Software-y-Ddraig
(Dragon Software)

Overview

Home

Contents

Search

User Feedback

Home

Up

About DarkManager

DarkManager is a utility for creating and maintaining a reduction frame library. Currently, it can automate the acquisition and management of both dark and bias frames, using CCDSoft version 5 and a supported CCD camera. This allows the reduction frame library to be built while the telescope is idle, perhaps during the day or while the owner is asleep.

Why Use a Reduction Frame Library?

Two reasons. First, convenience. Second, better quality results.

Convenience

DarkManager automates the acquisition of dark and bias frames, based on parameters that you specify ahead of time. Once the acquisition sequence has been started, it can be left unattended and will take dozens, even hundreds of frames without user intervention. This means you can take your dark frames at the end of an imaging session, while you get some sleep, or during the day, or on nights of bad weather when imaging is impossible. Simply, you maximize your imaging time and waste less time capturing reduction frames.

"But aren't reduction frames taken on-the-night supposed to be better?". Well, it depends on your technique. If you are going to take a few reduction frames with each image, then you may be better off taking your reduction frames on-the-night. However, techniques are changing and many imagers are finding that using a reduction frame library is more efficient and gives better quality results.

Better Quality

Building a comprehensive library of reduction frames enables the use of advanced statistical methods ("Sigma combine") which, when used with a large number of frames, can reject most of the noise and anomalies from frames and produce a high-quality master frame. Typically, fifty or more frames will be used. The resultant master frame is then used to reduce the light frames in the normal way. This technique produces results that outperform the standard "average" and "median" combines that are available in most CCD imaging software. Clearly, taking fifty dark and fifty bias frames during each night of imaging is impractical and a waste of valuable imaging time. DarkManager enables the library to be built at a more convenient time and ensures there are always sufficient frames available to produce high quality results.

Work Flow

Getting the best results from a reduction frame library requires a slightly different workflow to the traditional method. The technique described here is that used by me, Tim Long, the author of DarkManager. No originality is claimed; this technique was derived from various posts on various imaging newsgroups, notably on Ron Wodaski's CCD-Newastro group. DarkManager was written specifically to support this technique.

Build the Reduction Frame Library

My personal library consists of a minimum of fifty frames at each of 0C, -10C, -15C, -25C, -30C and at exposure times of 5 minutes (300 seconds) and 10 minutes (600 seconds). That's around 50 x 5 x 2 = 500 dark frames and 50 x 5 = 250 bias frames, a total of around 750 frames with a total exposure time (for dark frames) of about 3750 minutes, or about two and a half days. I do not try to build the whole library in one go. It is better to build smaller batches of frames on different nights. Depending on the season, some temperature set points will not be appropriate, for example, I can only cool to around 0C during high summer but in the winter I can often get down to -30C. So I concentrate on different temperatures at different times of year. Currently, I am not doing much colour imaging so I take all my reduction frames binned 1x1. If colour imaging is a requirement, then DarkManager can easily accommodate binned exposures. DarkManager automates this process as follows:

• Allows the user to describe their desired (or target) reduction frame library. I call this the "desired matrix". Once this has been done, DarkManager will display a list of all the frames that are "missing" from the library. Initially, of course, all of the frames will be missing.
• Displays the reduction frame library using a TreeView control. Branches in the tree represent folders on disk and leaf nodes represent FITS files (much like Windows Explorer). Each item has an icon that represents the item's status with repect to the desired matrix (missing, present, aged, surplus etc).
• Automatically acquires missing frames using CCDSoft (Software Bisque's camera control software).

Discard old frames

Camera characteristics tend to change slowly over time, perhaps over a few months. This will vary from one make and model to the next and between different individual cameras of the same make and model. So, after a certain time, old reduction frames will no longer be representative of the actual noise and defects present in the camera. The easiest way to remedy this is simply to discard (delete) old frames. New frames can then be taken to replace the old, deleted frames. By rolling in new frames occasionally, the library is maintained at full compliment, while ensuring that the frames are relatively up-to-date and representative of the actual camera characteristics. I discard frames older than 4 months from my library. DarkManager supports this in the following ways:

• The user is able to configure how old a frame should be before it is deleted from the library. DarkManager will detect "aged" frames and highlight them on screen. Aged frames can be easily selected and deleted. The frames may be deleted individually, or all of the aged frames in the entire library may be deleted in one fell swoop.
• DarkManager continually monitors how many frames of each temperature, exposure time and binning mode are in the library. Once old frames have been deleted, it will automatically know how many frames of each type are needed to replenish the library.

Produce Master Frames

The point of keeping all these reduction frames is to enable the production of high-quality master frames. Typically, for each new image, I will produce a new set of master frames from the library. Currently this must be done with a third party utility such as Ray Gralak's Sigma. A future version of DarkManager will incorporate this function. I use 50 dark frames and 50 bias frames, to produce two 'master reduction frames' (one dark and one bias). Combining this many frames needs a lot of memory and can take some time to complete. I store my master frames along with my image data.

DarkManager supports this in the following ways:

• Reduction frames are organised into a logical, easy-to-understand directory structure, making the frames easy to locate and use as input to Sigma.
• A planned future enhancement will enable the frames to be processed directly in DarkManager

Reduce the Image Data

This is done in the traditional way. In CCDSoft, a reduction group can be created containing the two master reduction frames produced in the previous section. This reduction group can then be easily applied to a whole folder of image frames. DarkManager plays no part in this process; it merely creates the master reduction frames which are then used in CCDSoft, or other image processing software. I think you will find that the resultant images are cleaner than those produced by traditional techniques.

Future Plans - Reduction Frame Utopia

DarkManager is a work-in-progress and will be improved over time. Ultimately, I hope to produce a product that will ensure that reduction frames are always available, on tap, whenever they are needed, with no human intervention whatsoever. I hope to produce a program that will automatically run at scheduled intervals, discarding old frames, acquiring new frames, combining the frames to produce master frames. As such, CCD imagers will never need to worry about producing reduction frames, because a fresh, up-to-date set will always be available ready for use.

The first part of this vision is already a reality. DarkManager is available for download today and can automate the acquisition of your reduction frame library and assist in recognising 'Aged' frames, missing frames and surplus frames.

The second part of this vision is currently under construction. A sigma-reject algorithm is being developed that will allow DarkManager to combine the frames in the library to produce the master frames needed. This will be able to operate on selected frames, or the whole library in one operation.

The third and final phase will concentrate on scheduling (using Windows task scheduler) and the complete automation of the reduction frame work flow.

-End-