PI Alternative To Stratton

[dave_galera]

With Nik's article in AN and Olly's images using Straton (nice work Olly) this got me thinking about PI and (as I am one) iMac users who can't use Straton........so after some investigation and a lot of javascript coding I have come up with two PI Scripts both which can be down loaded from:

http://www.qdigital-astro.com/pixinsightscripts

The scripts are:

StarReduction

LargeScaleStructureEnhance

I am just doing the instruction/information sheets at the moment so these will a few days before they are finished.

Both scripts are designed to work on monochrome images although I have tested them on RGB images and they do give acceptable results. They are both intended to be used within the normal PI workflow after Dynamic Crop, DBE and initial stretching with HT, but before combining into RGB, and, as Olly has done, only on the Ha and Luminance master channels

StarReduction

Does exactly what it says on the tin, it reduces the size of stars. The script can be run more than once to reduce the star size still further. It also has facilities to measure the mean FWHM and slightly sharpen the stars after reduction.

LargeScaleStructureEnhance

This script will enhance faint large structures such as Integrated Flux Nebulae (IFN) which are very close to the noise floor and is based upon the workflow as developed by Rogelio Bernal Andreo and fully described in the article (page 97) in Robert Gendler's book Lessons From The Masters.

From a technical standpoint, and being very pedantic, just removing the stars from an image using Straton is not the same as actually splitting an image into its Large and Small Scale Structures. I therefore decided not to go the Straton route but to adopt Rogelio workflow as being a more correct approach. I may develop a Straton look-alike script but have not yet decided.

Default Operation

With the check boxes set to their default settings (Extract Large and Small scale checked, all others un-checked) the script will split the image into its Large (LSView) and Small (SSView) Scale structures and these become available for further processing. Rogelio recommends a slight HT stretch on the SSView and a more aggressive stretch on the LSView.

Re-combine the two images by using PixelMath with the expression LSView + SSView after they have been processed and re-stretch the final image, you can use Auto Stretch or STF to give you a rough idea of how much stretch it needs.

You will find that the LSView just doesn't look right and is blurred, this is quite normal as you are seeing an image that has no small strucuture detail.

Optional Operation

Optionally you can let the script do the stretching of the LSView and SSView and then combine the two images into a final image, the LSView being stretched more aggressively than SSView. Complete the process by re-stretching this final image, you can use Auto Stretch or STF to give you a rough idea of how much stretch it needs. There is a choice of streching either view, both views or neither prior to creating the final image, this produces perfectly acceptable results, however, it is better to have direct control on the amount of stretching of the Large and Small scale views, and possibly perform other operations on them before re-combining using the Default Operation described above.

Examples

Master Ha image Dynamically Cropped, DBE and stretched with HT

Master Ha image after processing with LargeScaleStructureEnhance and one application of StarReduction.

If you try either of these scripts I would appreciate some feedback and let me know of any problems or enhancements.

Regards,

Dave

[Davey-T]

Looks interesting Dave, I'm in the process of trying to get all my astro needs transfered to Mac, starting with just image acquiring but hopefully more in the future.

Dave

[ollypenrice]

Nice one Dave.

I found Straton of little help in RGB, though this may be my fault. It thrives on the small tight stars of narrowband, it seems.

Olly

[dave_galera]

Nice one Dave.

I found Straton of little help in RGB, though this may be my fault. It thrives on the small tight stars of narrowband, it seems.

Olly

Hi Olly,

I am not surprised as RGB processing requires a different approach in the software. The norm is to split the the RGB into the channels, process them and stick them back together. In PI you can use Channel Extraction and Channel Combination to do this.

I am in two minds whether to bother with correct RGB processing as the normal workflow is to work on the master mono channels before RGB combination, and this is what I am trying to stick too, as I think you do, or did.

Dave

[steppenwolf]

I found Straton of little help in RGB, though this may be my fault. It thrives on the small tight stars of narrowband, it seems.

This has been my experience as well.