Recollimating the 16″

Accurate alignment of the optical elements, typically 2 mirrors, of your telescope is known as “collimation”. Proper collimation is the key to success in astroimaging. It was time, probably overdue, for re-collimation of the 16” (400mm) RiDK telescope. I had not collimated the scope since around first light which was now about 1 ½ years prior. My bad on that , but I had not noticed any issues with star quality until very recently. Since I had to do maintenance on the mount, I thought it was a good time.
The RiDK is a corrected Dall-Kirkham Cassegrain telescope or ‘CDK’. ‘Ri’ are the first 2 letters of Riccardi who is the optical engineer for the Italian version of the CDK and manufactured by Officina Stellare. I believe it is basically similar to a Planewave CDK except the corrector lens is closer to the focal plane theoretically reducing chromatic aberration to zero. I chose this design for the main telescope because it’s much easier to collimate than just about any other reflecting telescope! The corrected Dall Kirkham Cassegrain, CDK, or in this particular case RiDK is a 2 mirror telescope with a corrector lens. The primary main mirror is a “prolate ellipsoid” or something in between a spherical mirror and a paraboloid, found in the classical Newtonian telescopes. The secondary mirror is a convex spherical mirror. A doublet lens is placed at the level of the primary, in the center of the optical train, to eliminate distortions in stellar images created by the raw spherical configuration of the secondary and to create a “flat” image through the entire field, in other words perfectly round stars from “corner to corner”. This arrangement is much easier to align than for example the Ritchey Cretien “RC” design which employs a hyperbolic primary and hyperbolic secondary. It’s also easier than Newtonian telescopes in my opinion as usually you have to collimate the secondary mirror as well as the primary. Often the secondary is offset from the center of the primary which hugely complicates things! In the RiDK scope, ease of collimation was confirmed when I first did it a while back. It wasn’t too bad at all right out of the crate. Initially around 6 arc sec off, I was able to reduce that to less than 1 fairly easily with small 1/10-1/8 turns of the primary mirror collimation screws.
The method I have used time and time again is a software based algorithm. CCDWare’s CCD Inspector or ‘CCDI’ is the application. In order for this to work well I think a few things have to be true:
1) Collimation is “in the ball park” which is difficult to quantify but I would say if you have confirmed with a Cheshire eyepiece or similar it looks right and you are ready for star testing then you’re good. I’m going to guess that means it’s not more than 15 arc seconds off at most but that number is a guess
2) Only one mirror needs adjusting, either the primary or secondary but not both
3) Your telescope is not fast, i.e F/5 or lower
As an example a while back I tried unsuccessfully to collimate my Tak 180 using this method alone. This is a unique optical system with an offset secondary and a super fast F/2.8 focal ratio! So 2 of the 3 criteria were not met.
So how does this work? Before that I just want to point out the great thing about this method is that you are testing on an imaged star which is what the goal is anyway. I think it’s harder to do this visually with an eyepiece because the small changes are difficult to be certain of. CCDI gives you an actual number in arc seconds so you know exactly what your corrections are accomplishing quantitatively. Plus you can use the camera set up you are actually imaging with.
For best results you will need the following:
1) Ability to place a star in the true center of the image field. Most camera acquisition programs will have the option of adding a cross hair to the screen. This is necessary.
2) Knowledge of the focus position of your set up such that you can reach a defocused position yielding a star image which is a minimum of 200 pixels in diameter
3) At least average seeing. This is going to limit how much of a correction can be made.

First you will need to install CCD inspector here. There is a free trial period of I believe about 30 days

Open the program. Go to the settings tab and click on “default image properties”

Settings > Default image properties

Enter image scale and pixel size of your imager. Saturation level for most applications is not critical.

Here you will enter your main imager’s specs as above: image scale etc.

Next you will need to tell the program what camera control software you are going to use. If you are not using either CCDSoft or Maxim then click on ‘Generic’. It does not matter what program you use as long as you indicate to CCDI where the files are going to be saved. This is the key.

Use the “Generic” setting if not using either CCDSoft or Maxim

When you click on Generic, Windows explorer will open and you will be able to choose where CCDI needs to look for the saved image files. I would recommend doing this first in your camera program where you will need to tell it to autosave the files and select the folder where you want to save them. If you are using The Sky X autosave, my suggestion is to turn off the option to create a new folder for each date. This will be confusing to CCDI.

Clicking on “generic” pulls up the Windows explorer to select folder for “real time” images

We are now ready to begin the process! First slew to a medium bright to very bright star. This will take some trial and error to find the right combination of exposure time and magnitude for your optics. CCDI will tell you if you are on the right track. You will get an error message “star distorted or too bright” or “star too dim”. Remember to select “autodark” in your capture program and typically bin 1×1. For my 16″ F/7 scope I used a 4+ magnitude star for this with an exposure of 2.5 seconds to get consistent readings. You will need to get consistent readings without error messages to do this accurately. Also you might need to subframe your image depending on the size of your sensor. For example I am using a full frame 16803 sensor with 4096 x 4096 pixels. In The Sky X, I used the subframe option with 1/4 frame selected. This resulted in 1000 or so pixel square which worked well. If you have a medium sensor you might not need to use this feature. Just see what happens when you take several exposures. If , for example, you get 2 or 3 readings, then a bunch of error messages with “too bright”, then “too dim”, then “too bright” etc., you should definitely try subframing.

In The Sky X program you can select “subframe” and “size” . Note also “autosave” is checked. Once you select “size” you are brought to a screen where you can specify how small you want the frame down to 1/4 of the full size

Next confirm you are at or near focus and the star is dead center in the frame. I use the “closed loop slew” feature in The Sky X which will plate solve the image and move the star to the center automatically. Use a hand paddle if that’s what you have. You just need to make sure it’s centered.

Star is centered to begin the process. The live collimation viewer is seen on the left. The number is in arc sec.

Once it’s centered then you need to defocus the star image. Move to a focus position where the defocused image is no less than 200 pixel diameter. Now I would point out there is nothing wrong with defocusing the star first and centering that manually. You can continue to do that without having to refocus etc., but I decided to make use of the closed loop slew feature which requires focused stars in order to plate solve the image. Once the star is centered, go to CCDI program and open the collimation viewer.

Go to “collimation” > “Defocused Star Collimation Viewer”

 

You can also select “images to average” also under the “Real-Time” tab. I usually do just one to make sure the readings are consistent. The exposures are typically short.

Next take an exposure. You should have the collimation viewer open on your screen (see above). If you don’t get a reading or an error message, i.e. nothing happens in the viewer, that probably means you don’t have the file saving set up right or perhaps you have inadvertently selected multiple images to average. Again, make sure the folder you select in CCDI (click on the “generic” tab under camera control in settings) is the same as the the folder your camera control software is autosaving to.

Take several exposures to make sure the readings you are getting are consistent, i.e. the arrow is pointing in roughly the same direction every time and the collimation number in arc seconds is also about the same. For my first run I was getting 7-10 arc seconds in the same direction.

The collimation viewer is seen to the left reading 7.1. This was my initial starting point

Once you get several readings you are ready to make a correction. The goal is to tilt the primary mirror in the direction of the arrow. If you are working with a secondary mirror such as in an RC scope (e.g Astrotech model etc)  you might have to move the secondary mirror OPPOSITE the arrow. You will have to trial and error it to see which collimation screws move the mirror in which direction. Label the screws as you are doing it. Start with very small movements of the screw, not more than maybe 1/16 of a turn.

For my RiDK 16″, there is a collimation wrench for the primary screws. There are 3 set screws, one with each collimation screw (above the collimation screw and to the right slightly) which are loosened prior to collimation. One full turn of a collimation screw moves the mirror 1mm!)

After you make the small screw turn(s), take another image and you should see the defocused star move in the direction of the arrow if you have figured out your screws’ directions properly. Next what I did was refocused the star, recentered it, then defocused again. Take your image. You should see an improvement in the arc sec number. Perhaps a more efficient way of doing this if you are using The Sky X is to calibrate your main imager as an autoguider. Then you can right click on the center of the defocused star and it will recenter the star without you having to refocus, closed loop slew etc. I did not do that, but it really did not take more than 30 seconds to go through the routine I did.

After several iterations of this you should ideally be getting consistent readings at least below 5 arc sec, depending on seeing. In average seeing you should be able to get perhaps 3 or less. At a certain point you will start to see the arrow acting “weird” meaning it will point up after one exposure, then point down after a second exposure and then point to the side, etc. This means you are done!

After several iterations we got consistent readings less than around 2 arc sec but the arrow was flipping around so we are done for this session. Seeing was around 2-2.5 arc sec

And that is how I recollimated my RiDK , and also how you can collimate your reflector!

Thanks for reading!

Dr Dave

 

Roof control Redux

It’s been nearly a year since that day last July 7th when everything changed forever and I was finally able to open and close the observatory roof remotely! You can read that post how excited I was about that because from then on I had pretty much unlimited access to the sky since I did not have to physically be there (hence the term “remote”). A few weeks ago I went about my usual routine, about to start an imaging session, clicked on the “open roof” button, watched the progress bar indicating everything was fine, all systems operating normally, but one thing was different. The roof was clearly NOT OPENING! The system which was fairly easy to install and operate had failed after about 9 months. For the 2 weeks following I spent hours troubleshooting it. I started from the beginning, testing the board, checking the wire continuity, metering everything, even replacing all of the wiring I had going from the 12 volt relays to the board. For whatever reason when I tested the board, only the open relay would light up but very briefly. I even replaced the relays.This was without the roof motor connected. I tried a spare board, tried different software versions, even tried 2 different windows operating systems. Nothing. This really made no sense at all. I checked and rechecked the connections. There was no rhyme or reason to it. Unfortunately support was completely non existent. I was told to update the software which I did but no answers to any of the questions regarding the steps I had already taken. Frustrated to no end I began to realize that this was not only about the system failing but the fact that all indications were that it was working properly and zero support for troubleshooting this. If this happened once it would surely happen again even if I was able to sort it out which probably would have occurred eventually. The only option was to at least look at other “more robust” options for roof control if they existed. The most important function is to open your observatory reliably otherwise you can’t use it! Despite the early success I had with the Foster System board it looked like that was not going to be the long term answer.

I looked at it this way. Thousands of people open and close their garage doors every day from a phone app! Never fails. Those of you who are reading this probably have a garage door and never for a second wondered if it will open or not. A roll off roof is basically a heavy duty garage door. There has to be by now a nearly fullproof system for its operation. I understand any time you have a control system which requires usb to function and Windows you can only expect so much, but nevertheless I posted this question on the Cloudy Nights astronomy board and I received several responses from folks who apparently had success with a system called “Skyroof” which I hadn’t yet heard of. It’s produced by InteractiveAstronomy.com and apparently it is based on arduino control board technology. It looked interesting. More expensive than the Foster control board by around $100, so not too much more. It looked like a much more durable circuit board although I am no expert on these. Other features which were attractive were the fact that I could direct connect the motor wires to the board without additional relays, the board was a direct usb connection rather than serial to usb, and powered by usb, so it didn’t require a secondary 12 volt power supply. No cumbersome firmware updates or bootloader programs we had to use. I was going to give it some thought before purchasing yet another roof control system and figured I would try just a couple of other things to get the Foster board to work but actually got a phone call from Jim over at Interactive Astronomy and he answered the questions I had about his SkyRoof system over the phone that same day! He also emailed detailed instructions on how to wire the board. I asked him about their support system and he basically confirmed they responded by email or phone that same day, even on the weekend. No support forums required. Ok I’m sold! I bought the Skyroof kit and it arrived 2 days later. Free shipping! I went up to the observatory that weekend and in 15 minutes the roof was operational once again! All I had to do was connect the open and close motor wires to the board, no additional relays, as well as the common wire and on the other end connect the open and close roof sensors. The software was simple to install and everything worked pretty much right away. We’re back in business!

So in summary, this is the 3rd roof control system I have used. The first one which was installed as a bundled product by Backyard Observatory when the structure was built was M1OASYS ,which is basically the M1 Elk Gold home security system adapted to roof control as an afterthought. It was like trying to kill a fly with an assault weapon. Nearly 2 years I could never get it off the ground. I finally had someone who is using it at Deep Sky West hosting facility in Rowe NM confirm it is not a product for astronomy. They use it there because they have a lot of scopes and benefit from the security features.

The second system was the Foster Astro MC control board which worked great for 9 months but then as I have been detailing, stopped working. It is similar to the Skyroof system I have just started using but different in a few key ways as follows:

Foster                                                                                 Skyroof

$250-300 not including shipping                                          $395 with free shipping

12 volt power supply required                                       Direct usb which powers the board

Serial to usb adaptor required                                      No additional adaptors or drivers

Firmware updates/ bootloader app                             No regular firmware updates

*Warranty not documented                                            *Clearly stated 1year warranty

*Support essentially non existent                                  *Support by phone/email same day

Documentation not clear to inexperienced                   Documentation clear to anyone

 

Skyroof supports many additional features such as a Sky X driver as well as text message alerts if desired. Many standard features are suported on both systems such as emergency shut down for weather related and other conditions but for the Sky Roof these are activated by simply checking a box in the software set up. For Foster, it appears much more complex to set these up. Bottom line, if you are like me and “I’m not a mechanic Jim. I’m just an old country doctor!” then definitely Sky Roof would be for you, but I am not yet going to wax poetic about this new system right now since I have learned there is no such thing as a sure thing in this business!

Thanks for reading!

Dr Dave

The Sky Roof control board is basically an arduino microcontroller about 3″ x 5″. The grey cable labeled ‘A’ is the cable coming out of the roof motor to the panel. Only 3 wires inside that cable are needed to connect to the board: an ‘open’ wire, a ‘close’ wire and a yellow ground wire. You can figure out which is the open wire by connecting the wire directly to the yellow ground wire first and seeing if the roof opens. I did those tests before using the Foster board last year and lableed the open wire with green zip tie and close wire with a red zip tie (‘B’ and ‘C’) Yellow wire is the ground wire ‘D’. E is the blue cable which contains the wires for the open and close roof sensors. Those are connected on the bottom. The diagrams that were sent with the board are very clear and easy to follow.

 

Sky roof control panel and settings menu. Notice the options under “actions/alerts” in the settings menu give you several emergency features which are enabled just by clicking on a box!

 

SkyRoof control panel showing roof opening

 

 

Collimating the Tak 180ED- what I’ve learned

The very mention of the word “collimation” causes much anxiety amongst amateur astronomers. A google websearch for “collimation” produces literally hundreds of detailed accounts of blood and sweat with many a frustrated astroimager brought to the brink of despair. This is no such tale. In fact, like many other things in life, when I finally decided to take the plunge as it were, I found things a lot different than what they were “reported” to be. I decided to write this entry as a series of questions and answers as though you the reader were right here interviewing me. I felt that would cover everything sufficiently for a single blog post. I hope anyone who owns one of these or is planning to acquire one will find it useful

What is collimation?: Collimation is simply the process by which you align the components in your optical system to bring the light from the object to it’s best focus. Typically you’re talking about a reflector telescope which employs 2 mirrors of different geometrical configurations that are spaced apart varying distances inside of the optical tube. These mirrors have to be aligned in such a way as to enable all of the light to arrive at a single point.

What is unique about the Tak 180 ED?:

Tak 180 ED

The Tak 180ED employs a hyperbolic primary mirror, a flat secondary mirror and a doublet “corrector lens” to produce a wide flat field corrected for coma and spherical aberration. The focal length is extremely short, only 500mm (F 2.8) for a mirror 7 inches in diameter! This makes it more challenging to focus and collimate as the depth of field is extremely short. The secondary is offset from the primary center to enable more illumination of the field. It differs from the classic Newtonian telescope in that the primary of the newtonian is parabolic, not hyperbolic, rendering the field sizes smaller typically for similar aperture.

Why is there such a mystique behind collimating these telescopes?: I’m not sure about there being a “mystique” other than many “tales” that are out on the web regarding how difficult many people have found it. No question you do have to pay attention to details but that is no different than many other things we do in this hobby.

Was there anything different about your telescope that came into play during the collimating process?: Actually there was. I acquired this telescope from a friend who had the stock focuser removed and replaced with an extremely heavy duty focuser designed by Paul Van Slyke. Unfortunately these are no longer made but lucky for me I do have one on this scope and it is built like a tank. I was very concerned about even touching it for fear of causing some misalignment problem but clearly he demonstrated how his focuser did not alter the steps required for collimation. The corrector lens was on there pretty firmly and I did have to remove it carefully with a strap wrench but it was actually fairly easy to do.

Was there any special equipment, laser collimators or other tools you had to acquire?: That was actually the biggest surprise to me. Perhaps this was unique for the 180 but luckily my friend saved all of the collimation tools provided by Takahashi for this scope and I did not need anything other than an open 17mm wrench for the primary screws.

The items above were included and this diagram in the instruction manual shows the precise order of assembly. The “collimating tube” has a cross hair made of fine nylon threads at 90 degrees to each other that you tape down into slots in the tube. The eyepiece is a pinhole with a side port.

Corrector lens is simply unscrewed from the focuser. I used a strap wrench which was pretty simple. Not a ton of force was required

The lens is carefully removed.

Collimation tools are then assembled as in the diagram above and the assembly is screwed into the the focuser

 

Did you get any assistance from anyone during this process?: That is really the most important thing when doing anything for the first time and with any uncertainty, ALWAYS get help from someone more experienced. I contacted Texas Nautical Takahashi or “land sea and sky” like everyone else who has a question about their Tak scope in the US. There is a fellow named Fred who is the expert in these matters and has been there for years, but unfortunately he was out of the office. I spoke to one of his colleagues who basically told me to read the manual and follow the directions! So that’s exactly what I did! I found the manual to be really well written and directions easy to follow.

So how did this go down once you were told to “read the directions”?: Actually a very interesting thing I discovered before that happened occurred when I tried to collimate it by only adjusting the the secondary with the CCD inspector software, assuming collimation was close. This is my “goto” method for collimation which I have done for a variety of scopes and been successful over the past several years. What I realized was that even after the software seemed to indicate good collimation it didn’t look right. The secondary was offset as expected but the area surrounding the secondary was not concentric. I knew then that it wasn’t right.

By all accounts a collimation error of 3.7 arc sec is excellent. However if you look carefully you can see that distance “A” is shorter than “B” which is not right

After that I realized that I had to start from the beginning, reading the directions. When I assembled the collimation tools and screwed that in to the focuser I could see right away that the collimation was off. There is a dot on the secondary and a circle on the primary to assist you in making your adjustments. I really don’t have anything else special to report about the specifics other than I followed the directions, starting with adjusting the secondary to the focuser which involved both tilting the secondary and rotating it gradually until the spot was lined up, and then adjusting the primary to the secondary.

This is the initial view after I tried using CCD inspector, unsuccessfully, looking through the collimation set up. It’s WAY off! “A” is the spot on the secondary. “B” is the cross hair in the collimation tube. “C” is the circle on the primary

 

Can you provide any tips or any other important things you learned besides “read the directions”?: Yes certainly. I don’t mean to oversimplify things but honestly it really boiled down to exactly that. Here are a few tips you might find useful:

1. The directions: you may or may not have access to these especially if purchasing the scope used so please feel free to download the pdf file here
2. Don’t try to reinvent the wheel as it were. Ask an expert in the field.
3. Don’t be afraid to remove the corrector! It just screws on to the focuser!
4.  CCD inspector is a very valuable tool for collimation but I believe you need to be cautious relying on it exclusively for fast optical setups like this.
5. I recommend using a flat panel or other light box set up for aligning the secondary with the telescope mounted and horizontal to the ground. It was fairly easy to make the collimation adjustments while looking in the eyepiece and adjusting the secondary. It is really just trial and error. Small adjustments either rotating the secondary or turning the “push” screws. A “push” screw is just a screw that pushes against the secondary or primary cell to tilt it. I did not have to adjust the axial position of the secondary front to back at all.
6. The primary collimation screws are pretty coarse and difficult to separate and work on each one individually. I did not have the wrench provided by Tak so the one I used was perhaps fatter. You will need a 17mm open wrench and a screwdriver. There are 3 components. At the base is the locking nut and furthest back is the locking screw. The middle screw is what tilts the mirror. You have to loosen everything and then I found that turning the middle screw by hand was the most effective. Again you really have to trial and error it to see what works. Once it is lined up I tightened the locking nut by hand and a little more with the 17mm open wrench but then was VERY careful about tightening the back locking screw with the screwdriver because a couple of times I did that the middle screw was turned and threw the collimation off.
   

   3 components to the primary collimation screws. “A” is the locking nut. “C” is the locking screw and “B” is your push screw. The locking nut and locking screw have to be loosened first, then adjust the push screw to tilt the mirror. After you’re done lock the nut and screw

7. Once the secondary is aligned, put the scope vertical pointing to the sky and do the primary collimation in that position.

Ok. So can we see any results of what you did?: Of course, we’re all waiting for that! I confirmed the visual appearance to be correct based on the manual diagram, then tested the alignment on the Omega Centauri cluster. I believe you will agree that collimation is very good now!

Looks pretty spot on to me! Notice now the distances “A” and “B” are the same. The secondary is offset by design as described above. Compare to the initial view earlier in this post

 

This is the result with CCDI before redoing collimation from the beginning. Note the oblong stars on the left

Nearly the same star field in the same region after collimation was completed! Much improved star images throughout.

 

Omega Centauri before complete collimation. Notice the misshapen stars to the left of the image and it appears to be “soft” overall.

 

Omega Centauri after collimation is complete! I think you can agree it is a sharper image with very good stars corner to corner!

 

Any final comments about the process?: Listen to the experts. In this case they said read and follow the directions and use the tools provided, which obviously worked! Don’t be afraid to remove parts from your scope if they need to be. Most importantly, don’t believe the horror stories that are out there 🙂

Thanks for reading!

Dr Dave

 

 

 

 

 

 

Omega Glory

Tightly bound spherical collections of stars known as “globular clusters” exist in most galaxies, orbiting around their cores. Our own Milky Way galaxy has about 150 or so. Most astronomy enthusiasts who live in the Northern Hemisphere are quite familiar with the “Great Hercules Cluster” M13 located in the constellation Hercules. An impressive sight in any telescope, the Hercules cluster has a few hundred thousand stars and measures 145 light years in diameter. It is faintly visible to the naked eye. M13 is by far the largest globular cluster visible from the Northern Hemisphere, but it isn’t the largest in our galaxy. That title belongs to Omega Centauri, or NGC 5139, in the constellation Centaurus. Unfortunately Omega Centauri is visible “mainly” from the Southern Hemisphere. The reason I say “mainly” is that last year I accidentally came across it while casually sweeping the southern horizon with a pair of binoculars from the deck of the house up at Orion’s Belt Observatory! It turns out we actually do have a small 2 hour window of visibility from this location. Omega Centauri is quite visible with the naked eye from a dark site and looks to be about the same diameter as the full Moon! By contrast to M13, Omega has several million stars packed so densely they are on average only 0.1 light years apart from each other! At 180 light year diameter it is only slightly larger than M13 but has many times the number of stars.

While working on the “pier 2” equipment, I decided to take a couple of test images of the cluster during the brief window of visibility just to see what I could see! Of course this is not an easy target from this location, with a maximum elevation of 7-8 degrees above the horizon! The observatory has panels on the south wall that can be dropped down to access the horizon. This is exactly the situation they were designed for! I pointed the Tak 180 to Omega and tried a few 3 minute exposures. The results were much better than anticipated. The dark skies certainly help. On the screen was the glorious Omega Centauri cluster! The image I thought was pretty clean and stars well resolved for just a single exposure. The Tak 180 is a known bear to collimate and I did the best I could just for this test but it will need some work..a subject for a future post. Anyway I felt this was going to be a definite legitimate target in the future based on this preliminary test!

Screen shot from the “Stellarium” program showing the location currently of Omega Centauri (broken square icon due South) at its maximum elevation during transit of a wopping 7 degrees above the horizon! Some of the brighter stars in the constellation Centaurus are seen to the left

 

NGC 5139 or Omega Centauri is the largest globular cluster in our galaxy! This is a single raw image taken from Orion’s Belt Remote Observatory using a Tak 180ED on a Paramount MX+. 3 minute exposure. Canon 60Da camera. I did a quick collimation with CCD inspector and it’s actually pretty decent but if you look to the left side of the image the stars are a little elongated. I was quite thrilled with this quick test and can certainly add the cluster to the list of obtainable targets from this location!

Thanks for reading!

Dr Dave

News from Orion’s Belt Remote Observatory

We installed a “flat panel” for the second pier (Pier 2). “Flat field correction” is another facet of image calibration in astrophotography and attempts to remove artifacts and other aberrations that occur as light travels though your specific optical train. These could be anything from dust to stray light or shadows occurring due to various equipment components and set-ups. The goal is to achieve a uniform field across the target image. The “flat field” is itself an image of a uniformly lit area which when exposure is adjusted properly, yields a “flat image” showing the artifacts present in your set up. When you subtract this image from the actual astronomical image, the result will be (hopefully) a clean image of your galaxy, nebula etc. Several options exist for a uniformly lit target to produce adequate flat frames. For pier 1 up here at the observatory I use what are called ‘sky flats’. At dusk or dawn if it is clear there is a region of sky where brightness is uniform. The 16″ scope routinely takes these sky flats before and after the imaging session. They work very well- when it’s clear. The other potential issue is that you have to keep changing exposure times as the sky continues to darken or brighten during dusk or dawn in order to maintain the same peak intensity of your flat, typically 40-70% of the saturation point of your sensor. While this exposure adjustment happens automatically with the control software for Pier 1, at some point it either becomes too dark or too light in the sky to continue so you will always be limited with the numbers of flats you can acquire using the sky flat method.

Flat panel which is placed in this case in front of the FSQ106 refractor for obtaining flat frames

Enter the “flat panel”! This is a really neat device which is a uniformly lit artificial light panel where you can adjust the brightness until you have reached the desired level for your equipment. Then you can fire away and take as many flats as you like whenever you want! I decided to try this out for pier 2 and I have to say it works very nicely. Now it may be difficult to automate this feature because the flats in this case are obtained when the telescope is in the parked position and the imaging session is completed. However it is easily carried out through a PC software interface and thus can be obtained remotely!

“Spika flat fielder” is available through a company called All Pro Software and is very easy to set up and use. I purchased their standard panel which is about 15″ and can be used for scopes up to 12″ in diameter. Larger ones can be purchased for bigger scopes if needed. I modified an artists easel to accommodate the panel and this enables me to move the panel around if needed. A standard AC power converter enables a usb connection to your PC and also powers the panel. The software control is quite simple. An example of a flat frame is shown below.

The flat panel is mounted onto an easel. The entire assembly is fairly light weight and can be easily positioned in front of your telescope

 

This is all there is in the control panel! Just click up or down on the panel brightness. 

 

An example of a flat frame. The round spheres are dust particles in the optical train. Notice also the darker “shadows” in the corners. This frame is subtracted from the target image to arrive at a uniform result without the artifacts present in your system 

 

Happy Flat Fielding!

Thanks for reading!

Dr Dave

A Bubble in Space

It’s been about a year and a half now since the 16” scope up at Orion’s Belt saw first light. Now, finally, the very first imaging project with that telescope is completed! The “Bubble” nebula, or NGC 7635, it’s official designation in the “New General Catalog” of deep space objects, is a true bubble in space, about 7000 light years from us in the constellation Cassiopeia! Not science fiction! The bubble is created by “stellar wind” or gas ejected from the upper atmosphere of a very hot centrally located star in the nebula. Surrounding the bubble is a huge cloud of hydrogen gas. Since all of you who have been reading these posts are now experts in spectroscopy 😊 you know that hydrogen gas after it absorbs radiation can then emit that radiation back into space in the “hydrogen alpha” wavelength of light which is the visible red portion of the spectrum.

This data was a total bear to process and I almost bailed on the whole thing when I realized that the bubble was hidden in a dense star field! (see below). We forget how many stars are out there. 250 billion in our galaxy alone! It really seemed to me that they were all congregated right there in one frame. The processing challenge was to somehow remove the stars to bring out the nebulosity. Folks, if you are going to image a nebula in a dense star field my suggestion is to use narrow band filters all the way. I felt I could get the natural color of the nebula using just the hydrogen alpha filter, which is narrow band ( in other words just allows the light in the H alpha region to pass through), but only one channel, combined with the regular band width red-green-blue filters. Problem is the regular band width filters do not filter out the stars! The details of the processing are beyond the scope of this one post but I tried to show the essence of what was done in the images shown here. Star removal is still not a perfect “science”. I have yet to see one method that does a complete job with no artifacts, but the method I used seemed to work pretty well. The software Pixinsight has a number of processes that in combination can achieve the desired result. The “star mask process” can be applied repeatedly to “mask” different sized stars which can then be removed by applying the mask back to the image as a “defect map” (another process) to do the actual removal. Anyway it took weeks of trial and error to get it right enough where the residual artifacts were manageable!

If you click on the thumbnail under “My Astro images” on the right side of this blog (scroll down a bit) you will go to the full resolution version.

This is the Hydrogen alpha version. Notice it is black and white. You have to combine this with one or more other channels for color. The detail in the nebula is able to be detected with this filter since it is only letting the narrow 3nm width of light through in the H alpha region of the spectrum. Most of the star field is filtered out however to achieve the red color you have to combine this with another channel, typically red ,but in my case it was the regular band width red channel which is full of stars!

 

This is the standard “tricolor” RGB image! This is how we obtain color images in general with astroimaging, combining individual red, green and blue filtered images. Amazing to see how the nebula is really hidden in this image behind all of these stars! Also notice the slightly pink color of the nebula which is from the blue channel

 

Going through the star removal process for the RGB image above, then combining that with the Hydrogen alpha image (first one above) I was able to get to this workable solution which then had to be tweaked some to arrive at the final result many hours later. I did look at trying to sharpen things up a bit as it does get a little soft at full res, but that wasn’t possible due to the processing artifact in the background. Overall I think this was a decent salvage effort!

 

Thanks for reading!

Dr Dave

 

Behind the lines

This was my second spectrum accepted into the BeSS database. It is a high resolution spectrum I just obtained of the H alpha region of the Be star (B emission) HD43544. This is a 6^th magnitude star in the well known Winter constellation Canis Major. Captured with the Lhires III spectrograph, C14, Atik 460EX, Paramount ME from Las Cruces, NM. Many B spectral type stars exhibit emission lines rather than absorption (meaning they spike upward , not downward ). These are actively researched now because it is believed that several, possibly most B stars, are actually part of close binary systems with a small “type O subdwarf” ,or white dwarf ( or in some cases neutron star or even black hole!) accounting for mass transfer between the 2 stars which gives rise to the rapid rotation of the B star and formation of a gaseous disc surrounding the star. This is the reason for the emission feature. Close analysis of the emission profile changes over time in the HA region or other regions can yield a solution for orbital parameters of a binary system, thus proving they exist! The peaked nature of the Ha emission is dependent on line of sight from earth as shown in this schematic:

Position ‘A’ viewed from Earth would be looking directly ‘above’ the star where we see only an emission peak. Position ‘B’ is an oblique view where we would see 2 peaks with central dip depending on degree of obliquity but this “dip” gradually increases in depth toward the “C” position where we see the full absorption component (peak pointing down) at the H alpha 6563 angstrom point when we are looking at the star edge on!

I think one of the compelling things about amateur spectroscopy is that this kind of data that I have shown here, despite being obtained with amateur equipment, is still highly sought after by professional researchers. Once these spectra are “validated” by the BeSS administrators then they are considered accurate enough to be used in astrophysical research. Remember that most astronomers do not have daily access to equipment like we do! They have to vie for telescope time with many others and perhaps they will get a day or a few days once every year or more, if they are lucky, on one of the big telescopes either on Earth or in space but this resource that we provide enables researchers to acquire data any time anywhere at will! Spectroscopy or the analysis of light from stars and other objects in space is the way that we have figured out how stars work and evolve over time as well as the large scale structure of the universe. It is also a very viable way that we amateurs, you and me , can still participate and be relevant in new discoveries out there in our universe! For more information on astronomical spectroscopy you can read my primer that I wrote for ‘Reflector” magazine in this post

Thanks for reading!

Dr Dave