Manfrotto static tripod. Images are captured at ISO 100 in burst mode (3 images per burst). If the sky stays clear during the capturing session, the aim is to capture 150 images. Allowing for the images to be saved to a fast SD card, it takes about 2.5 minutes to capture all of the images. During this time interval, there is virtually no image rotation, so stacking does not require derotation.

The first example was to image the 99.6% Moon.

At ISO 100 and 1/1000s exposure at 60 x optical zoom at f/5.9 and F=215mm. With the sensor used in the camera, this is the equivalent of 1200mm with a 35mm sensor. The 150 images are captured as the highest quality jpgs and a captured image has dimensions of 4608 x 3456 pixels

Precisely cropping and registering the images in AstroCrop.

In order to get very precise and well registered crops it is necessary to set the crop search perimeter (the distance searched away from the crop box) to a large number. We used 35 pixels . The larger the search perimeter value, the longer the cropping/registering will take. With a modern computer, the time taken is acceptable.

A single frame from the camera

The reference frame is selected and the crop box is sized and positioned

Cropping and registering in progress

Single cropped image

The cropped/registered images sum-stacked in Siril

Wavelet sharpened in waveSharp

Levels adjusted in waveSharp and saved

The image was post processed in Gimp 2.10

99.6% Moon

The second example was to image the Sun

Lumix DMC-FZ72 fitted with an ICE ND 100000 filter

A Panasonic Lumix DMCFZ72, 60x optical zoom bridge camera fitted with an ICE ND100000 solar filter and mounted on a static tripod was used to capture 69 images of the Sun at ISO 100 and 1/2000s. The images were precisely cropped/registered in Nicola Mackin's AstroCrop before being sum stacked in Siril, wavelet processed in waveSharp and post processed in the Gimp 2.10.

Reference frame set and crop-box sized and positioned

Cropping/registering in progress

Sum-stacking the cropped/registered images

Wavelet sharpening in waveSharp

The image was post processed in Gimp 2.10

The Sun showing the huge sunspot group AR3590

These two examples demonstrate that when images are captured in a sufficiently short period of time that image rotation is negligible by a camera on a static tripod, AstroCrop can crop and register the images so that a simple sum-stacking is all that is required to produce a 16 bit stacked image ready for processing. We find that capturing 150 images is sufficient to give a worthwhile image with an increase in signal to noise ratio of  12.25 compared with a single image. The camera we use is quite old and it is possible that more modern versions may capture faster and more images in burst mode. There is, however, a law of diminishing returns for the increase in Signal to Noise ratio as a function of the number of images stacked. However, the more images that are stacked, the higher the signal to noise ratio in the final image.

The 150 images does give a respectable S/N which allows for wavelet enhancing of the fine details without enhancing the noise.

Creating every synthetic narrowband image palette from One Shot Colour (OSC) RGB images

Creating every synthetic narrowband image palette from One Shot Colour (OSC) RGB images using the PixInsight Toolbox Script CreateHubblePaletteFromOSC. 

We have previously discussed the use of this script with respect to creating a synthetic Hubble Palette SHO image. I use the term ‘synthetic narrowband image palette’ because, coming from an OSC RGB image, all of the necessary Ha, SII and OIII data are simply not available as separate (or separatable) colour channels as they are in true narrowband or (Ha, OIII) and (SII, OIII) pairs of dualband images.

The CreateHubblePaletteFromOSC script takes an OSC RGB image and produces reasonable approximations of all of the narrowband palettes with the exception of SOH (which I feel must be an oversight). It also produces an approximation to the HOO two element palette.

Fortunately there is a Utility script called Swap channels that allows us to take an ‘OSH’ image produced by CreateHubblePaletteFromOSC and swap the R and G channels to produce an ‘SOH’ image.

Whilst many astro-imagers regard PixInsight or Siril etc to be ‘one stop shops’ for all of their image processing, my approach is to use whatever software I feel is most appropriate to my workflow, which includes a number of programs such as the Gimp, Neat Image and Photoshop CS2.

As previously, we start with an OSC image of the Elephant trunk nebula captured using AstroDMx Capture, through an Altair Starwave 60 ED refractor with an 0.8 reducer/flattener and 2" magnetic filter holder with an Altair 2" Quadband filter, and an SVBONY SV605CC OSC 14 bit CMOS camera. 

It is best to work with starless images for processing nebulae and this was done with StarXterminator in PixInsight.

Invoking the StarXterminator Procedure

Starless image an a stars image produced

Launching the CreateHubblePaletteFromOSC script

Producing a 'HOS' image for example

Synthetic narrowband images produced by the CreateHubblePaletteFromOSC script

All of the starless, synthetic narrowband palette images were saved along with the stars image.

Using the Swap channels script on the 'OSH' image produces an 'SOH' image which was saved

The starless images were stretched by the same amount, below are three example

The images were denoised using the same settings in Neat Image as the three images below illustrate

Table showing the synthetic palettes in the montage of images below with the OSC RGB original at the bottom right

Montage of synthetic palettes as shown in the table above

The CreateHubblePaletteFromOSC Toolbox PixInsight script produces sufficiently good analogues to narrowband palettes that allow an imager using an OSC camera producing RGB images to be able to present their data in a way other than simple RGB.

Test of a PixInsight Toolbox Script to build a synthetic SHO image from an OSC image taken through an Altair Quadband filter.

The image was captured using AstroDMx Capture, through an Altair Starwave 60 ED refractor with an 0.8 reducer/flattener and 2" magnetic filter holder with an Altair 2" Quadband filter, and an SVBONY SV605CC OSC 14 bit CMOS camera.

The Elephant trunk nebula

RGB image

It is best to remove the stars before the nebulosity is processed. Star removal can be done in PixInsight, Gimp or Siril.

The starless RGB image was processed with a PixInsight scripts:

Scripts  ► Toolbox  ►  CreateHubblePaletteFromOSC

PixInsight with the synthetic SHO image on the left

Starless synthetic SHO image

This image doesn’t look like a typical published SHO image but it does look like many SHO images when they are first constructed. It is usual for SHO images, which are very green, to be selective colour processed to reduce the green and to introduce the familiar golden tones of the Hubble palette. PixInsight, Photoshop, Affinity Photo and other software are all able to do selective colour processing.

Synthetic SHO image loaded into Photoshop CS2 for selective colour processing

Selective colour processing in Photohop CS2 to bring out the Golden hues and the blue.

Curves processing in  Photoshop CS2

Neat Image was used to denoise the image .

Starless synthetic Hubble palette

Stars that were removed from the original image

Synthetic Hubble palette image with the stars replaced

The Wizard nebula with Altair Quadband filter data

Processed RGB image

The stars in the original RGB image were removed by starXterminator in PixInsight. However, they could have been removed by Starnet++ in PixInsight, the Gimp or Siril.

Then:

Scripts  ► Toolbox  ►  CreateHubblePaletteFromOSC

The CreateHubblePaletteFromOSC script was applied to the starless RGB image which produced a very green synthetic SHO image.

The image was saved along with the stars image, and the synthetic SHO image was denoised in Neat Image. (of course, denoising could have been done in PixInsight or other software such as the GMIC plugin in the Gimp.

Then the denoised image was loaded into PhotoShop CS2 and subject to selective colour processing to reduce the green and introduce the desired golden hues characteristic of published Hubble Palette images.

Progressive stages of selective colour processing

Final stage of selective colour processing

Curves were used to further process the synthetic SHO image.

The image was then denoised again in Neat Image.

Synthetic SHO starless image

Synthetic SHO image with stars back

In conclusion, the PixInsight Toolbo script CreateHubblePaletteFromOSC is very capable and with an OSC image produces a result that can, by selective colour processing, produce a reasonable approximation to a true SHO image. It should be noticed that the same script can produce its own version of every possible palette that can be constructed from Ha, SII and OIII data except SOH, which must be an oversight. We shall explore these in the future.