Tuesday, October 6, 2009

Michael A. Phillips' Planetary Processing Routine version 8.1

Was this tutorial helpful?

Yes, thanks very much!
 
 8

I think I like the v7 better.
 
 0

Yes, but I'd like additional information and will post in the comments below.
 
 0

No thanks, I like my own routine.
 
 3

No, I have something I'd like to add in the comments below.
 
 0















HINT: To download this tutorial as a Microsoft Word Doc, click here - http://www.mediafire.com/?yj4em1omomn




Introduction

1. This is by no means a replacement to the tutorials I learned the most from such as Mike Salway's (http://www.mikesalway.com.au/2008/08/26/planetary-imaging-and-image-processing) or Paul Haese's (http://paulhaese.net/planetaryprocessing.html). I am simply tuning, tweaking and automating some of the steps to suit my own taste and hope you find some ideas here to use for yourself!


    1. Incorporating a large portion of the v7 tutorial (seen here), I was able to skip the Registax application altogether. The advantage here is the I now stack in Iris direct from Ninox and Iris has a very robust scripting engine to automate stacking and saving of raw files.


    1. All the routines in this HowTo use applications running on Linux, even the windows only software packages run under Wine (http://www.winehq.org) running under Linux. All of the video screencasts are recorded and produced using applications running on Linux. I am confident that all software and processes detailed in this HowTo can be run on MS Windows with one exception. The capture software I use, Coriander, will run only on Linux. Even some of the bash scripts may be run under Cygwin (http://www.cygwin.com/) although I've not tried myself. Never fear, the main method of this HowTo still applies to users running on MS Windows.


    1. OK, HERE'S the SHORT METHOD!

All videos are best viewed in HD (click the button in the lower right) and full screen!!

      1. Acquisition: Using Anthony Wesley's custom coriander 1.0.0. This works well for me as I have a firewire camera and motorized TruTek Color Filter Wheel with PC control cable. - Part 1 Video (http://www.youtube.com/watch?v=IF70aenFIq4)


      1. Sorting by quality, rescaling, centering and cropping in ninox: (http://acquerra.com.au/astro/software/ninox/ - A tool for processing planetary images by Anthony Wesley) – Part 2 Video (http://www.youtube.com/watch?v=NWZ7Gc8kmTM)


      1. Stacking in Iris: Major change in this version 8, is the absence or skipping of Registax for stacking. Move Ninox sorted and cropped images into Iris for stacking.


Stacking ninox sorted data right into Iris - Part 3 Video (http://www.youtube.com/watch?v=tjPwY5NQYnw)


      1. Processing in Iris: RGB combine, whitebalance, black point, sharpening and processing. - Part 4 Video (http://www.youtube.com/watch?v=HKlshHH3Zhg)




  1. Acquisition

      1. My equipment consists of

        1. I use a Celestron C8i SE (orange tube)

        2. Lenovo T61 laptop running Ubuntu 9.04 Linux and custom coriander (for firewire cameras – Thanks Anthony Wesley! )

        3. DMK 21AF04 (Firewire)

        4. 2.5x PowerMate

        5. Astronomik LRGB filters

        6. True Tek Color Filter Wheel with visu diag and PC serial cable (USB-to-Serial)

        7. Homemade Hand Controller extender or PC mount control for drift corrections

        8. some shots here:























http://picasaweb.google.com/maphilli14/MichaelAPhillipsAstronomyEquipment#5244213374458800210





























http://picasaweb.google.com/maphilli14/MichaelAPhillipsAstronomyEquipment#5244213381800624706



      1. Once everything is connected up I follow the 'custom coriander 1.0.0' modified by Anthony Wesley (aka Bird - http://acquerra.com.au/astro/software/coriander/ ).

      2. Before beginning capture, I suggest you get some basics covered - Michael A. Phillips' Astronomy Lesson on Seeing, Collimation and Focusing (http://maphilli14.multiply.com/journal/item/70/)

      3. Here is a screencast that demonstrates a regular capture routine that I follow for Jupiter. - http://www.youtube.com/watch?v=IF70aenFIq4

      4. After completing your captures you may move onto the next section that moves data from 'transit' to 'sorted'. It is important to note that this coriander outputs a static image stream of .fit files not a .avi or video file. This saves me a step extracting the movie to images later on!






Sorting by quality, rescaling, centering and cropping in ninox

      1. I prepare all captured source images that are really .fit files by passing it through ninox (http://acquerra.com.au/astro/software/ninox/) to sort and crop the images and then output to a new subdirectory called 'sorted'

      2. An optional, short getting started video on downloading, installing and using ninox - http://www.youtube.com/watch?v=FpOJQxzyTnw

      3. Additional changes to the version 8 routine include batch processing of all the nights images in a single command as well as upscaling to 200% or 300% of raw images before stacking.

      4. The full script is a bash script and can be found here:

        1. Generate list of nights captures - http://www.mediafire.com/file/nfyimywxnny/MAPgenplanetlist.sh.zip


Execute looping ninox:

        1. In version 8 of my routine, I now have a simple loop that reads all the nights captures and allows you to selectively run Ninox against all or some of the captures. This script that will read the capture directories and use them to create all the sorted, Ninox folders. Per these guidelines (http://maphilli14.blogspot.com/2008/01/tips-on-organizing-your-computer.html).

      1. The key ninox syntax is:

        1. ninox -width=300 -height=300 -cutx=300 -cuty=300 -qestimator -qrenumber -outdir=Astronomy/Sorted/2-Scratch/Jupiter-$1/$2 /Astronomy/Transit/1-Corianders/$1/$2

          • This, “-width=300 -height=300 -cutx=300 -cuty=300” crops the images to a square 300x300 pixel size

          • This “-qestimator -qrenumber” sorts the data and renumbers the images based on quality

          • This “-outdir=Astronomy/Sorted/2-Scratch/Jupiter-$1/$2” allows me to specify a new folder in a sorted (not transit) folder for safer keeping. The $1 and $2 are the arguments passed to the script at runtime in which $1 is the ISO date that coriander writes ( IE 20081122) and $2 is the UTC time (IE 001122UTC).

          • There is another path following the outdir, /Astronomy/Transit/1-Corianders/$1/$2 this specifies the source files if not in the current directory.

          • Upscaling is in the format of -resample=3/1 scales all the subframes to 300% the original size after cropping. A tip from Anthony Wesley citing that it helps to remove artifacts made me want to try for myself and it in fact does work quite well on good data!! I've gotten fantastic results running a raw stacked, 3/1 upscaled image through AstraImage, which is part of my long routine, v8.2





Stacking in Iris: notes on skipping Registax and how to get 'Iris-ready'

      1. Why skip Registax? Only because I can automate stacking and saving of Red, Green and Blue with a single click!

      2. Skipping Registax is a two step process. First ensure the proper naming of files. Second, use the command window or Iris script to automate the stacking.

      3. The key to taking Ninox sorted data straight to Iris is in the naming. Iris has a fairly rigid naming scheme for any sequences of data that it works with. In particular if you are going to stack a series of raw monochrome data in the red channel it will NEED the numbering to appear as follows, as this is how Iris requires the images to be numbered, IE no leading zeros.

      • R1.fit

      • R2.fit

      • R10.fit

      • R99.fit

      • R100.fit

      • and so on...

This is not the same scheme that most image capture programs work with and I was able to solve this with the Linux package for batch copy / move called 'mmv' (mmv is a program to move/copy/ append/link multiple files according to a set of wildcard patterns)

In a nutshell your files need to get renamed from:

xxxx-yyy-q001999.fit → R1999.fit

or

xxxx-yyy-q00239.fit → xxxx-yyy-R239.fit

where all all leading 0's get stripped off as Iris does not like that.

I've not tested yet but for windows users this suggestion came across the Yahoo Group for Iris Software via Christian Viladrich:

The numbering rule in Iris is :
1, 2, ..., 10, 11, ..., 100, 101,

So, if you have a numbering such as :
0001, 0002, etc
you have to renumber your files.

You may renumber your files with "remultfile" software:
http://www.ptorris.com/windows.php

This is a French software, but very easy to understand.”” - Christian Viladrich



The fun part is starting the automated stacking and walking away for something more exiting!

Once the files are in the right format you may use this Iris script (http://www.mediafire.com/file/xzy3ihqamwt/MAPPlanet81.pgm.zip)
to convert and or stack the raw images in a manner similar to Registax's stacking scheme. Before using, test with one or both of these commands:



Non-FIT users need to convert to fit first! (For BMP users, use PIC settings not FIT settings)

BMP2PIC red redpic $1


add_norm red $1

Where $1 is simply the number of frames 'located in the iris working directory' that are to be stacked.

Some notes about the final image. First it is a After stacking you may find the brightness to be clipped or oversaturated. This is actually not the case and you should notice that Iris works in 48-bit space so you may fix this issue by using the following command:

mult $2

Where $2 is a number less than 1 to which all pixels are multiplied by. If you save the raw stack you may reload at anytime before you find the final desired brightness.







  • This image was a normal brightness image which I used mult 1.5 several times to give the over exposed view on, but using mult 0.8 or mult 0.5 a few times will bring it back to this. If it does NOT bring it back then you have gone beyond the 48-bit max value point and I've never encountered this before with my 8-bit camera and anywhere between 500-600 frame stacks.


























Processing in Iris

Flow – I like to follow a framework of a routine, that worked for me in the past under varying conditions. Here's my flow that I load from a text file and then paste into the Iris command window.















>run mapplanet81 500



>load r0



>--processing / wavelets-- 1,5,10,3,1



>unsharp 3 2 1



>save r81



>load g0



>save g81



>load b0



>save b81



>--view / rgb-- x81--



>white



>black



>save rgb81a



>scale 2 0.67 0.67



>--processing / wavelets-- 1.1,1.9,1.7,1,1



>save rgb81b



>--Saturation 1.3--



>gauss 0.5



>--processing / blur filter--



>visu 25000 200



>save rgb8a



>run rgb8a



>load rgb8a



>scale 2 0.75 0.75



>



Anything with a > is the 'prompt' and not a command

Anything with a >-- is just a note to find that in the menu system as it's not a command (not that I can see)







Step by Step Explanation:



stack per step 3 above with Iris script

Load each raw file (EG R0)

apply wavelets to taste

apply unsharp mask (optional and mixed success)

save as new filename (EG R81)

repeat for Green and Blue

RGB combine (EG Trichro – located in View RGB) - GREAT explanation of Whitebalance and black point by Sebastien Leboutte http://www.skyimaging.com/filter-wheel.php

whitepoint “

blackpoint “

save 'backup'

scale down to intermediate step (EG from 3x to 2x)

more wavelets

save new 'backup'

saturation increase to 1.3x

smooth with gauss

smooth with blur (again to taste)

adjust levels with visu

save as final

this final gets picked up by next iris script which saves as a variety of formats to get copied by bash script

At this point you're done and you can use Gimp or Photoshop or other similar editor to brush up and put in stats, logos etc...



Some additional Iris resources are found here:



Sebastien Leboutte - http://www.skyimaging.com/tutorials.php

and of course Christian Buil, author of Iris! - http://www.astrosurf.com/buil/us/iris/iris.htm (Scroll past halfway to find the TUTORIAL section)



TO DO's:



Include more MS Windows automation

Add Astra Image per my v8.2 routine.

Figure out how to use Deconvolution in Iris, in lieu of AstraImage.



HTH!



Mike




Monday, October 5, 2009

Jupiter 20091002



Yes, my C8i SE Single Armed Fork decided to take a break.  Ironically just after I got the CGE mount back!  The power wasn't getting to the hand controller just at the end of this session and the mount is on it's way to Celestron from out of warranty repair.  :(
Good news is the seeing was pretty good.
Not much else going on other than the Z-shaped / Zorro in the upper left!

Wish my mount a speedy recovery, The OTA will ride the CGE for a while... *sob* *sob* I'll have to one scope it for a while!  :)

Saturday, October 3, 2009

Waxing Gibbous Moon from 20090929

After my Pluto shots, I tried for the moon.  I've never really gotten a good full disk photo of the moon before.  Not one I liked at least.  I was even able to load the Canon raws (.CR2) into Registax 5 and stack 12 of the best ones.  AstraImaged for most of the detail and some touches in Iris and Gimp.  Hope you enjoy!


Blog Archive