As mentioned previously, I always have an IDAS LP filter in the optical train. This filter cuts out emissions from common light-polluting sources (sodium and mercury vapour lamps) and as a bonus I find I do not need to radically alter the colour balance of my one-shot colour images. I have tried other "nebula" or "light pollution" filters and have found it necessary to make substantial colour balance changes to get a good colour-balanced image.
Also mentioned previously was the Hydrogen Alpha narrowband filter. The H-alpha filter works at a wavelength of 656.3nm, where a nanometre (nm) is 10-9m. This is in the red part of the visible spectrum and it is the emission line associated with singly ionised hydrogen (HII), which is the main light emitting species in emission nebulae such as the Orion nebula, the Monkey Head nebula, and the Rosette nebula. Since we are using just a very narrow band of wavelengths we do not need to use refractors with a good degree of chromatic aberration compensation; this opens up the possibility of using good quality camera lenses instead of high cost apochromatic refractors! Also, since we are only collecting photons of a single wavelength, it becomes pointless to use a one-shot colour CCD for imaging, and we typically use monochrome CCD cameras for narrowband imaging. As an added bonus the monochrome camera provides higher resolution than a one-shot colour camera based on the same CCD chip.
Again recall that narrowband imaging allows us to image in areas that suffer from light pollution, providing that the pollution is not too severe, and we can also image with the Moon up! This narrowband imaging really seems like a win-win situation, and in many ways, it is. There are downsides however. If you want to create a false colour image, you need to take several sets of data of the same image using different narrowband filters, and this pushes up your total imaging time. You also need to use long sub-exposure times as you have severely cut down the number of photons reaching your detector with the narrowband filter. This also greatly increases your total imaging time on an object. You will not get those "true colour" pretty pictures that you get with film or one-shot colour CCDs, but you can create nice "false colour" images. You can create false colour images by combining the images from several different narrowband filters. Other readily available narrowband filters include:
Hydrogen-Beta filters operating at 486.1nm. This is in the blue region of the spectrum and this emission line is associated with doubly ionised hydrogen. Note this is NOT the blue emission you see from reflection nebulae which is a broadband blue associated with the scattering of starlight from a dust cloud. A classic example of reflection nebulosity is the blue nebulosity associated with the Pleiades star cluster M45.
Oxygen-III filters operating at 500.7nm in the green part of the visible spectrum. O-III emission is associated with doubly ionised oxygen atoms and is often the dominant emission line from planetary nebulae.
Sulphur-II filters operating in the deep red region of the visible spectrum at 672.4nm. Singly ionised sulphur is also a common emission line in deep-sky objects.
And finally Nitrogen-II filters operating at 658.4nm also in the red region of the visible spectrum (we have 3 narrowband filters all operating in the red!).
False colour images can be formed by using various combinations of the above narrowband filters as the Red Green and Blue channels of the image to be formed. These different possible "mixtures" are called palettes, and the Hubble "palette" is one of many possible alternatives. The Hubble tricolour palette assigns S-II to the Red channel, H-alpha to the Green channel, and O-III to the blue channel. There are a number of iconic Hubble images using this particular false colour scheme, the "Pillars of Creation" being one very well known example.
So there in a nutshell are the basics of what's involved in narrowband imaging. You can carry out high resolution, deep-sky imaging in a relatively light-polluted environment (you can of course always take even better images from a dark sky site!). You can use lower specification equipment than that necessary for single shot colour imaging, this allows the use of short focal length camera lenses for those really wide field shots. You can also differentiate your work by creating your own "custom" palette.
What I haven't discussed in any detail is the fun and games you will have in overlaying and combining the different narrowband channels. You will find the stars in your separate images will have slightly different diameters according to the narrowband filter used, and this will make the formation of the final false colour image quite a processing challenge.
Computational Considerations -Data Acquisition and Image Processing
You will be acquiring and manipulating digital data in your new hobby, and this means you will need to use one or more computers as an integral part of your overall imaging system. There are many ways of handling the computer hardware and software issues and I will outline just a few.
Many people, myself included, start off by taking their laptop out to the observatory and using that to acquire the CCD data and to look after the scope driving and autoguiding. The software for downloading the CCD data, and autoguiding can be the native camera software, or it can be specialist software specifically written for carrying out the tasks, including image processing. I have used AstroArt [http://www.msb-astroart.com/] to great effect (AstroArt also has a great photometry package amongst many other goodies), but I currently use Maxim DL [http://www.cyanogen.com/] for all data acquisition, autoguiding, colour conversion and stacking. Returning to hardware considerations, any current model laptop is going to be good enough to carry out these tasks provided it has a fast USB 2.0 interface available for downloading the data from the CCD camera. Any reasonable size hard drive will also be more than adequate for storing a night's imaging data, the screen brightness is readily reduced to save your night vision, and it is very easy to carry a laptop around your observatory for the best access in what are usually cramped conditions. There is only one negative as far as I am concerned. I do not like taking a very expensive laptop out to a freezing observatory in the depths of winter and then bringing it back indoors to work on, or download the data. I know you can carefully wrap the laptop up in a plastic bag before bringing it indoors so that it can warm up slowly without filling itself with condensation, but I still feel it's a risky thing to do on a regular basis. I much prefer to have a very basic, small footprint computer, left full time in the observatory, and that is now my current arrangement. I also think that an LCD monitor is preferable over the CRT variety in the highly variable environmental conditions you get year round in an observatory. Again from experience, a little 15" CRT monitor didn't last one season before giving up the ghost, whereas my current 17" LCD doesn't seem to be suffering any problems at all. You can use a standard desktop (or server) configuration in the observatory, but space being a premium (especially if you also have a electrical greenhouse heater and a small dehumidifier unit taking up floor space as I have) means I use a mini ITX system with a 1GHz processor that has worked flawlessly - so far - http://www.mini-itx.com/store/ . The mini ITX system downloads the sub-exposures from the CCD, stores them in a file set up for the evening's imaging, and looks after the autoguiding of the Nexstar 11 GPS using Maxim DL software. With the SXV-H9C camera I also used the Celestron NexRemote software on the mini ITX to initialise the scope alignment and for the goto functions. An added bonus of using NexRemote is that you can connect up a wireless gamepad to control the scope movement. This makes life very easy during the star align routine, and when centring your object, since you are not tied to the Celestron keypad and its connection to the base of the telescope. I found the joystick gamepad control of the telescope movement much more intuitive (and enjoyable!) than the keypad, and I am not one for playing computer games. After acquiring the evening's data on the mini ITX, I transfer all the files onto a 4 GB USB pen and take that indoors for image processing on a separate desktop computer.
In the few weeks since writing the above, I had a nasty incident that has caused me to think a little more carefully about the observatory electrical system. The supply from the house has an earth leakage circuit breaker for safety, and you really must fit one of these as the last thing we want is for you to be incapacitated in a remote observatory after receiving an electrical shock, and nobody knowing you are in trouble! That wasn't the nasty incident by the way. The nasty incident was some power surges from a failing power supply spiking the mains supply. These spikes successfully took out the motor control board on the Celestron Nexstar 11 GPS telescope, and two complete sets of memory in two different computers! This as you can gather was a very expensive glitch. In order to minimise this potential problem occurring again in the future I have fitted the Celestron "Power Tank" to the telescope [http://www.celestron.com/c2/product.php?CatID=51&ProdID=371], this is basically a 12 volt rechargeable battery, so the scope is now completely isolated from the mains. By the way, the 12-volt connector to the base of the scope can become intermittent due to poor electrical contact, a fact that is well-documented on Mike Swanson's superb site http://www.nexstarsite.com/. I wrapped a single layer of aluminium foil around the outside (earth) part of the connector from the power supply and this seems to have improved things enormously, this is in addition to carefully "splaying" the +12V connector pins on the telescope connector using a small screwdriver. I have also hooked up the computer and monitor to an uninterruptible power supply (UPS) to add an extra bit of buffering protection between the mains and the computer system. Finally I have replaced the ordinary extension leads with Belkin surge protection extension cables. I'm not sure there is a great deal more I can do to try and protect the sensitive electronics from something nasty on the mains supply.
My main (desktop) processing computer is a 2.8GHz homebuilt system with 250Gb of internal hard drives, a further 1.0Tb of external networked hard drive, and 2Gb of RAM, which is located in the house. The main processor is dual core. I also use a dual 19" CRT monitor system so that I can pull all the software processing menus across onto one screen leaving the second screen completely clear for the image I am processing. Again, you don't need an enormously powerful computer for your image processing, but I have found that PhotoShop actions on the 2.8GHz system are carried out very much faster than on the 850MHz system I used previously for over a year. This isn't to say a sub-lGHz system isn't perfectly adequate - I believe it is - but processing an image takes quite a long time as it is and it can be quite frustrating having to wait a while for a simple processing action to be carried out. So the sequence of events on the indoor desktop computer is as follows. Download the previous evening's data from the USB pen into a named and dated file on the desktop computer. Convert the raw 16-bit FITS sub-exposures into FITS floating point IEEE format. Convert these FITS files into RGB (i.e. colour images) using the View/Command Sequence Window in Maxim DL. Manually look through each of the colour sub-exposure images and delete any that have satellite or plane trails, download glitches, or movement glitches. Combine the remaining (good) sub-exposures using SD Mask or Sigma Clip. Lightly crop the outer edge of the combined image and mirror the image (if necessary) in order to get the correct image orientation from the optical system. And that's about it. However, there are lots of variations and combinations you might want to try. My home computer system can be seen in Figure 4.1.
As I suggested originally, you can do all your astronomical work on a reasonable laptop computer, this could be your "observatory" computer and your indoor desktop computer for image processing. You can have a separate "observatory" computer and indoor desktop computer, the system I currently use, and transfer the data between the two using either a USB pen, or even by writing the raw FITS data to a CD/DVD in the observatory using a CD/DVD burner. This latter approach is a good idea, as you will automatically generate a "hard copy" of your night's imaging that won't be easily lost due to some electronic glitch in your computer system. For obvious safety reasons you should always burn your original FITS data onto DVD or CD so that you never lose this hard won data to something as trivial as a hard disk failure. Fine words, but lately I have not actually done this myself! I used to burn all original data to DVD, but with two 0.5Tb external networked hard drives I now pull all my data (processed and unprocessed) off the desktop computer's internal hard drive, and place all that data onto the huge network drive. The idea is to fill the network drive and then remove it from the system and install another network drive. However it is quite possible that at some point I turn "chicken" and spend a month or two burning all the original FITS data onto DVDs. I highly recommend you get into the habit of doing this from the beginning of your imaging career and do not follow the lazy route I have chosen, which I am sure will one day end in tears.
You can actually connect your observatory computer to your indoor desktop computer using either cable or wireless if the distance is not too great. This allows you to use your indoor computer as a remote desktop for your observatory computer so that you can see what is going on in the observatory from the luxury of your study armchair. The only reason I have not bothered to go this extra mile is that I have not automated the dome rotation of my observatory, so every half an hour I have to go outside to rotate the dome so that I am not restricting the FOV of my scope with the dome aperture. If I had automatic dome rotation I would probably remotely control the observatory computer from indoors and do all my imaging from my study with some nice appropriate music playing in the background. In fact that sounds so nice I think I'd better start looking into automatic dome rotation!
Most other computer system configurations for astronomical imaging and processing are variations on the above options. It really comes down to where your telescope is situated, and how far away it is from your home computer, that decides on the configuration that will work best for you.
To summarise my "observatory" and my indoor "image processing" computers are as follows:
The Observatory Computer System
An Epia 1.2GHz motherboard in a black Venus (beautiful!) case. 512 Mb of RAM (single card). 150 Gb single hard disk drive.
CD/DVD writer and single floppy drive. 17" LCD monitor with in-built speakers. Standard keyboard and mouse.
Wireless Gamepad controller for joystick control of the scope movement camera only, not the SXVF-M25C. Windows XP, Maxim DL, CCDInspector.
The main function of the observatory computer is for downloading the imaging CCD camera data, and for autoguiding.
The Indoor Image Processing Computer System
An ASUS motherboard with 2.8GHz dual-core Pentium 4 processor. Silent 500W power supply. 2Gb of RAM (4X512 Mb).
An 8XAGP graphics card with dual monitor control.
2x150Gb internal SCSI hard drives.
Tsunami Dream Case (beautiful again!).
CD/DVD writer and floppy drive.
Wireless keyboard and mouse.
2x19" Hyundai CRT monitors.
Internet access, Windows XP, PhotoShop CS2 plus Noel's actions, Maxim DL, Neat Image, Paint Shop Pro 7, Rus Croman's Gradient Xterminator, and a whole bunch of other processing software both freeware and purchased.
The main function of the indoor computer is to process the astronomical images and print the results on an A1 format inkjet printer [HP Designjet 130].
Always keep in mind that for what you get, computer power is exceptionally cheap. If you have a problem that can be solved by buying an extra computer, this is likely to be the quickest and cheapest way of solving your problem.
My long-term project goal is to create my own micro-WASP http ://www.super wasp .org/ observatory. This will consist of four imaging systems (short focal length) with four large format CCD cameras so that a large area of sky can be imaged at one time. For a Sky 90/SXVF-M25C micro-WASP array consisting of four imagers, an area of sky 4.5 by 6.5 degrees would be covered in each image! It is likely that I will hook each CCD up to its own computer so that I will avoid data conflicts from so much data transfer, and so that there is a certain amount of redundancy available. So, my future observatory is likely to be overrun with computers - at least they will keep me warm in winter.
Was this article helpful?