 Today I took some shots of the Sun while it’s not covered by clouds. To really see details on surface and solar prominence, special H-a (Hydrogen Alpha) telescope is needed. Most affordable one is Coronado P.S.T. (Personal solar telescope). Unfortunately it doesn’t allow to take prime focus photographs because of extremely short focus.
It is possible to modify a Meade barlow to achieve focus with DSLRs and I have done so. However I wasn’t happy with image quality and ended up using positive projection though Baader Hyperion Zoom Eyepiece. After some experimenting I ended up shooting at 20mm with resolution of about 1.5 arcseconds per pixel which is roughly equivalent to 880mm telephoto lens.
I’d have to say that H-a filter does eat up a lot of sunlight. At ISO400 my shutter speed was mere 1/125s. I had to take the additional H-alpha etalon off the telescope because despite providing better picture for visual observing it ate up even more light, bringing shutter speed down to 1/25s, making it very hard to focus and degrading image quality overall.
Here’s the photo of my setup.
 Messier 3 (NGC5272) After several months of futile attempts to eliminate flexure hidden somewhere in my new setup I decided to give up and go with off-axis guider setup instead. I had Orion Deluxe Off-Axis Guider laying around gathering dust for quite a while already and I finally got around to using it as outlined in the previous post. As I was shooting from my front-yard I decided to pick an easier target to test out the new setup. Here’s how it came out. It definitely can (and will) be improved upon. M3 was discovered by Messier in 1764. It is one of the brightest and largest globulars, containing aroudn 500,000 stars at a about 34,000 light-years away from Earth.
Setting up off-axis guider for the first time can be tricky. It certainly was so for me. I spent several hours in the dark trying to make it work with no results to show for it and ended up giving up.
The reason it is hard is that guide-camera has to be precisely in focus for it to stand a chance of seeing guidestars. If it is slightly out of focus, what little light was there would be smeared all over the sensor and become completely undetectable. Trying to find focus in the night is a futile endeavor because most guide stars that make it in the guide-cam though average f/10 scope are so faint that 2 second exposures are required to see them. Worse yet, most prism positions will yield frames with no detectable guidestars at all.
So it pays to be prepared and roughly focus your guide-cam during daylight. Setup your telescope on a mount but don’t power it on. The mount should not move. Using clutches point the telescope at something far enough that it can be focused on but not too far. Ideally it should have sharp easily detectable pattern. In my case I used a broom. Attach your camera to the scope and focus it on the object. Since you are doing this in daylight, use shortest exposure your camera supports. I used 0.001s on my Orion G3 Monochrome.
Now you can adjust your off-axis guider to focus the guide camera. Don’t forget to adjust the prism tilt screw if your guider has one.
This will give your a rough focus that would be good enough for the guide-cam to detect brightest stars (ones that you can see with unaided eye) on exposures around 0.5s. Find such bright star with your telescope and focus your imaging camera precisely using a Bachtinov mask. Notice where your guide-cam prism is respective to the picture and move the mount slowly in that direction until you see the star in your guide-camera. It should be bright enough for Bachtinov pattern to be visible. Use it to precisely focus your guide camera.
You are now set. Find your target, setup the guider and shot away.
Several things to keep in mind:
- Finding a nice guide star can be tough. Use longer exposures on the guide-cam (1.5s..3s) and rotate the prism slowly.
- Calibrate your guiding software each time. Even slight adjustment of the prism will mess up RA/Dec orientation.
- Remember nice calm graphs you had when guiding through 162mm FL scope with RMS of less than 0.1? You can forget about them. Graphs will jump all over the place. But worry not – there will be very little blur and definitely zero drift. I used Orion G3 Monochrome Imager as a guide camera and Canon 5D Mark II as imaging camera on 2800mm FL scope. G3 pixel is 1.3X larger than pixel in the Canon. This means that with 2 RMS you will have 2..3 pixel blur on the image. It is quite good. Remember that there’s still atmospheric seeing that will still blur your image somewhat regardless of the quality of your guiding.
In the past I did some astrophotography with Canon 300mm lens (NGC6960, NGC6888, Andromeda Galaxy). There were some minor problems though:
- Canon 300mm F/2.8L is a huge lens and with tripod shoe being the only point of support point it easily shifts around and can mess up the picture.
- It is impossible to use any kind of standard filters with this lens as it is. So even something as simple as light pollution filter is not an option.
I’ve tried to address both of these problems and it seems like I’ve succeeded.
1) After some tinkering in Google SketchUp I came up with a model of a part that would complement ADM Accessories vixen camera mount and secure the 140mm lens shade to the vixen dovetail. The model is freely available on shapeways.com. Perhaps it is a bit thicker than necessary, but better safe than sorry.
 …
…
2) I got my hands on 52mm drop-in filter holder for Canon 300mm lens and modded it with a adapter ring so that it will take 2″ astronomical filters (corresponds to 48mm)
Now I just need a clear dark night to test this out. If the filter experiment works well, the grand plan is to use H-alpha, OIII and SII filters and then combine the images to achive effect similar to Hubble Space Telescope.
You can download the original 3D model here.
Standard Canon 5D Mark II Li-Ion battery pack (LP-E6) has 1800mAh capacity at 7.2V. This translates into roughly 13Wh of stored power. When shooting 10 minute exposures, the camera draws 3W of power on average. When shorter exposures are used, power usage goes up since more power is required every time image is read from the sesor, processed and written to the memory card. This translates into roughtly 4 hours of shooting before the battery goes out.
This may seem like a lot. However if you take a 90 minute trip (one way and not counding packing/unpacking) to a dark location, you may want to shoot for the whole night. Having a spare battery may of may not be enough. There is a nice Hanken ACK-E6 Replacement AC Power Adapter that costs under $20 that can be used to power your Canon 5D Mark II DSLR. The down side is that it runs off 110V AC. Initially I ran it off inverter plugged into my main 60Ah deep-cycle battery. However inverter draws around 4Wh by itself. Small things add up and when you add heaters to the equation, even 66Ah battery may not last until dawn.
With no 12V power solution seemingly available from Canon or other companies, I’ve decided to put one together myself. I used the battery connector pack from the Hanken adapter, 12V car power plug and 25W adjusable switching regulator to put in between then. 25W should be enough to account even for peak power usage when camera processes and stores taken image.
While switching power adapters are effecient, the downside is that they produce slight high-frequency noise in the output power. I was afraid that this noise may adversely affect image quality so I decided to run some tests. I’ve put the camera into dark room, covered the lens and set it to shoot 6 10 minute darks with 1 minute delay in between to compare noise level when running on battery and on 12V switching power supply. I then used ImageMagick to compute image histogram and analyze frequency components of the image to see if PSU introduced some regular noise patterns.
Initially I tuned the regulator to output 8V just like the Hanken. However after I ran the dark tests it turned out that the camera produced measurably more noise when running on PSU versus the battery. There weren’t any regular pattern to the extra noise. There was just more of it.
I tried to lower the output voltate to 7.2V to match the battery, but the camera wouldn’t power on at all. I kept gradually increasing the voltage until the camera came to life at 7.4V. I did another set of measurements and found that noise amount and patterns didn’t have any detectable differences between PSU and battery shots.
So I can conclude that using switching regulator doesn’t introduce detectable amount of noise into long exposure photography with Canon 5D Mark II. In the initial test excess voltage from the PSU caused additional heating of the sensor that resulted in more noise. It should be noted that even in that case the amount was minor: 1.3% average frame luminance versus 1.14% in battery and 7.4 volt PSU shots. In all likelyhood it will not matter for daytime photography and have only minor impact on long-exposure astrophotography.
Fast Fourier Transform is a nice way to analyze images. ImageMagick can use fftw library to run FFT on images. Unfortunately for Windows users, this capability is not compiled into the program as of version 6.7.6.
If you try to run something like convert source.png -fft source_fft.png, you will get the following error message: convert: delegate library support not built-in ‘logo.gif’ (FTTW) @ fourier.c/ForwardFourierTransformImage/611 (or 640, or some other number depending on your software version)
Users are advised to recompile from source. However if you didn’t take computer science classes, this may prove to be an impossible task. I had to spend a few hours getting everything to compile on Windows.
To save other people the trouble I’m making the binaries available here for download. Binaries are 32-bit (will also run on 64-bit Windows) statically linked multi-threaded. I’ve used fftw version 3.3.2 that was kindly precompiled by David Svoboda. Thank you, David. Thanks also go to el_supremo for providing clear and useful instructions on how to compile ImageMagick in Visual Studio.
Without further ado, here’s the link to the binaries: http://blog.astrophotographytargets.com/pictures/imagemagick-with-fftw.zip
To make sure that it all works correctly I ran the FFT on the moon picture from previous post, and then ran the inverse transform on resulting two images. Recombined image looked just fine.
Due to a combination of factors (environmental, personal and work-related) an amazing piece of optics that Celestron 1100HD is has been gathering dust for a few months. I finally had a chance to take it for a spin, but as luck would have it Moon was high and bright so I had little choice but to image the Moon. Here’s how it came out. At 2800mm Moon is large enough not to fit within 35mm sensor so I had to glue two shots together. You can download full version (13 megabytes) here.
For your convenience I’ve also separated out soutern pole region since it had the most intetesting and contrast features.
…
 M76 Little Dumbbell Nebula Final image taken with my CG5 mount. After that it plain refuses to track accurately enough for astrophotography.
- Target: Messier 76 (aka Little Dumbbell Nebula, NGC650, NGC651)
- Apparent magnitude: 10.1
- Apparent dimensions: 3′ x 2′
- Date: 9/03/2011
- Location: Woodinville, WA (Red light pollution level)
- Conditions: Average transparency, Average seeing
- Equipment: Celestron C8-A XLT (2032mm f/10), Celestron CG5, Canon EOS 5D mark II, Orion Starshoot USB Autoguider, Orion 162mm f/3.2 Mini-Guidescope, Baader Planetarium Skyglow Filter
- Software: DeepSkyStacker 3.3.2, Adobe Lightroom 3, Adobe Photoshop CS5
- Images: Lights – 4 300 second (ISO1600), 9 darks, no flats and offsets.
 M71 Globular Cluster in Sagitta Another picture from my first night out with 1100 EdgeHD. This one is a single frame of Messier 71 globular cluster in the constellation of Sagitta.
 M57 Ring Nebula - HD 1100 M57, The Ring Nebula is very easy to find in the constellation of Lyra. It is high in the Summer sky and is compact and bright. So it is an easy target even for binoculars. These two images are taken with two different setups. First one is more recent. It is the first picture I took with my new Celestron CGEM DX/EdgeHD 1100 setup. Being a decent picture, it isn’t anything to write home about but since it is “first light” from a very nice scope that I like it is special to me.
- Target: Messier 57 (aka Ring Nebula, NGC6720)
- Apparent magnitude: 8.8
- Apparent dimensions: 4′ x 4′
- Date: 9/28/2011
- Location: Woodinville, WA (Red light pollution level)
- Conditions: Average transparency, Average seeing
- Equipment: Celestron EdgeHD 1100 (2800mm f/10), Celestron CGEM DX, Canon EOS 5D mark II, Orion Starshoot USB Autoguider, Orion 162mm f/3.2 Mini-Guidescope, Astronomik CLS LP filter.
- Software: Deep Sky Stacker 3.3.2, Adobe Lightroom 3, Adobe Photoshop CS5
- Images: Lights – 4 x 600 second (ISO800), no darks, flats and offsets.
Second image was taken about 1 month prior with my old C8/CG5 setup just as the CG5 mount was failing. It developed a significant wobble in the Dec axis to the point of being unusable for astrophotography shortly after. I decided to switch to CGEM DX but still indend to use C8 scope when I’m in need of shorter focal length.
 M57 Ring Nebula - C8/CG5
- Date: 9/03/2011
- Location: Woodinville, WA (Red light pollution level)
- Conditions: Below average transparency, below average visibility
- Equipment: Celestron C8-A XLT (2032mm f/10), Celestron CG5, Canon EOS 5D mark II, Orion Starshoot USB Autoguider, Orion 162mm f/3.2 Mini-Guidescope, Baader Planetarium Skyglow Filter
- Software: DeepSkyStacker 3.3.2, Adobe Lightroom 3, Adobe Photoshop CS5
- Images: Lights – 4 300 second (ISO1600), 9 darks, no flats and offsets.
|
|
Today I took some shots of the Sun while it’s not covered by clouds. To really see details on surface and solar prominence, special H-a (Hydrogen Alpha) telescope is needed. Most affordable one is Coronado P.S.T. (Personal solar telescope). Unfortunately it doesn’t allow to take prime focus photographs because of extremely short focus.
